| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.002102738690980681, |
| "eval_steps": 500, |
| "global_step": 15000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0, |
| "eval_loss": 10.238426208496094, |
| "eval_runtime": 111.932, |
| "eval_samples_per_second": 204.079, |
| "eval_steps_per_second": 1.599, |
| "step": 0 |
| }, |
| { |
| "epoch": 1.4018257939871205e-06, |
| "grad_norm": 235.0, |
| "learning_rate": 2.6999999999999996e-06, |
| "loss": 10.1619, |
| "step": 10 |
| }, |
| { |
| "epoch": 2.803651587974241e-06, |
| "grad_norm": 187.0, |
| "learning_rate": 5.7e-06, |
| "loss": 10.0993, |
| "step": 20 |
| }, |
| { |
| "epoch": 4.205477381961362e-06, |
| "grad_norm": 203.0, |
| "learning_rate": 8.7e-06, |
| "loss": 9.9129, |
| "step": 30 |
| }, |
| { |
| "epoch": 5.607303175948482e-06, |
| "grad_norm": 178.0, |
| "learning_rate": 1.17e-05, |
| "loss": 9.5407, |
| "step": 40 |
| }, |
| { |
| "epoch": 7.009128969935603e-06, |
| "grad_norm": 230.0, |
| "learning_rate": 1.47e-05, |
| "loss": 9.1468, |
| "step": 50 |
| }, |
| { |
| "epoch": 8.410954763922724e-06, |
| "grad_norm": 290.0, |
| "learning_rate": 1.7699999999999997e-05, |
| "loss": 8.5916, |
| "step": 60 |
| }, |
| { |
| "epoch": 9.812780557909844e-06, |
| "grad_norm": 398.0, |
| "learning_rate": 2.07e-05, |
| "loss": 8.1624, |
| "step": 70 |
| }, |
| { |
| "epoch": 1.1214606351896964e-05, |
| "grad_norm": 340.0, |
| "learning_rate": 2.3699999999999997e-05, |
| "loss": 8.0425, |
| "step": 80 |
| }, |
| { |
| "epoch": 1.2616432145884086e-05, |
| "grad_norm": 368.0, |
| "learning_rate": 2.6699999999999995e-05, |
| "loss": 7.2555, |
| "step": 90 |
| }, |
| { |
| "epoch": 1.4018257939871206e-05, |
| "grad_norm": 346.0, |
| "learning_rate": 2.97e-05, |
| "loss": 7.3853, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.5420083733858326e-05, |
| "grad_norm": 167.0, |
| "learning_rate": 3.2699999999999995e-05, |
| "loss": 6.8838, |
| "step": 110 |
| }, |
| { |
| "epoch": 1.6821909527845447e-05, |
| "grad_norm": 292.0, |
| "learning_rate": 3.5699999999999994e-05, |
| "loss": 6.7212, |
| "step": 120 |
| }, |
| { |
| "epoch": 1.822373532183257e-05, |
| "grad_norm": 253.0, |
| "learning_rate": 3.87e-05, |
| "loss": 6.3265, |
| "step": 130 |
| }, |
| { |
| "epoch": 1.9625561115819688e-05, |
| "grad_norm": 133.0, |
| "learning_rate": 4.17e-05, |
| "loss": 6.2292, |
| "step": 140 |
| }, |
| { |
| "epoch": 2.102738690980681e-05, |
| "grad_norm": 79.5, |
| "learning_rate": 4.4699999999999996e-05, |
| "loss": 5.1204, |
| "step": 150 |
| }, |
| { |
| "epoch": 2.2429212703793928e-05, |
| "grad_norm": 38.5, |
| "learning_rate": 4.7699999999999994e-05, |
| "loss": 4.9199, |
| "step": 160 |
| }, |
| { |
| "epoch": 2.383103849778105e-05, |
| "grad_norm": 27.875, |
| "learning_rate": 5.07e-05, |
| "loss": 4.1897, |
| "step": 170 |
| }, |
| { |
| "epoch": 2.523286429176817e-05, |
| "grad_norm": 23.375, |
| "learning_rate": 5.369999999999999e-05, |
| "loss": 3.1582, |
| "step": 180 |
| }, |
| { |
| "epoch": 2.663469008575529e-05, |
| "grad_norm": 11.75, |
| "learning_rate": 5.6699999999999996e-05, |
| "loss": 2.0433, |
| "step": 190 |
| }, |
| { |
| "epoch": 2.803651587974241e-05, |
| "grad_norm": 4.625, |
| "learning_rate": 5.97e-05, |
| "loss": 1.3245, |
| "step": 200 |
| }, |
| { |
| "epoch": 2.9438341673729533e-05, |
| "grad_norm": 3.1875, |
| "learning_rate": 6.269999999999999e-05, |
| "loss": 1.1166, |
| "step": 210 |
| }, |
| { |
| "epoch": 3.084016746771665e-05, |
| "grad_norm": 1.40625, |
| "learning_rate": 6.57e-05, |
| "loss": 0.884, |
| "step": 220 |
| }, |
| { |
| "epoch": 3.224199326170377e-05, |
| "grad_norm": 2.40625, |
| "learning_rate": 6.87e-05, |
| "loss": 0.8963, |
| "step": 230 |
| }, |
| { |
| "epoch": 3.3643819055690895e-05, |
| "grad_norm": 1.8125, |
| "learning_rate": 7.17e-05, |
| "loss": 0.9096, |
| "step": 240 |
| }, |
| { |
| "epoch": 3.504564484967802e-05, |
| "grad_norm": 0.9921875, |
| "learning_rate": 7.47e-05, |
| "loss": 0.8575, |
| "step": 250 |
| }, |
| { |
| "epoch": 3.644747064366514e-05, |
| "grad_norm": 1.4375, |
| "learning_rate": 7.769999999999999e-05, |
| "loss": 0.791, |
| "step": 260 |
| }, |
| { |
| "epoch": 3.7849296437652253e-05, |
| "grad_norm": 1.046875, |
| "learning_rate": 8.07e-05, |
| "loss": 0.8676, |
| "step": 270 |
| }, |
| { |
| "epoch": 3.9251122231639375e-05, |
| "grad_norm": 1.5859375, |
| "learning_rate": 8.37e-05, |
| "loss": 0.8645, |
| "step": 280 |
| }, |
| { |
| "epoch": 4.06529480256265e-05, |
| "grad_norm": 0.84765625, |
| "learning_rate": 8.669999999999998e-05, |
| "loss": 0.7618, |
| "step": 290 |
| }, |
| { |
| "epoch": 4.205477381961362e-05, |
| "grad_norm": 1.1640625, |
| "learning_rate": 8.969999999999998e-05, |
| "loss": 0.8603, |
| "step": 300 |
| }, |
| { |
| "epoch": 4.345659961360074e-05, |
| "grad_norm": 0.64453125, |
| "learning_rate": 9.269999999999999e-05, |
| "loss": 0.7988, |
| "step": 310 |
| }, |
| { |
| "epoch": 4.4858425407587855e-05, |
| "grad_norm": 0.94921875, |
| "learning_rate": 9.57e-05, |
| "loss": 0.8073, |
| "step": 320 |
| }, |
| { |
| "epoch": 4.626025120157498e-05, |
| "grad_norm": 0.76171875, |
| "learning_rate": 9.87e-05, |
| "loss": 0.8298, |
| "step": 330 |
| }, |
| { |
| "epoch": 4.76620769955621e-05, |
| "grad_norm": 1.5390625, |
| "learning_rate": 0.00010169999999999999, |
| "loss": 0.6355, |
| "step": 340 |
| }, |
| { |
| "epoch": 4.906390278954922e-05, |
| "grad_norm": 1.7890625, |
| "learning_rate": 0.00010469999999999998, |
| "loss": 0.848, |
| "step": 350 |
| }, |
| { |
| "epoch": 5.046572858353634e-05, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.00010769999999999999, |
| "loss": 0.5291, |
| "step": 360 |
| }, |
| { |
| "epoch": 5.1867554377523464e-05, |
| "grad_norm": 1.578125, |
| "learning_rate": 0.0001107, |
| "loss": 0.6325, |
| "step": 370 |
| }, |
| { |
| "epoch": 5.326938017151058e-05, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.00011369999999999999, |
| "loss": 0.7049, |
| "step": 380 |
| }, |
| { |
| "epoch": 5.46712059654977e-05, |
| "grad_norm": 1.109375, |
| "learning_rate": 0.00011669999999999999, |
| "loss": 0.7385, |
| "step": 390 |
| }, |
| { |
| "epoch": 5.607303175948482e-05, |
| "grad_norm": 1.0078125, |
| "learning_rate": 0.0001197, |
| "loss": 0.7195, |
| "step": 400 |
| }, |
| { |
| "epoch": 5.7474857553471944e-05, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.00012269999999999997, |
| "loss": 0.7844, |
| "step": 410 |
| }, |
| { |
| "epoch": 5.8876683347459066e-05, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.0001257, |
| "loss": 0.677, |
| "step": 420 |
| }, |
| { |
| "epoch": 6.027850914144619e-05, |
| "grad_norm": 1.703125, |
| "learning_rate": 0.00012869999999999998, |
| "loss": 0.704, |
| "step": 430 |
| }, |
| { |
| "epoch": 6.16803349354333e-05, |
| "grad_norm": 1.21875, |
| "learning_rate": 0.00013169999999999998, |
| "loss": 0.5956, |
| "step": 440 |
| }, |
| { |
| "epoch": 6.308216072942043e-05, |
| "grad_norm": 1.015625, |
| "learning_rate": 0.0001347, |
| "loss": 0.769, |
| "step": 450 |
| }, |
| { |
| "epoch": 6.448398652340755e-05, |
| "grad_norm": 1.1484375, |
| "learning_rate": 0.00013769999999999999, |
| "loss": 0.6908, |
| "step": 460 |
| }, |
| { |
| "epoch": 6.588581231739466e-05, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.00014069999999999998, |
| "loss": 0.6306, |
| "step": 470 |
| }, |
| { |
| "epoch": 6.728763811138179e-05, |
| "grad_norm": 0.9765625, |
| "learning_rate": 0.00014369999999999997, |
| "loss": 0.6421, |
| "step": 480 |
| }, |
| { |
| "epoch": 6.86894639053689e-05, |
| "grad_norm": 1.71875, |
| "learning_rate": 0.0001467, |
| "loss": 0.7901, |
| "step": 490 |
| }, |
| { |
| "epoch": 7.009128969935603e-05, |
| "grad_norm": 1.6640625, |
| "learning_rate": 0.00014969999999999998, |
| "loss": 0.6785, |
| "step": 500 |
| }, |
| { |
| "epoch": 7.009128969935603e-05, |
| "eval_loss": 0.27128955721855164, |
| "eval_runtime": 113.5365, |
| "eval_samples_per_second": 201.195, |
| "eval_steps_per_second": 1.577, |
| "step": 500 |
| }, |
| { |
| "epoch": 7.149311549334315e-05, |
| "grad_norm": 1.0859375, |
| "learning_rate": 0.0001527, |
| "loss": 0.6427, |
| "step": 510 |
| }, |
| { |
| "epoch": 7.289494128733028e-05, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.0001557, |
| "loss": 0.7077, |
| "step": 520 |
| }, |
| { |
| "epoch": 7.429676708131739e-05, |
| "grad_norm": 1.3203125, |
| "learning_rate": 0.00015869999999999998, |
| "loss": 0.6865, |
| "step": 530 |
| }, |
| { |
| "epoch": 7.569859287530451e-05, |
| "grad_norm": 1.140625, |
| "learning_rate": 0.0001617, |
| "loss": 0.7239, |
| "step": 540 |
| }, |
| { |
| "epoch": 7.710041866929164e-05, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.0001647, |
| "loss": 0.6667, |
| "step": 550 |
| }, |
| { |
| "epoch": 7.850224446327875e-05, |
| "grad_norm": 1.34375, |
| "learning_rate": 0.0001677, |
| "loss": 0.8085, |
| "step": 560 |
| }, |
| { |
| "epoch": 7.990407025726588e-05, |
| "grad_norm": 1.3046875, |
| "learning_rate": 0.00017069999999999998, |
| "loss": 0.7019, |
| "step": 570 |
| }, |
| { |
| "epoch": 8.1305896051253e-05, |
| "grad_norm": 1.796875, |
| "learning_rate": 0.00017369999999999997, |
| "loss": 0.7378, |
| "step": 580 |
| }, |
| { |
| "epoch": 8.270772184524011e-05, |
| "grad_norm": 0.98828125, |
| "learning_rate": 0.00017669999999999999, |
| "loss": 0.6776, |
| "step": 590 |
| }, |
| { |
| "epoch": 8.410954763922724e-05, |
| "grad_norm": 1.2578125, |
| "learning_rate": 0.00017969999999999998, |
| "loss": 0.6166, |
| "step": 600 |
| }, |
| { |
| "epoch": 8.551137343321435e-05, |
| "grad_norm": 0.75, |
| "learning_rate": 0.00018269999999999997, |
| "loss": 0.5741, |
| "step": 610 |
| }, |
| { |
| "epoch": 8.691319922720148e-05, |
| "grad_norm": 1.1484375, |
| "learning_rate": 0.0001857, |
| "loss": 0.7397, |
| "step": 620 |
| }, |
| { |
| "epoch": 8.83150250211886e-05, |
| "grad_norm": 1.265625, |
| "learning_rate": 0.00018869999999999998, |
| "loss": 0.6753, |
| "step": 630 |
| }, |
| { |
| "epoch": 8.971685081517571e-05, |
| "grad_norm": 0.9375, |
| "learning_rate": 0.0001917, |
| "loss": 0.689, |
| "step": 640 |
| }, |
| { |
| "epoch": 9.111867660916284e-05, |
| "grad_norm": 1.1484375, |
| "learning_rate": 0.0001947, |
| "loss": 0.5332, |
| "step": 650 |
| }, |
| { |
| "epoch": 9.252050240314995e-05, |
| "grad_norm": 1.234375, |
| "learning_rate": 0.00019769999999999998, |
| "loss": 0.7178, |
| "step": 660 |
| }, |
| { |
| "epoch": 9.392232819713708e-05, |
| "grad_norm": 0.8203125, |
| "learning_rate": 0.0002007, |
| "loss": 0.5924, |
| "step": 670 |
| }, |
| { |
| "epoch": 9.53241539911242e-05, |
| "grad_norm": 1.5, |
| "learning_rate": 0.0002037, |
| "loss": 0.6734, |
| "step": 680 |
| }, |
| { |
| "epoch": 9.672597978511133e-05, |
| "grad_norm": 0.9921875, |
| "learning_rate": 0.00020669999999999996, |
| "loss": 0.6536, |
| "step": 690 |
| }, |
| { |
| "epoch": 9.812780557909844e-05, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.00020969999999999997, |
| "loss": 0.5084, |
| "step": 700 |
| }, |
| { |
| "epoch": 9.952963137308556e-05, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.00021269999999999997, |
| "loss": 0.7349, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.00010093145716707268, |
| "grad_norm": 0.6875, |
| "learning_rate": 0.00021569999999999998, |
| "loss": 0.519, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.0001023332829610598, |
| "grad_norm": 1.3671875, |
| "learning_rate": 0.00021869999999999998, |
| "loss": 0.6678, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.00010373510875504693, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00022169999999999997, |
| "loss": 0.6112, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.00010513693454903404, |
| "grad_norm": 1.09375, |
| "learning_rate": 0.0002247, |
| "loss": 0.6505, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.00010653876034302116, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.00022769999999999998, |
| "loss": 0.6256, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.00010794058613700829, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.0002307, |
| "loss": 0.6422, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.0001093424119309954, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.0002337, |
| "loss": 0.6357, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.00011074423772498253, |
| "grad_norm": 1.25, |
| "learning_rate": 0.00023669999999999998, |
| "loss": 0.6263, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.00011214606351896965, |
| "grad_norm": 1.1171875, |
| "learning_rate": 0.0002397, |
| "loss": 0.6531, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.00011354788931295677, |
| "grad_norm": 1.1875, |
| "learning_rate": 0.0002427, |
| "loss": 0.5806, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.00011494971510694389, |
| "grad_norm": 1.1953125, |
| "learning_rate": 0.00024569999999999995, |
| "loss": 0.5925, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.000116351540900931, |
| "grad_norm": 1.1640625, |
| "learning_rate": 0.0002487, |
| "loss": 0.5465, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.00011775336669491813, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.0002517, |
| "loss": 0.7182, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.00011915519248890525, |
| "grad_norm": 1.3203125, |
| "learning_rate": 0.00025469999999999996, |
| "loss": 0.6767, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.00012055701828289238, |
| "grad_norm": 1.2890625, |
| "learning_rate": 0.0002577, |
| "loss": 0.6315, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.00012195884407687949, |
| "grad_norm": 0.71484375, |
| "learning_rate": 0.0002607, |
| "loss": 0.6724, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.0001233606698708666, |
| "grad_norm": 0.9765625, |
| "learning_rate": 0.00026369999999999996, |
| "loss": 0.6528, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.00012476249566485372, |
| "grad_norm": 1.15625, |
| "learning_rate": 0.0002667, |
| "loss": 0.5295, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.00012616432145884086, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.0002697, |
| "loss": 0.4982, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.00012756614725282798, |
| "grad_norm": 1.2734375, |
| "learning_rate": 0.00027269999999999996, |
| "loss": 0.6562, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.0001289679730468151, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.0002757, |
| "loss": 0.675, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.0001303697988408022, |
| "grad_norm": 1.2890625, |
| "learning_rate": 0.0002787, |
| "loss": 0.6327, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.00013177162463478932, |
| "grad_norm": 1.203125, |
| "learning_rate": 0.00028169999999999996, |
| "loss": 0.7426, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.00013317345042877646, |
| "grad_norm": 0.94921875, |
| "learning_rate": 0.0002847, |
| "loss": 0.5876, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.00013457527622276358, |
| "grad_norm": 0.625, |
| "learning_rate": 0.00028769999999999995, |
| "loss": 0.6968, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.0001359771020167507, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.00029069999999999996, |
| "loss": 0.6377, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.0001373789278107378, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.0002937, |
| "loss": 0.515, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.00013878075360472492, |
| "grad_norm": 0.95703125, |
| "learning_rate": 0.00029669999999999995, |
| "loss": 0.6966, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.00014018257939871207, |
| "grad_norm": 0.376953125, |
| "learning_rate": 0.00029969999999999997, |
| "loss": 0.5426, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.00014018257939871207, |
| "eval_loss": 0.2417471706867218, |
| "eval_runtime": 111.9274, |
| "eval_samples_per_second": 204.088, |
| "eval_steps_per_second": 1.599, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.00014158440519269918, |
| "grad_norm": 1.1015625, |
| "learning_rate": 0.0002999999999988214, |
| "loss": 0.579, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.0001429862309866863, |
| "grad_norm": 1.0625, |
| "learning_rate": 0.0002999999999947473, |
| "loss": 0.5958, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.0001443880567806734, |
| "grad_norm": 1.1953125, |
| "learning_rate": 0.0002999999999877632, |
| "loss": 0.5812, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.00014578988257466055, |
| "grad_norm": 1.1171875, |
| "learning_rate": 0.000299999999977869, |
| "loss": 0.6251, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.00014719170836864767, |
| "grad_norm": 1.125, |
| "learning_rate": 0.0002999999999650648, |
| "loss": 0.6301, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.00014859353416263478, |
| "grad_norm": 0.71484375, |
| "learning_rate": 0.0002999999999493505, |
| "loss": 0.6041, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.0001499953599566219, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.0002999999999307262, |
| "loss": 0.5958, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.00015139718575060901, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00029999999990919175, |
| "loss": 0.5158, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.00015279901154459616, |
| "grad_norm": 0.93359375, |
| "learning_rate": 0.00029999999988474733, |
| "loss": 0.5815, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.00015420083733858327, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.00029999999985739286, |
| "loss": 0.5338, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.00015560266313257039, |
| "grad_norm": 0.546875, |
| "learning_rate": 0.0002999999998271283, |
| "loss": 0.3756, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.0001570044889265575, |
| "grad_norm": 0.86328125, |
| "learning_rate": 0.0002999999997939537, |
| "loss": 0.5302, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.00015840631472054462, |
| "grad_norm": 0.625, |
| "learning_rate": 0.000299999999757869, |
| "loss": 0.6399, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.00015980814051453176, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.0002999999997188743, |
| "loss": 0.5165, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.00016120996630851887, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.0002999999996769696, |
| "loss": 0.611, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.000162611792102506, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.00029999999963215473, |
| "loss": 0.4649, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.0001640136178964931, |
| "grad_norm": 0.474609375, |
| "learning_rate": 0.0002999999995844299, |
| "loss": 0.5931, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.00016541544369048022, |
| "grad_norm": 1.515625, |
| "learning_rate": 0.000299999999533795, |
| "loss": 0.605, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.00016681726948446736, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.00029999999948025, |
| "loss": 0.5483, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.00016821909527845447, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.000299999999423795, |
| "loss": 0.5798, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.0001696209210724416, |
| "grad_norm": 0.93359375, |
| "learning_rate": 0.0002999999993644299, |
| "loss": 0.5735, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.0001710227468664287, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.00029999999930215476, |
| "loss": 0.5838, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.00017242457266041582, |
| "grad_norm": 0.44921875, |
| "learning_rate": 0.0002999999992369696, |
| "loss": 0.495, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.00017382639845440296, |
| "grad_norm": 0.8984375, |
| "learning_rate": 0.00029999999916887433, |
| "loss": 0.576, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.00017522822424839008, |
| "grad_norm": 1.015625, |
| "learning_rate": 0.00029999999909786906, |
| "loss": 0.5682, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.0001766300500423772, |
| "grad_norm": 0.91796875, |
| "learning_rate": 0.00029999999902395374, |
| "loss": 0.609, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.0001780318758363643, |
| "grad_norm": 0.9609375, |
| "learning_rate": 0.0002999999989471283, |
| "loss": 0.5911, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.00017943370163035142, |
| "grad_norm": 1.078125, |
| "learning_rate": 0.0002999999988673929, |
| "loss": 0.6365, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.00018083552742433856, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.00029999999878474736, |
| "loss": 0.4741, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.00018223735321832568, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.0002999999986991918, |
| "loss": 0.6288, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.0001836391790123128, |
| "grad_norm": 0.5, |
| "learning_rate": 0.0002999999986107262, |
| "loss": 0.4648, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.0001850410048062999, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.00029999999851935055, |
| "loss": 0.5761, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.00018644283060028705, |
| "grad_norm": 1.4375, |
| "learning_rate": 0.00029999999842506486, |
| "loss": 0.5775, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.00018784465639427417, |
| "grad_norm": 0.96875, |
| "learning_rate": 0.00029999999832786906, |
| "loss": 0.5382, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.00018924648218826128, |
| "grad_norm": 1.078125, |
| "learning_rate": 0.00029999999822776327, |
| "loss": 0.5922, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.0001906483079822484, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.0002999999981247474, |
| "loss": 0.4537, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.0001920501337762355, |
| "grad_norm": 1.328125, |
| "learning_rate": 0.00029999999801882146, |
| "loss": 0.5119, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.00019345195957022265, |
| "grad_norm": 0.86328125, |
| "learning_rate": 0.0002999999979099855, |
| "loss": 0.478, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.00019485378536420977, |
| "grad_norm": 1.9296875, |
| "learning_rate": 0.00029999999779823944, |
| "loss": 0.4734, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.00019625561115819688, |
| "grad_norm": 1.2421875, |
| "learning_rate": 0.0002999999976835834, |
| "loss": 0.4924, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.000197657436952184, |
| "grad_norm": 2.046875, |
| "learning_rate": 0.00029999999756601727, |
| "loss": 0.5205, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.0001990592627461711, |
| "grad_norm": 1.3984375, |
| "learning_rate": 0.0002999999974455411, |
| "loss": 0.5122, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.00020046108854015825, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999999732215483, |
| "loss": 0.5492, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.00020186291433414537, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.00029999999719585856, |
| "loss": 0.5266, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.00020326474012813248, |
| "grad_norm": 0.85546875, |
| "learning_rate": 0.00029999999706665224, |
| "loss": 0.5307, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.0002046665659221196, |
| "grad_norm": 0.53515625, |
| "learning_rate": 0.0002999999969345358, |
| "loss": 0.6882, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.00020606839171610671, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.0002999999967995094, |
| "loss": 0.5092, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.00020747021751009386, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.0002999999966615729, |
| "loss": 0.4554, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.00020887204330408097, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.0002999999965207263, |
| "loss": 0.5944, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.0002102738690980681, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.0002999999963769697, |
| "loss": 0.5038, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.0002102738690980681, |
| "eval_loss": 0.22094833850860596, |
| "eval_runtime": 112.0927, |
| "eval_samples_per_second": 203.787, |
| "eval_steps_per_second": 1.597, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.0002116756948920552, |
| "grad_norm": 0.8125, |
| "learning_rate": 0.00029999999623030307, |
| "loss": 0.5026, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.00021307752068604232, |
| "grad_norm": 0.390625, |
| "learning_rate": 0.0002999999960807263, |
| "loss": 0.5545, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.00021447934648002946, |
| "grad_norm": 1.109375, |
| "learning_rate": 0.0002999999959282396, |
| "loss": 0.5133, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.00021588117227401657, |
| "grad_norm": 1.2578125, |
| "learning_rate": 0.0002999999957728428, |
| "loss": 0.5241, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.0002172829980680037, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.00029999999561453587, |
| "loss": 0.4591, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.0002186848238619908, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.00029999999545331896, |
| "loss": 0.5348, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.00022008664965597792, |
| "grad_norm": 1.171875, |
| "learning_rate": 0.00029999999528919195, |
| "loss": 0.5408, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.00022148847544996506, |
| "grad_norm": 1.1171875, |
| "learning_rate": 0.000299999995122155, |
| "loss": 0.616, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.00022289030124395218, |
| "grad_norm": 1.53125, |
| "learning_rate": 0.0002999999949522079, |
| "loss": 0.585, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.0002242921270379393, |
| "grad_norm": 0.373046875, |
| "learning_rate": 0.0002999999947793507, |
| "loss": 0.4513, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.0002256939528319264, |
| "grad_norm": 1.0078125, |
| "learning_rate": 0.0002999999946035836, |
| "loss": 0.5646, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.00022709577862591355, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.0002999999944249063, |
| "loss": 0.4162, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.00022849760441990066, |
| "grad_norm": 0.462890625, |
| "learning_rate": 0.00029999999424331903, |
| "loss": 0.5288, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.00022989943021388778, |
| "grad_norm": 0.26953125, |
| "learning_rate": 0.0002999999940588217, |
| "loss": 0.5398, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.0002313012560078749, |
| "grad_norm": 1.109375, |
| "learning_rate": 0.00029999999387141427, |
| "loss": 0.5167, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.000232703081801862, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.00029999999368109684, |
| "loss": 0.5953, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.00023410490759584915, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.00029999999348786935, |
| "loss": 0.5765, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.00023550673338983626, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.00029999999329173176, |
| "loss": 0.5683, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.00023690855918382338, |
| "grad_norm": 0.8203125, |
| "learning_rate": 0.00029999999309268416, |
| "loss": 0.4832, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.0002383103849778105, |
| "grad_norm": 0.369140625, |
| "learning_rate": 0.0002999999928907265, |
| "loss": 0.4763, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.0002397122107717976, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.0002999999926858588, |
| "loss": 0.5735, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.00024111403656578475, |
| "grad_norm": 1.078125, |
| "learning_rate": 0.000299999992478081, |
| "loss": 0.5352, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.00024251586235977187, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.0002999999922673932, |
| "loss": 0.5497, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.00024391768815375898, |
| "grad_norm": 0.310546875, |
| "learning_rate": 0.00029999999205379535, |
| "loss": 0.5335, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.0002453195139477461, |
| "grad_norm": 1.03125, |
| "learning_rate": 0.0002999999918372874, |
| "loss": 0.476, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.0002467213397417332, |
| "grad_norm": 0.86328125, |
| "learning_rate": 0.0002999999916178694, |
| "loss": 0.5105, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.0002481231655357203, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.0002999999913955414, |
| "loss": 0.5929, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.00024952499132970744, |
| "grad_norm": 0.419921875, |
| "learning_rate": 0.00029999999117030334, |
| "loss": 0.4957, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.0002509268171236946, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.0002999999909421552, |
| "loss": 0.5539, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.0002523286429176817, |
| "grad_norm": 0.96484375, |
| "learning_rate": 0.000299999990711097, |
| "loss": 0.4833, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.00025373046871166884, |
| "grad_norm": 0.55078125, |
| "learning_rate": 0.00029999999047712877, |
| "loss": 0.5147, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.00025513229450565596, |
| "grad_norm": 1.3671875, |
| "learning_rate": 0.0002999999902402505, |
| "loss": 0.4975, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.00025653412029964307, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.0002999999900004621, |
| "loss": 0.5936, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.0002579359460936302, |
| "grad_norm": 1.25, |
| "learning_rate": 0.0002999999897577637, |
| "loss": 0.5696, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.0002593377718876173, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.0002999999895121553, |
| "loss": 0.507, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.0002607395976816044, |
| "grad_norm": 0.37890625, |
| "learning_rate": 0.0002999999892636368, |
| "loss": 0.4285, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.00026214142347559153, |
| "grad_norm": 1.1328125, |
| "learning_rate": 0.00029999998901220824, |
| "loss": 0.5559, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.00026354324926957865, |
| "grad_norm": 0.57421875, |
| "learning_rate": 0.0002999999887578696, |
| "loss": 0.5271, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.0002649450750635658, |
| "grad_norm": 0.484375, |
| "learning_rate": 0.00029999998850062094, |
| "loss": 0.4832, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.00026634690085755293, |
| "grad_norm": 1.328125, |
| "learning_rate": 0.00029999998824046224, |
| "loss": 0.6106, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.00026774872665154004, |
| "grad_norm": 1.828125, |
| "learning_rate": 0.0002999999879773935, |
| "loss": 0.5356, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.00026915055244552716, |
| "grad_norm": 1.2265625, |
| "learning_rate": 0.0002999999877114147, |
| "loss": 0.5922, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.0002705523782395143, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.00029999998744252576, |
| "loss": 0.4454, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.0002719542040335014, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.00029999998717072685, |
| "loss": 0.4555, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.0002733560298274885, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.0002999999868960179, |
| "loss": 0.5104, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.0002747578556214756, |
| "grad_norm": 0.9140625, |
| "learning_rate": 0.00029999998661839886, |
| "loss": 0.5298, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.00027615968141546273, |
| "grad_norm": 0.3828125, |
| "learning_rate": 0.0002999999863378698, |
| "loss": 0.511, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.00027756150720944985, |
| "grad_norm": 0.98046875, |
| "learning_rate": 0.00029999998605443067, |
| "loss": 0.5218, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.000278963333003437, |
| "grad_norm": 1.1796875, |
| "learning_rate": 0.0002999999857680815, |
| "loss": 0.5483, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.00028036515879742413, |
| "grad_norm": 0.494140625, |
| "learning_rate": 0.00029999998547882226, |
| "loss": 0.5184, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.00028036515879742413, |
| "eval_loss": 0.2110401690006256, |
| "eval_runtime": 113.7235, |
| "eval_samples_per_second": 200.864, |
| "eval_steps_per_second": 1.574, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.00028176698459141125, |
| "grad_norm": 0.76171875, |
| "learning_rate": 0.00029999998518665297, |
| "loss": 0.5142, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.00028316881038539836, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.0002999999848915736, |
| "loss": 0.6032, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.0002845706361793855, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.0002999999845935842, |
| "loss": 0.5468, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.0002859724619733726, |
| "grad_norm": 0.380859375, |
| "learning_rate": 0.00029999998429268475, |
| "loss": 0.4213, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.0002873742877673597, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.00029999998398887526, |
| "loss": 0.4999, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.0002887761135613468, |
| "grad_norm": 1.0703125, |
| "learning_rate": 0.0002999999836821557, |
| "loss": 0.4975, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.00029017793935533394, |
| "grad_norm": 0.396484375, |
| "learning_rate": 0.0002999999833725261, |
| "loss": 0.4618, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.0002915797651493211, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.00029999998305998645, |
| "loss": 0.6292, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.0002929815909433082, |
| "grad_norm": 1.1015625, |
| "learning_rate": 0.0002999999827445367, |
| "loss": 0.4893, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.00029438341673729534, |
| "grad_norm": 0.921875, |
| "learning_rate": 0.000299999982426177, |
| "loss": 0.4749, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.00029578524253128245, |
| "grad_norm": 1.21875, |
| "learning_rate": 0.0002999999821049071, |
| "loss": 0.5639, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.00029718706832526957, |
| "grad_norm": 0.275390625, |
| "learning_rate": 0.00029999998178072726, |
| "loss": 0.4799, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.0002985888941192567, |
| "grad_norm": 1.1875, |
| "learning_rate": 0.00029999998145363734, |
| "loss": 0.5267, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.0002999907199132438, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.00029999998112363737, |
| "loss": 0.5218, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.0003013925457072309, |
| "grad_norm": 0.98046875, |
| "learning_rate": 0.00029999998079072735, |
| "loss": 0.5824, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.00030279437150121803, |
| "grad_norm": 0.3359375, |
| "learning_rate": 0.00029999998045490727, |
| "loss": 0.5229, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.00030419619729520514, |
| "grad_norm": 0.490234375, |
| "learning_rate": 0.00029999998011617714, |
| "loss": 0.42, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.0003055980230891923, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.00029999997977453696, |
| "loss": 0.5678, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.0003069998488831794, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.0002999999794299867, |
| "loss": 0.5489, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.00030840167467716654, |
| "grad_norm": 0.40625, |
| "learning_rate": 0.00029999997908252644, |
| "loss": 0.4916, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.00030980350047115366, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.0002999999787321561, |
| "loss": 0.4691, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.00031120532626514077, |
| "grad_norm": 0.71484375, |
| "learning_rate": 0.0002999999783788757, |
| "loss": 0.4792, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.0003126071520591279, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.00029999997802268526, |
| "loss": 0.4996, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.000314008977853115, |
| "grad_norm": 1.0625, |
| "learning_rate": 0.0002999999776635848, |
| "loss": 0.4695, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.0003154108036471021, |
| "grad_norm": 0.31640625, |
| "learning_rate": 0.00029999997730157426, |
| "loss": 0.4476, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.00031681262944108923, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.0002999999769366536, |
| "loss": 0.5829, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.00031821445523507635, |
| "grad_norm": 0.98046875, |
| "learning_rate": 0.000299999976568823, |
| "loss": 0.521, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.0003196162810290635, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.0002999999761980823, |
| "loss": 0.5269, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.00032101810682305063, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.00029999997582443147, |
| "loss": 0.5492, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.00032241993261703775, |
| "grad_norm": 0.466796875, |
| "learning_rate": 0.00029999997544787066, |
| "loss": 0.5568, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.00032382175841102486, |
| "grad_norm": 0.2734375, |
| "learning_rate": 0.00029999997506839984, |
| "loss": 0.5229, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.000325223584205012, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.0002999999746860189, |
| "loss": 0.4528, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.0003266254099989991, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.00029999997430072794, |
| "loss": 0.5159, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.0003280272357929862, |
| "grad_norm": 0.640625, |
| "learning_rate": 0.0002999999739125269, |
| "loss": 0.4858, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.0003294290615869733, |
| "grad_norm": 1.109375, |
| "learning_rate": 0.00029999997352141584, |
| "loss": 0.5768, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.00033083088738096044, |
| "grad_norm": 0.859375, |
| "learning_rate": 0.0002999999731273947, |
| "loss": 0.5316, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.0003322327131749476, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.0002999999727304635, |
| "loss": 0.4754, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.0003336345389689347, |
| "grad_norm": 0.86328125, |
| "learning_rate": 0.00029999997233062233, |
| "loss": 0.5883, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.00033503636476292183, |
| "grad_norm": 0.341796875, |
| "learning_rate": 0.00029999997192787104, |
| "loss": 0.5788, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.00033643819055690895, |
| "grad_norm": 0.376953125, |
| "learning_rate": 0.0002999999715222097, |
| "loss": 0.4608, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.00033784001635089606, |
| "grad_norm": 0.427734375, |
| "learning_rate": 0.00029999997111363836, |
| "loss": 0.4271, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.0003392418421448832, |
| "grad_norm": 0.984375, |
| "learning_rate": 0.0002999999707021569, |
| "loss": 0.5311, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.0003406436679388703, |
| "grad_norm": 0.71484375, |
| "learning_rate": 0.00029999997028776535, |
| "loss": 0.528, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.0003420454937328574, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.00029999996987046385, |
| "loss": 0.5641, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.0003434473195268445, |
| "grad_norm": 1.0078125, |
| "learning_rate": 0.00029999996945025224, |
| "loss": 0.597, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.00034484914532083164, |
| "grad_norm": 1.1015625, |
| "learning_rate": 0.00029999996902713063, |
| "loss": 0.4643, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.0003462509711148188, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.0002999999686010989, |
| "loss": 0.3736, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.0003476527969088059, |
| "grad_norm": 0.439453125, |
| "learning_rate": 0.00029999996817215715, |
| "loss": 0.5049, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.00034905462270279304, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.00029999996774030533, |
| "loss": 0.3526, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.00035045644849678015, |
| "grad_norm": 0.91796875, |
| "learning_rate": 0.0002999999673055435, |
| "loss": 0.5538, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.00035045644849678015, |
| "eval_loss": 0.20391806960105896, |
| "eval_runtime": 112.0697, |
| "eval_samples_per_second": 203.829, |
| "eval_steps_per_second": 1.597, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.00035185827429076727, |
| "grad_norm": 0.353515625, |
| "learning_rate": 0.0002999999668678716, |
| "loss": 0.4682, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.0003532601000847544, |
| "grad_norm": 0.421875, |
| "learning_rate": 0.0002999999664272896, |
| "loss": 0.5603, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.0003546619258787415, |
| "grad_norm": 0.52734375, |
| "learning_rate": 0.0002999999659837976, |
| "loss": 0.4438, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.0003560637516727286, |
| "grad_norm": 0.328125, |
| "learning_rate": 0.00029999996553739555, |
| "loss": 0.4908, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.00035746557746671573, |
| "grad_norm": 1.046875, |
| "learning_rate": 0.0002999999650880834, |
| "loss": 0.4753, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.00035886740326070284, |
| "grad_norm": 0.99609375, |
| "learning_rate": 0.0002999999646358613, |
| "loss": 0.4611, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.00036026922905469, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.00029999996418072903, |
| "loss": 0.3998, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.00036167105484867713, |
| "grad_norm": 0.359375, |
| "learning_rate": 0.00029999996372268673, |
| "loss": 0.4206, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.00036307288064266424, |
| "grad_norm": 1.1484375, |
| "learning_rate": 0.00029999996326173444, |
| "loss": 0.5298, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.00036447470643665136, |
| "grad_norm": 0.255859375, |
| "learning_rate": 0.00029999996279787204, |
| "loss": 0.4437, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.00036587653223063847, |
| "grad_norm": 0.40625, |
| "learning_rate": 0.00029999996233109964, |
| "loss": 0.4519, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.0003672783580246256, |
| "grad_norm": 0.427734375, |
| "learning_rate": 0.00029999996186141713, |
| "loss": 0.3585, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.0003686801838186127, |
| "grad_norm": 0.486328125, |
| "learning_rate": 0.00029999996138882463, |
| "loss": 0.4805, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.0003700820096125998, |
| "grad_norm": 0.94140625, |
| "learning_rate": 0.000299999960913322, |
| "loss": 0.549, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.00037148383540658693, |
| "grad_norm": 1.1796875, |
| "learning_rate": 0.0002999999604349094, |
| "loss": 0.4753, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.0003728856612005741, |
| "grad_norm": 0.4609375, |
| "learning_rate": 0.00029999995995358674, |
| "loss": 0.5911, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.0003742874869945612, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.00029999995946935397, |
| "loss": 0.5762, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.00037568931278854833, |
| "grad_norm": 0.404296875, |
| "learning_rate": 0.0002999999589822112, |
| "loss": 0.5504, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.00037709113858253545, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.0002999999584921583, |
| "loss": 0.435, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.00037849296437652256, |
| "grad_norm": 1.3984375, |
| "learning_rate": 0.00029999995799919545, |
| "loss": 0.4766, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.0003798947901705097, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.00029999995750332246, |
| "loss": 0.5923, |
| "step": 2710 |
| }, |
| { |
| "epoch": 0.0003812966159644968, |
| "grad_norm": 0.48828125, |
| "learning_rate": 0.0002999999570045395, |
| "loss": 0.5589, |
| "step": 2720 |
| }, |
| { |
| "epoch": 0.0003826984417584839, |
| "grad_norm": 0.365234375, |
| "learning_rate": 0.00029999995650284645, |
| "loss": 0.5023, |
| "step": 2730 |
| }, |
| { |
| "epoch": 0.000384100267552471, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.0002999999559982433, |
| "loss": 0.5224, |
| "step": 2740 |
| }, |
| { |
| "epoch": 0.00038550209334645814, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.00029999995549073017, |
| "loss": 0.4, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.0003869039191404453, |
| "grad_norm": 0.396484375, |
| "learning_rate": 0.0002999999549803069, |
| "loss": 0.4416, |
| "step": 2760 |
| }, |
| { |
| "epoch": 0.0003883057449344324, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.0002999999544669737, |
| "loss": 0.4588, |
| "step": 2770 |
| }, |
| { |
| "epoch": 0.00038970757072841954, |
| "grad_norm": 0.6015625, |
| "learning_rate": 0.0002999999539507304, |
| "loss": 0.5128, |
| "step": 2780 |
| }, |
| { |
| "epoch": 0.00039110939652240665, |
| "grad_norm": 0.640625, |
| "learning_rate": 0.00029999995343157703, |
| "loss": 0.4519, |
| "step": 2790 |
| }, |
| { |
| "epoch": 0.00039251122231639377, |
| "grad_norm": 0.75, |
| "learning_rate": 0.00029999995290951357, |
| "loss": 0.553, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.0003939130481103809, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.0002999999523845401, |
| "loss": 0.4382, |
| "step": 2810 |
| }, |
| { |
| "epoch": 0.000395314873904368, |
| "grad_norm": 0.8984375, |
| "learning_rate": 0.0002999999518566566, |
| "loss": 0.6044, |
| "step": 2820 |
| }, |
| { |
| "epoch": 0.0003967166996983551, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.00029999995132586304, |
| "loss": 0.5411, |
| "step": 2830 |
| }, |
| { |
| "epoch": 0.0003981185254923422, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.0002999999507921594, |
| "loss": 0.48, |
| "step": 2840 |
| }, |
| { |
| "epoch": 0.00039952035128632934, |
| "grad_norm": 0.96484375, |
| "learning_rate": 0.00029999995025554576, |
| "loss": 0.5314, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.0004009221770803165, |
| "grad_norm": 1.0859375, |
| "learning_rate": 0.000299999949716022, |
| "loss": 0.5695, |
| "step": 2860 |
| }, |
| { |
| "epoch": 0.0004023240028743036, |
| "grad_norm": 1.0078125, |
| "learning_rate": 0.0002999999491735882, |
| "loss": 0.4795, |
| "step": 2870 |
| }, |
| { |
| "epoch": 0.00040372582866829074, |
| "grad_norm": 0.46484375, |
| "learning_rate": 0.0002999999486282444, |
| "loss": 0.3841, |
| "step": 2880 |
| }, |
| { |
| "epoch": 0.00040512765446227785, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.0002999999480799905, |
| "loss": 0.4925, |
| "step": 2890 |
| }, |
| { |
| "epoch": 0.00040652948025626497, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.00029999994752882657, |
| "loss": 0.4798, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.0004079313060502521, |
| "grad_norm": 0.9453125, |
| "learning_rate": 0.0002999999469747526, |
| "loss": 0.5165, |
| "step": 2910 |
| }, |
| { |
| "epoch": 0.0004093331318442392, |
| "grad_norm": 0.70703125, |
| "learning_rate": 0.00029999994641776855, |
| "loss": 0.5801, |
| "step": 2920 |
| }, |
| { |
| "epoch": 0.0004107349576382263, |
| "grad_norm": 0.3359375, |
| "learning_rate": 0.00029999994585787446, |
| "loss": 0.5142, |
| "step": 2930 |
| }, |
| { |
| "epoch": 0.00041213678343221343, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.0002999999452950703, |
| "loss": 0.4823, |
| "step": 2940 |
| }, |
| { |
| "epoch": 0.0004135386092262006, |
| "grad_norm": 0.294921875, |
| "learning_rate": 0.0002999999447293561, |
| "loss": 0.4351, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.0004149404350201877, |
| "grad_norm": 0.396484375, |
| "learning_rate": 0.0002999999441607319, |
| "loss": 0.5292, |
| "step": 2960 |
| }, |
| { |
| "epoch": 0.00041634226081417483, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999999435891976, |
| "loss": 0.458, |
| "step": 2970 |
| }, |
| { |
| "epoch": 0.00041774408660816194, |
| "grad_norm": 0.294921875, |
| "learning_rate": 0.00029999994301475327, |
| "loss": 0.4933, |
| "step": 2980 |
| }, |
| { |
| "epoch": 0.00041914591240214906, |
| "grad_norm": 1.140625, |
| "learning_rate": 0.0002999999424373988, |
| "loss": 0.4954, |
| "step": 2990 |
| }, |
| { |
| "epoch": 0.0004205477381961362, |
| "grad_norm": 1.5234375, |
| "learning_rate": 0.0002999999418571344, |
| "loss": 0.4878, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.0004205477381961362, |
| "eval_loss": 0.20102550089359283, |
| "eval_runtime": 112.0673, |
| "eval_samples_per_second": 203.833, |
| "eval_steps_per_second": 1.597, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.0004219495639901233, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.0002999999412739599, |
| "loss": 0.5205, |
| "step": 3010 |
| }, |
| { |
| "epoch": 0.0004233513897841104, |
| "grad_norm": 0.40625, |
| "learning_rate": 0.0002999999406878753, |
| "loss": 0.4147, |
| "step": 3020 |
| }, |
| { |
| "epoch": 0.0004247532155780975, |
| "grad_norm": 0.4921875, |
| "learning_rate": 0.0002999999400988807, |
| "loss": 0.424, |
| "step": 3030 |
| }, |
| { |
| "epoch": 0.00042615504137208463, |
| "grad_norm": 0.93359375, |
| "learning_rate": 0.000299999939506976, |
| "loss": 0.4885, |
| "step": 3040 |
| }, |
| { |
| "epoch": 0.0004275568671660718, |
| "grad_norm": 1.2265625, |
| "learning_rate": 0.00029999993891216136, |
| "loss": 0.517, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.0004289586929600589, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.0002999999383144366, |
| "loss": 0.5413, |
| "step": 3060 |
| }, |
| { |
| "epoch": 0.00043036051875404603, |
| "grad_norm": 0.5390625, |
| "learning_rate": 0.00029999993771380175, |
| "loss": 0.4722, |
| "step": 3070 |
| }, |
| { |
| "epoch": 0.00043176234454803315, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.0002999999371102569, |
| "loss": 0.5205, |
| "step": 3080 |
| }, |
| { |
| "epoch": 0.00043316417034202026, |
| "grad_norm": 0.314453125, |
| "learning_rate": 0.000299999936503802, |
| "loss": 0.407, |
| "step": 3090 |
| }, |
| { |
| "epoch": 0.0004345659961360074, |
| "grad_norm": 1.2109375, |
| "learning_rate": 0.00029999993589443705, |
| "loss": 0.495, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.0004359678219299945, |
| "grad_norm": 0.81640625, |
| "learning_rate": 0.000299999935282162, |
| "loss": 0.4925, |
| "step": 3110 |
| }, |
| { |
| "epoch": 0.0004373696477239816, |
| "grad_norm": 0.9921875, |
| "learning_rate": 0.0002999999346669769, |
| "loss": 0.42, |
| "step": 3120 |
| }, |
| { |
| "epoch": 0.0004387714735179687, |
| "grad_norm": 0.3984375, |
| "learning_rate": 0.0002999999340488818, |
| "loss": 0.3723, |
| "step": 3130 |
| }, |
| { |
| "epoch": 0.00044017329931195584, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.0002999999334278766, |
| "loss": 0.5267, |
| "step": 3140 |
| }, |
| { |
| "epoch": 0.000441575125105943, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999999328039614, |
| "loss": 0.4224, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.0004429769508999301, |
| "grad_norm": 1.3203125, |
| "learning_rate": 0.0002999999321771361, |
| "loss": 0.5182, |
| "step": 3160 |
| }, |
| { |
| "epoch": 0.00044437877669391724, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.0002999999315474008, |
| "loss": 0.5608, |
| "step": 3170 |
| }, |
| { |
| "epoch": 0.00044578060248790435, |
| "grad_norm": 0.482421875, |
| "learning_rate": 0.00029999993091475545, |
| "loss": 0.3993, |
| "step": 3180 |
| }, |
| { |
| "epoch": 0.00044718242828189147, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.0002999999302792, |
| "loss": 0.4892, |
| "step": 3190 |
| }, |
| { |
| "epoch": 0.0004485842540758786, |
| "grad_norm": 0.92578125, |
| "learning_rate": 0.00029999992964073447, |
| "loss": 0.5218, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.0004499860798698657, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.00029999992899935895, |
| "loss": 0.3582, |
| "step": 3210 |
| }, |
| { |
| "epoch": 0.0004513879056638528, |
| "grad_norm": 0.53515625, |
| "learning_rate": 0.0002999999283550734, |
| "loss": 0.5051, |
| "step": 3220 |
| }, |
| { |
| "epoch": 0.0004527897314578399, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.00029999992770787776, |
| "loss": 0.4788, |
| "step": 3230 |
| }, |
| { |
| "epoch": 0.0004541915572518271, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.00029999992705777203, |
| "loss": 0.43, |
| "step": 3240 |
| }, |
| { |
| "epoch": 0.0004555933830458142, |
| "grad_norm": 0.9921875, |
| "learning_rate": 0.00029999992640475636, |
| "loss": 0.5673, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.0004569952088398013, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.0002999999257488305, |
| "loss": 0.4597, |
| "step": 3260 |
| }, |
| { |
| "epoch": 0.00045839703463378844, |
| "grad_norm": 0.94921875, |
| "learning_rate": 0.0002999999250899947, |
| "loss": 0.5661, |
| "step": 3270 |
| }, |
| { |
| "epoch": 0.00045979886042777556, |
| "grad_norm": 0.427734375, |
| "learning_rate": 0.0002999999244282488, |
| "loss": 0.5277, |
| "step": 3280 |
| }, |
| { |
| "epoch": 0.00046120068622176267, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.00029999992376359287, |
| "loss": 0.4895, |
| "step": 3290 |
| }, |
| { |
| "epoch": 0.0004626025120157498, |
| "grad_norm": 0.96875, |
| "learning_rate": 0.0002999999230960269, |
| "loss": 0.4452, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.0004640043378097369, |
| "grad_norm": 0.55078125, |
| "learning_rate": 0.00029999992242555083, |
| "loss": 0.4958, |
| "step": 3310 |
| }, |
| { |
| "epoch": 0.000465406163603724, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.0002999999217521647, |
| "loss": 0.4742, |
| "step": 3320 |
| }, |
| { |
| "epoch": 0.00046680798939771113, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.00029999992107586853, |
| "loss": 0.4959, |
| "step": 3330 |
| }, |
| { |
| "epoch": 0.0004682098151916983, |
| "grad_norm": 0.640625, |
| "learning_rate": 0.0002999999203966623, |
| "loss": 0.5154, |
| "step": 3340 |
| }, |
| { |
| "epoch": 0.0004696116409856854, |
| "grad_norm": 0.486328125, |
| "learning_rate": 0.00029999991971454607, |
| "loss": 0.5168, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.00047101346677967253, |
| "grad_norm": 0.455078125, |
| "learning_rate": 0.00029999991902951976, |
| "loss": 0.5385, |
| "step": 3360 |
| }, |
| { |
| "epoch": 0.00047241529257365964, |
| "grad_norm": 0.47265625, |
| "learning_rate": 0.00029999991834158345, |
| "loss": 0.5004, |
| "step": 3370 |
| }, |
| { |
| "epoch": 0.00047381711836764676, |
| "grad_norm": 0.2421875, |
| "learning_rate": 0.00029999991765073704, |
| "loss": 0.4502, |
| "step": 3380 |
| }, |
| { |
| "epoch": 0.0004752189441616339, |
| "grad_norm": 0.4765625, |
| "learning_rate": 0.00029999991695698057, |
| "loss": 0.4549, |
| "step": 3390 |
| }, |
| { |
| "epoch": 0.000476620769955621, |
| "grad_norm": 0.4453125, |
| "learning_rate": 0.00029999991626031405, |
| "loss": 0.4557, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.0004780225957496081, |
| "grad_norm": 0.302734375, |
| "learning_rate": 0.00029999991556073753, |
| "loss": 0.4765, |
| "step": 3410 |
| }, |
| { |
| "epoch": 0.0004794244215435952, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.0002999999148582509, |
| "loss": 0.4911, |
| "step": 3420 |
| }, |
| { |
| "epoch": 0.00048082624733758233, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.00029999991415285423, |
| "loss": 0.5774, |
| "step": 3430 |
| }, |
| { |
| "epoch": 0.0004822280731315695, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.0002999999134445475, |
| "loss": 0.4328, |
| "step": 3440 |
| }, |
| { |
| "epoch": 0.0004836298989255566, |
| "grad_norm": 0.37109375, |
| "learning_rate": 0.00029999991273333077, |
| "loss": 0.4692, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.00048503172471954373, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.00029999991201920393, |
| "loss": 0.4571, |
| "step": 3460 |
| }, |
| { |
| "epoch": 0.00048643355051353085, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.00029999991130216704, |
| "loss": 0.5374, |
| "step": 3470 |
| }, |
| { |
| "epoch": 0.00048783537630751796, |
| "grad_norm": 0.98046875, |
| "learning_rate": 0.00029999991058222015, |
| "loss": 0.5196, |
| "step": 3480 |
| }, |
| { |
| "epoch": 0.0004892372021015051, |
| "grad_norm": 0.44921875, |
| "learning_rate": 0.0002999999098593632, |
| "loss": 0.4814, |
| "step": 3490 |
| }, |
| { |
| "epoch": 0.0004906390278954922, |
| "grad_norm": 1.21875, |
| "learning_rate": 0.00029999990913359616, |
| "loss": 0.5114, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.0004906390278954922, |
| "eval_loss": 0.19615845382213593, |
| "eval_runtime": 113.7184, |
| "eval_samples_per_second": 200.873, |
| "eval_steps_per_second": 1.574, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.0004920408536894794, |
| "grad_norm": 0.6875, |
| "learning_rate": 0.00029999990840491906, |
| "loss": 0.5248, |
| "step": 3510 |
| }, |
| { |
| "epoch": 0.0004934426794834664, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.00029999990767333196, |
| "loss": 0.4625, |
| "step": 3520 |
| }, |
| { |
| "epoch": 0.0004948445052774536, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999999069388348, |
| "loss": 0.4906, |
| "step": 3530 |
| }, |
| { |
| "epoch": 0.0004962463310714407, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.00029999990620142755, |
| "loss": 0.3924, |
| "step": 3540 |
| }, |
| { |
| "epoch": 0.0004976481568654278, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.0002999999054611103, |
| "loss": 0.5016, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.0004990499826594149, |
| "grad_norm": 0.87890625, |
| "learning_rate": 0.000299999904717883, |
| "loss": 0.4974, |
| "step": 3560 |
| }, |
| { |
| "epoch": 0.000500451808453402, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.0002999999039717456, |
| "loss": 0.5082, |
| "step": 3570 |
| }, |
| { |
| "epoch": 0.0005018536342473892, |
| "grad_norm": 0.279296875, |
| "learning_rate": 0.00029999990322269814, |
| "loss": 0.5716, |
| "step": 3580 |
| }, |
| { |
| "epoch": 0.0005032554600413763, |
| "grad_norm": 0.734375, |
| "learning_rate": 0.00029999990247074073, |
| "loss": 0.548, |
| "step": 3590 |
| }, |
| { |
| "epoch": 0.0005046572858353635, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.00029999990171587315, |
| "loss": 0.4553, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.0005060591116293505, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.0002999999009580956, |
| "loss": 0.5029, |
| "step": 3610 |
| }, |
| { |
| "epoch": 0.0005074609374233377, |
| "grad_norm": 0.46484375, |
| "learning_rate": 0.0002999999001974079, |
| "loss": 0.4318, |
| "step": 3620 |
| }, |
| { |
| "epoch": 0.0005088627632173247, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.00029999989943381027, |
| "loss": 0.5442, |
| "step": 3630 |
| }, |
| { |
| "epoch": 0.0005102645890113119, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.00029999989866730253, |
| "loss": 0.4907, |
| "step": 3640 |
| }, |
| { |
| "epoch": 0.000511666414805299, |
| "grad_norm": 0.30078125, |
| "learning_rate": 0.00029999989789788475, |
| "loss": 0.4758, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.0005130682405992861, |
| "grad_norm": 0.96484375, |
| "learning_rate": 0.0002999998971255569, |
| "loss": 0.4656, |
| "step": 3660 |
| }, |
| { |
| "epoch": 0.0005144700663932732, |
| "grad_norm": 0.431640625, |
| "learning_rate": 0.000299999896350319, |
| "loss": 0.5217, |
| "step": 3670 |
| }, |
| { |
| "epoch": 0.0005158718921872604, |
| "grad_norm": 0.44921875, |
| "learning_rate": 0.00029999989557217107, |
| "loss": 0.5127, |
| "step": 3680 |
| }, |
| { |
| "epoch": 0.0005172737179812475, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.0002999998947911131, |
| "loss": 0.5713, |
| "step": 3690 |
| }, |
| { |
| "epoch": 0.0005186755437752346, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.00029999989400714507, |
| "loss": 0.4652, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.0005200773695692218, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.00029999989322026697, |
| "loss": 0.5029, |
| "step": 3710 |
| }, |
| { |
| "epoch": 0.0005214791953632088, |
| "grad_norm": 0.21484375, |
| "learning_rate": 0.0002999998924304788, |
| "loss": 0.4575, |
| "step": 3720 |
| }, |
| { |
| "epoch": 0.000522881021157196, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.00029999989163778066, |
| "loss": 0.4947, |
| "step": 3730 |
| }, |
| { |
| "epoch": 0.0005242828469511831, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.0002999998908421724, |
| "loss": 0.4046, |
| "step": 3740 |
| }, |
| { |
| "epoch": 0.0005256846727451702, |
| "grad_norm": 0.9609375, |
| "learning_rate": 0.0002999998900436541, |
| "loss": 0.5298, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.0005270864985391573, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999988924222577, |
| "loss": 0.4179, |
| "step": 3760 |
| }, |
| { |
| "epoch": 0.0005284883243331445, |
| "grad_norm": 0.291015625, |
| "learning_rate": 0.00029999988843788734, |
| "loss": 0.5142, |
| "step": 3770 |
| }, |
| { |
| "epoch": 0.0005298901501271316, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.0002999998876306389, |
| "loss": 0.4722, |
| "step": 3780 |
| }, |
| { |
| "epoch": 0.0005312919759211187, |
| "grad_norm": 0.92578125, |
| "learning_rate": 0.0002999998868204804, |
| "loss": 0.5202, |
| "step": 3790 |
| }, |
| { |
| "epoch": 0.0005326938017151059, |
| "grad_norm": 1.1328125, |
| "learning_rate": 0.00029999988600741187, |
| "loss": 0.5537, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.0005340956275090929, |
| "grad_norm": 0.33984375, |
| "learning_rate": 0.0002999998851914333, |
| "loss": 0.5072, |
| "step": 3810 |
| }, |
| { |
| "epoch": 0.0005354974533030801, |
| "grad_norm": 1.140625, |
| "learning_rate": 0.0002999998843725446, |
| "loss": 0.4999, |
| "step": 3820 |
| }, |
| { |
| "epoch": 0.0005368992790970672, |
| "grad_norm": 0.7265625, |
| "learning_rate": 0.0002999998835507459, |
| "loss": 0.4916, |
| "step": 3830 |
| }, |
| { |
| "epoch": 0.0005383011048910543, |
| "grad_norm": 0.82421875, |
| "learning_rate": 0.0002999998827260372, |
| "loss": 0.6081, |
| "step": 3840 |
| }, |
| { |
| "epoch": 0.0005397029306850414, |
| "grad_norm": 0.671875, |
| "learning_rate": 0.0002999998818984184, |
| "loss": 0.5145, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.0005411047564790285, |
| "grad_norm": 1.4765625, |
| "learning_rate": 0.00029999988106788956, |
| "loss": 0.4729, |
| "step": 3860 |
| }, |
| { |
| "epoch": 0.0005425065822730157, |
| "grad_norm": 0.81640625, |
| "learning_rate": 0.0002999998802344506, |
| "loss": 0.5762, |
| "step": 3870 |
| }, |
| { |
| "epoch": 0.0005439084080670028, |
| "grad_norm": 0.90234375, |
| "learning_rate": 0.0002999998793981017, |
| "loss": 0.5645, |
| "step": 3880 |
| }, |
| { |
| "epoch": 0.00054531023386099, |
| "grad_norm": 0.3203125, |
| "learning_rate": 0.0002999998785588427, |
| "loss": 0.5388, |
| "step": 3890 |
| }, |
| { |
| "epoch": 0.000546712059654977, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.00029999987771667366, |
| "loss": 0.4449, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.0005481138854489642, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.00029999987687159455, |
| "loss": 0.5345, |
| "step": 3910 |
| }, |
| { |
| "epoch": 0.0005495157112429512, |
| "grad_norm": 0.53125, |
| "learning_rate": 0.00029999987602360545, |
| "loss": 0.319, |
| "step": 3920 |
| }, |
| { |
| "epoch": 0.0005509175370369384, |
| "grad_norm": 0.427734375, |
| "learning_rate": 0.0002999998751727062, |
| "loss": 0.3845, |
| "step": 3930 |
| }, |
| { |
| "epoch": 0.0005523193628309255, |
| "grad_norm": 0.87890625, |
| "learning_rate": 0.00029999987431889697, |
| "loss": 0.4358, |
| "step": 3940 |
| }, |
| { |
| "epoch": 0.0005537211886249126, |
| "grad_norm": 0.76171875, |
| "learning_rate": 0.00029999987346217765, |
| "loss": 0.424, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.0005551230144188997, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.00029999987260254834, |
| "loss": 0.5856, |
| "step": 3960 |
| }, |
| { |
| "epoch": 0.0005565248402128869, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.00029999987174000886, |
| "loss": 0.4395, |
| "step": 3970 |
| }, |
| { |
| "epoch": 0.000557926666006874, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.00029999987087455944, |
| "loss": 0.4667, |
| "step": 3980 |
| }, |
| { |
| "epoch": 0.0005593284918008611, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.0002999998700061999, |
| "loss": 0.5008, |
| "step": 3990 |
| }, |
| { |
| "epoch": 0.0005607303175948483, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.0002999998691349304, |
| "loss": 0.5067, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.0005607303175948483, |
| "eval_loss": 0.1947367787361145, |
| "eval_runtime": 112.0665, |
| "eval_samples_per_second": 203.834, |
| "eval_steps_per_second": 1.597, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.0005621321433888353, |
| "grad_norm": 0.8984375, |
| "learning_rate": 0.0002999998682607508, |
| "loss": 0.4297, |
| "step": 4010 |
| }, |
| { |
| "epoch": 0.0005635339691828225, |
| "grad_norm": 0.22265625, |
| "learning_rate": 0.0002999998673836611, |
| "loss": 0.3309, |
| "step": 4020 |
| }, |
| { |
| "epoch": 0.0005649357949768096, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999998665036614, |
| "loss": 0.3924, |
| "step": 4030 |
| }, |
| { |
| "epoch": 0.0005663376207707967, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.00029999986562075163, |
| "loss": 0.4433, |
| "step": 4040 |
| }, |
| { |
| "epoch": 0.0005677394465647838, |
| "grad_norm": 0.458984375, |
| "learning_rate": 0.00029999986473493184, |
| "loss": 0.4218, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.000569141272358771, |
| "grad_norm": 0.39453125, |
| "learning_rate": 0.000299999863846202, |
| "loss": 0.5819, |
| "step": 4060 |
| }, |
| { |
| "epoch": 0.0005705430981527581, |
| "grad_norm": 0.71484375, |
| "learning_rate": 0.0002999998629545621, |
| "loss": 0.5292, |
| "step": 4070 |
| }, |
| { |
| "epoch": 0.0005719449239467452, |
| "grad_norm": 0.55078125, |
| "learning_rate": 0.00029999986206001215, |
| "loss": 0.4745, |
| "step": 4080 |
| }, |
| { |
| "epoch": 0.0005733467497407324, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.00029999986116255214, |
| "loss": 0.5194, |
| "step": 4090 |
| }, |
| { |
| "epoch": 0.0005747485755347194, |
| "grad_norm": 0.5546875, |
| "learning_rate": 0.0002999998602621821, |
| "loss": 0.5094, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.0005761504013287066, |
| "grad_norm": 0.416015625, |
| "learning_rate": 0.00029999985935890203, |
| "loss": 0.4252, |
| "step": 4110 |
| }, |
| { |
| "epoch": 0.0005775522271226936, |
| "grad_norm": 0.53125, |
| "learning_rate": 0.0002999998584527118, |
| "loss": 0.439, |
| "step": 4120 |
| }, |
| { |
| "epoch": 0.0005789540529166808, |
| "grad_norm": 0.6953125, |
| "learning_rate": 0.0002999998575436116, |
| "loss": 0.5719, |
| "step": 4130 |
| }, |
| { |
| "epoch": 0.0005803558787106679, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.00029999985663160133, |
| "loss": 0.4448, |
| "step": 4140 |
| }, |
| { |
| "epoch": 0.000581757704504655, |
| "grad_norm": 0.484375, |
| "learning_rate": 0.00029999985571668107, |
| "loss": 0.4758, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.0005831595302986422, |
| "grad_norm": 0.91796875, |
| "learning_rate": 0.0002999998547988507, |
| "loss": 0.4819, |
| "step": 4160 |
| }, |
| { |
| "epoch": 0.0005845613560926293, |
| "grad_norm": 0.45703125, |
| "learning_rate": 0.0002999998538781103, |
| "loss": 0.4092, |
| "step": 4170 |
| }, |
| { |
| "epoch": 0.0005859631818866164, |
| "grad_norm": 0.29296875, |
| "learning_rate": 0.00029999985295445984, |
| "loss": 0.4887, |
| "step": 4180 |
| }, |
| { |
| "epoch": 0.0005873650076806035, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.0002999998520278993, |
| "loss": 0.5011, |
| "step": 4190 |
| }, |
| { |
| "epoch": 0.0005887668334745907, |
| "grad_norm": 1.140625, |
| "learning_rate": 0.0002999998510984288, |
| "loss": 0.4195, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.0005901686592685777, |
| "grad_norm": 0.8984375, |
| "learning_rate": 0.0002999998501660482, |
| "loss": 0.4402, |
| "step": 4210 |
| }, |
| { |
| "epoch": 0.0005915704850625649, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.0002999998492307575, |
| "loss": 0.4796, |
| "step": 4220 |
| }, |
| { |
| "epoch": 0.000592972310856552, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.0002999998482925568, |
| "loss": 0.4227, |
| "step": 4230 |
| }, |
| { |
| "epoch": 0.0005943741366505391, |
| "grad_norm": 0.87109375, |
| "learning_rate": 0.000299999847351446, |
| "loss": 0.4143, |
| "step": 4240 |
| }, |
| { |
| "epoch": 0.0005957759624445262, |
| "grad_norm": 0.46875, |
| "learning_rate": 0.0002999998464074252, |
| "loss": 0.4414, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.0005971777882385134, |
| "grad_norm": 0.361328125, |
| "learning_rate": 0.0002999998454604944, |
| "loss": 0.4819, |
| "step": 4260 |
| }, |
| { |
| "epoch": 0.0005985796140325005, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.00029999984451065353, |
| "loss": 0.3205, |
| "step": 4270 |
| }, |
| { |
| "epoch": 0.0005999814398264876, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.0002999998435579026, |
| "loss": 0.443, |
| "step": 4280 |
| }, |
| { |
| "epoch": 0.0006013832656204748, |
| "grad_norm": 0.76953125, |
| "learning_rate": 0.00029999984260224157, |
| "loss": 0.4969, |
| "step": 4290 |
| }, |
| { |
| "epoch": 0.0006027850914144618, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.0002999998416436705, |
| "loss": 0.5088, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.000604186917208449, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.00029999984068218935, |
| "loss": 0.544, |
| "step": 4310 |
| }, |
| { |
| "epoch": 0.0006055887430024361, |
| "grad_norm": 0.482421875, |
| "learning_rate": 0.00029999983971779824, |
| "loss": 0.4043, |
| "step": 4320 |
| }, |
| { |
| "epoch": 0.0006069905687964232, |
| "grad_norm": 0.94921875, |
| "learning_rate": 0.000299999838750497, |
| "loss": 0.4532, |
| "step": 4330 |
| }, |
| { |
| "epoch": 0.0006083923945904103, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999998377802858, |
| "loss": 0.4442, |
| "step": 4340 |
| }, |
| { |
| "epoch": 0.0006097942203843975, |
| "grad_norm": 0.52734375, |
| "learning_rate": 0.0002999998368071645, |
| "loss": 0.4439, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.0006111960461783846, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.00029999983583113316, |
| "loss": 0.429, |
| "step": 4360 |
| }, |
| { |
| "epoch": 0.0006125978719723717, |
| "grad_norm": 0.30078125, |
| "learning_rate": 0.00029999983485219173, |
| "loss": 0.4306, |
| "step": 4370 |
| }, |
| { |
| "epoch": 0.0006139996977663589, |
| "grad_norm": 0.2890625, |
| "learning_rate": 0.0002999998338703403, |
| "loss": 0.4207, |
| "step": 4380 |
| }, |
| { |
| "epoch": 0.0006154015235603459, |
| "grad_norm": 1.1953125, |
| "learning_rate": 0.00029999983288557877, |
| "loss": 0.4525, |
| "step": 4390 |
| }, |
| { |
| "epoch": 0.0006168033493543331, |
| "grad_norm": 0.7265625, |
| "learning_rate": 0.00029999983189790724, |
| "loss": 0.3968, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.0006182051751483201, |
| "grad_norm": 0.3203125, |
| "learning_rate": 0.00029999983090732565, |
| "loss": 0.3892, |
| "step": 4410 |
| }, |
| { |
| "epoch": 0.0006196070009423073, |
| "grad_norm": 0.255859375, |
| "learning_rate": 0.00029999982991383396, |
| "loss": 0.3471, |
| "step": 4420 |
| }, |
| { |
| "epoch": 0.0006210088267362944, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.00029999982891743226, |
| "loss": 0.5504, |
| "step": 4430 |
| }, |
| { |
| "epoch": 0.0006224106525302815, |
| "grad_norm": 0.330078125, |
| "learning_rate": 0.0002999998279181205, |
| "loss": 0.434, |
| "step": 4440 |
| }, |
| { |
| "epoch": 0.0006238124783242687, |
| "grad_norm": 1.09375, |
| "learning_rate": 0.0002999998269158987, |
| "loss": 0.5546, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.0006252143041182558, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.00029999982591076687, |
| "loss": 0.4573, |
| "step": 4460 |
| }, |
| { |
| "epoch": 0.0006266161299122429, |
| "grad_norm": 1.0078125, |
| "learning_rate": 0.00029999982490272497, |
| "loss": 0.4395, |
| "step": 4470 |
| }, |
| { |
| "epoch": 0.00062801795570623, |
| "grad_norm": 0.48828125, |
| "learning_rate": 0.000299999823891773, |
| "loss": 0.4691, |
| "step": 4480 |
| }, |
| { |
| "epoch": 0.0006294197815002172, |
| "grad_norm": 0.38671875, |
| "learning_rate": 0.00029999982287791106, |
| "loss": 0.4948, |
| "step": 4490 |
| }, |
| { |
| "epoch": 0.0006308216072942042, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.000299999821861139, |
| "loss": 0.3723, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.0006308216072942042, |
| "eval_loss": 0.19076591730117798, |
| "eval_runtime": 112.1557, |
| "eval_samples_per_second": 203.672, |
| "eval_steps_per_second": 1.596, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.0006322234330881914, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999998208414569, |
| "loss": 0.4265, |
| "step": 4510 |
| }, |
| { |
| "epoch": 0.0006336252588821785, |
| "grad_norm": 1.078125, |
| "learning_rate": 0.00029999981981886477, |
| "loss": 0.506, |
| "step": 4520 |
| }, |
| { |
| "epoch": 0.0006350270846761656, |
| "grad_norm": 0.59375, |
| "learning_rate": 0.0002999998187933626, |
| "loss": 0.4459, |
| "step": 4530 |
| }, |
| { |
| "epoch": 0.0006364289104701527, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.00029999981776495033, |
| "loss": 0.506, |
| "step": 4540 |
| }, |
| { |
| "epoch": 0.0006378307362641399, |
| "grad_norm": 0.275390625, |
| "learning_rate": 0.000299999816733628, |
| "loss": 0.4536, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.000639232562058127, |
| "grad_norm": 0.86328125, |
| "learning_rate": 0.0002999998156993957, |
| "loss": 0.554, |
| "step": 4560 |
| }, |
| { |
| "epoch": 0.0006406343878521141, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.0002999998146622533, |
| "loss": 0.5602, |
| "step": 4570 |
| }, |
| { |
| "epoch": 0.0006420362136461013, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.0002999998136222009, |
| "loss": 0.471, |
| "step": 4580 |
| }, |
| { |
| "epoch": 0.0006434380394400883, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.0002999998125792384, |
| "loss": 0.49, |
| "step": 4590 |
| }, |
| { |
| "epoch": 0.0006448398652340755, |
| "grad_norm": 0.86328125, |
| "learning_rate": 0.00029999981153336586, |
| "loss": 0.4289, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.0006462416910280626, |
| "grad_norm": 0.82421875, |
| "learning_rate": 0.00029999981048458327, |
| "loss": 0.4727, |
| "step": 4610 |
| }, |
| { |
| "epoch": 0.0006476435168220497, |
| "grad_norm": 0.40625, |
| "learning_rate": 0.0002999998094328906, |
| "loss": 0.4323, |
| "step": 4620 |
| }, |
| { |
| "epoch": 0.0006490453426160368, |
| "grad_norm": 0.921875, |
| "learning_rate": 0.00029999980837828793, |
| "loss": 0.4516, |
| "step": 4630 |
| }, |
| { |
| "epoch": 0.000650447168410024, |
| "grad_norm": 0.828125, |
| "learning_rate": 0.0002999998073207752, |
| "loss": 0.4945, |
| "step": 4640 |
| }, |
| { |
| "epoch": 0.0006518489942040111, |
| "grad_norm": 0.87109375, |
| "learning_rate": 0.00029999980626035244, |
| "loss": 0.4593, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.0006532508199979982, |
| "grad_norm": 0.6953125, |
| "learning_rate": 0.0002999998051970196, |
| "loss": 0.3685, |
| "step": 4660 |
| }, |
| { |
| "epoch": 0.0006546526457919854, |
| "grad_norm": 0.466796875, |
| "learning_rate": 0.00029999980413077674, |
| "loss": 0.4393, |
| "step": 4670 |
| }, |
| { |
| "epoch": 0.0006560544715859724, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.0002999998030616238, |
| "loss": 0.4165, |
| "step": 4680 |
| }, |
| { |
| "epoch": 0.0006574562973799596, |
| "grad_norm": 0.52734375, |
| "learning_rate": 0.0002999998019895608, |
| "loss": 0.4665, |
| "step": 4690 |
| }, |
| { |
| "epoch": 0.0006588581231739466, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.0002999998009145878, |
| "loss": 0.544, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.0006602599489679338, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.0002999997998367047, |
| "loss": 0.5659, |
| "step": 4710 |
| }, |
| { |
| "epoch": 0.0006616617747619209, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.0002999997987559116, |
| "loss": 0.4516, |
| "step": 4720 |
| }, |
| { |
| "epoch": 0.000663063600555908, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999997976722084, |
| "loss": 0.4434, |
| "step": 4730 |
| }, |
| { |
| "epoch": 0.0006644654263498952, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.0002999997965855952, |
| "loss": 0.4576, |
| "step": 4740 |
| }, |
| { |
| "epoch": 0.0006658672521438823, |
| "grad_norm": 0.396484375, |
| "learning_rate": 0.0002999997954960719, |
| "loss": 0.4261, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.0006672690779378694, |
| "grad_norm": 0.55078125, |
| "learning_rate": 0.0002999997944036386, |
| "loss": 0.477, |
| "step": 4760 |
| }, |
| { |
| "epoch": 0.0006686709037318565, |
| "grad_norm": 0.671875, |
| "learning_rate": 0.0002999997933082952, |
| "loss": 0.4698, |
| "step": 4770 |
| }, |
| { |
| "epoch": 0.0006700727295258437, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999997922100418, |
| "loss": 0.4822, |
| "step": 4780 |
| }, |
| { |
| "epoch": 0.0006714745553198307, |
| "grad_norm": 0.5390625, |
| "learning_rate": 0.00029999979110887834, |
| "loss": 0.4478, |
| "step": 4790 |
| }, |
| { |
| "epoch": 0.0006728763811138179, |
| "grad_norm": 0.373046875, |
| "learning_rate": 0.0002999997900048048, |
| "loss": 0.4292, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.000674278206907805, |
| "grad_norm": 0.4921875, |
| "learning_rate": 0.00029999978889782127, |
| "loss": 0.5094, |
| "step": 4810 |
| }, |
| { |
| "epoch": 0.0006756800327017921, |
| "grad_norm": 0.337890625, |
| "learning_rate": 0.0002999997877879277, |
| "loss": 0.4333, |
| "step": 4820 |
| }, |
| { |
| "epoch": 0.0006770818584957792, |
| "grad_norm": 0.474609375, |
| "learning_rate": 0.00029999978667512403, |
| "loss": 0.4495, |
| "step": 4830 |
| }, |
| { |
| "epoch": 0.0006784836842897664, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.0002999997855594103, |
| "loss": 0.4684, |
| "step": 4840 |
| }, |
| { |
| "epoch": 0.0006798855100837535, |
| "grad_norm": 0.263671875, |
| "learning_rate": 0.0002999997844407865, |
| "loss": 0.3566, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.0006812873358777406, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.0002999997833192527, |
| "loss": 0.4387, |
| "step": 4860 |
| }, |
| { |
| "epoch": 0.0006826891616717278, |
| "grad_norm": 1.2265625, |
| "learning_rate": 0.0002999997821948089, |
| "loss": 0.4308, |
| "step": 4870 |
| }, |
| { |
| "epoch": 0.0006840909874657148, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.0002999997810674549, |
| "loss": 0.5234, |
| "step": 4880 |
| }, |
| { |
| "epoch": 0.000685492813259702, |
| "grad_norm": 0.234375, |
| "learning_rate": 0.000299999779937191, |
| "loss": 0.5019, |
| "step": 4890 |
| }, |
| { |
| "epoch": 0.000686894639053689, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.000299999778804017, |
| "loss": 0.4867, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.0006882964648476762, |
| "grad_norm": 0.9921875, |
| "learning_rate": 0.00029999977766793294, |
| "loss": 0.5128, |
| "step": 4910 |
| }, |
| { |
| "epoch": 0.0006896982906416633, |
| "grad_norm": 1.921875, |
| "learning_rate": 0.0002999997765289389, |
| "loss": 0.5317, |
| "step": 4920 |
| }, |
| { |
| "epoch": 0.0006911001164356504, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.0002999997753870347, |
| "loss": 0.5652, |
| "step": 4930 |
| }, |
| { |
| "epoch": 0.0006925019422296376, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.0002999997742422205, |
| "loss": 0.4115, |
| "step": 4940 |
| }, |
| { |
| "epoch": 0.0006939037680236247, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.00029999977309449626, |
| "loss": 0.5186, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.0006953055938176118, |
| "grad_norm": 0.93359375, |
| "learning_rate": 0.000299999771943862, |
| "loss": 0.5305, |
| "step": 4960 |
| }, |
| { |
| "epoch": 0.0006967074196115989, |
| "grad_norm": 0.9140625, |
| "learning_rate": 0.0002999997707903176, |
| "loss": 0.4705, |
| "step": 4970 |
| }, |
| { |
| "epoch": 0.0006981092454055861, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.0002999997696338632, |
| "loss": 0.4838, |
| "step": 4980 |
| }, |
| { |
| "epoch": 0.0006995110711995731, |
| "grad_norm": 0.40625, |
| "learning_rate": 0.0002999997684744988, |
| "loss": 0.5191, |
| "step": 4990 |
| }, |
| { |
| "epoch": 0.0007009128969935603, |
| "grad_norm": 0.478515625, |
| "learning_rate": 0.00029999976731222435, |
| "loss": 0.4744, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.0007009128969935603, |
| "eval_loss": 0.1905738264322281, |
| "eval_runtime": 113.7667, |
| "eval_samples_per_second": 200.788, |
| "eval_steps_per_second": 1.573, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.0007023147227875474, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.00029999976614703976, |
| "loss": 0.5915, |
| "step": 5010 |
| }, |
| { |
| "epoch": 0.0007037165485815345, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.00029999976497894516, |
| "loss": 0.5237, |
| "step": 5020 |
| }, |
| { |
| "epoch": 0.0007051183743755217, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.0002999997638079406, |
| "loss": 0.5251, |
| "step": 5030 |
| }, |
| { |
| "epoch": 0.0007065202001695088, |
| "grad_norm": 0.439453125, |
| "learning_rate": 0.0002999997626340259, |
| "loss": 0.4624, |
| "step": 5040 |
| }, |
| { |
| "epoch": 0.0007079220259634959, |
| "grad_norm": 0.953125, |
| "learning_rate": 0.0002999997614572012, |
| "loss": 0.4527, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.000709323851757483, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.0002999997602774664, |
| "loss": 0.4864, |
| "step": 5060 |
| }, |
| { |
| "epoch": 0.0007107256775514702, |
| "grad_norm": 1.046875, |
| "learning_rate": 0.00029999975909482157, |
| "loss": 0.4791, |
| "step": 5070 |
| }, |
| { |
| "epoch": 0.0007121275033454572, |
| "grad_norm": 0.2109375, |
| "learning_rate": 0.0002999997579092667, |
| "loss": 0.4468, |
| "step": 5080 |
| }, |
| { |
| "epoch": 0.0007135293291394444, |
| "grad_norm": 0.318359375, |
| "learning_rate": 0.0002999997567208018, |
| "loss": 0.4285, |
| "step": 5090 |
| }, |
| { |
| "epoch": 0.0007149311549334315, |
| "grad_norm": 0.30078125, |
| "learning_rate": 0.0002999997555294268, |
| "loss": 0.4075, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.0007163329807274186, |
| "grad_norm": 0.8125, |
| "learning_rate": 0.0002999997543351418, |
| "loss": 0.4852, |
| "step": 5110 |
| }, |
| { |
| "epoch": 0.0007177348065214057, |
| "grad_norm": 0.94140625, |
| "learning_rate": 0.0002999997531379467, |
| "loss": 0.4541, |
| "step": 5120 |
| }, |
| { |
| "epoch": 0.0007191366323153929, |
| "grad_norm": 0.9375, |
| "learning_rate": 0.00029999975193784165, |
| "loss": 0.5359, |
| "step": 5130 |
| }, |
| { |
| "epoch": 0.00072053845810938, |
| "grad_norm": 0.8125, |
| "learning_rate": 0.0002999997507348265, |
| "loss": 0.507, |
| "step": 5140 |
| }, |
| { |
| "epoch": 0.0007219402839033671, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.00029999974952890125, |
| "loss": 0.5468, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.0007233421096973543, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.00029999974832006603, |
| "loss": 0.3978, |
| "step": 5160 |
| }, |
| { |
| "epoch": 0.0007247439354913413, |
| "grad_norm": 0.2109375, |
| "learning_rate": 0.0002999997471083207, |
| "loss": 0.3386, |
| "step": 5170 |
| }, |
| { |
| "epoch": 0.0007261457612853285, |
| "grad_norm": 0.396484375, |
| "learning_rate": 0.0002999997458936653, |
| "loss": 0.4484, |
| "step": 5180 |
| }, |
| { |
| "epoch": 0.0007275475870793155, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.00029999974467609993, |
| "loss": 0.4583, |
| "step": 5190 |
| }, |
| { |
| "epoch": 0.0007289494128733027, |
| "grad_norm": 0.33203125, |
| "learning_rate": 0.0002999997434556245, |
| "loss": 0.4929, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.0007303512386672898, |
| "grad_norm": 0.92578125, |
| "learning_rate": 0.00029999974223223895, |
| "loss": 0.4117, |
| "step": 5210 |
| }, |
| { |
| "epoch": 0.0007317530644612769, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.0002999997410059434, |
| "loss": 0.4758, |
| "step": 5220 |
| }, |
| { |
| "epoch": 0.0007331548902552641, |
| "grad_norm": 0.5546875, |
| "learning_rate": 0.00029999973977673787, |
| "loss": 0.38, |
| "step": 5230 |
| }, |
| { |
| "epoch": 0.0007345567160492512, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.0002999997385446222, |
| "loss": 0.4684, |
| "step": 5240 |
| }, |
| { |
| "epoch": 0.0007359585418432383, |
| "grad_norm": 0.7109375, |
| "learning_rate": 0.0002999997373095965, |
| "loss": 0.4076, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.0007373603676372254, |
| "grad_norm": 0.439453125, |
| "learning_rate": 0.0002999997360716608, |
| "loss": 0.5375, |
| "step": 5260 |
| }, |
| { |
| "epoch": 0.0007387621934312126, |
| "grad_norm": 0.546875, |
| "learning_rate": 0.000299999734830815, |
| "loss": 0.4483, |
| "step": 5270 |
| }, |
| { |
| "epoch": 0.0007401640192251996, |
| "grad_norm": 1.1328125, |
| "learning_rate": 0.0002999997335870592, |
| "loss": 0.4744, |
| "step": 5280 |
| }, |
| { |
| "epoch": 0.0007415658450191868, |
| "grad_norm": 0.6953125, |
| "learning_rate": 0.0002999997323403933, |
| "loss": 0.4671, |
| "step": 5290 |
| }, |
| { |
| "epoch": 0.0007429676708131739, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999997310908174, |
| "loss": 0.4767, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.000744369496607161, |
| "grad_norm": 1.015625, |
| "learning_rate": 0.0002999997298383314, |
| "loss": 0.5018, |
| "step": 5310 |
| }, |
| { |
| "epoch": 0.0007457713224011482, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.0002999997285829354, |
| "loss": 0.4238, |
| "step": 5320 |
| }, |
| { |
| "epoch": 0.0007471731481951353, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.00029999972732462936, |
| "loss": 0.4321, |
| "step": 5330 |
| }, |
| { |
| "epoch": 0.0007485749739891224, |
| "grad_norm": 0.271484375, |
| "learning_rate": 0.00029999972606341323, |
| "loss": 0.4522, |
| "step": 5340 |
| }, |
| { |
| "epoch": 0.0007499767997831095, |
| "grad_norm": 0.5078125, |
| "learning_rate": 0.00029999972479928706, |
| "loss": 0.4748, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.0007513786255770967, |
| "grad_norm": 0.357421875, |
| "learning_rate": 0.00029999972353225083, |
| "loss": 0.4502, |
| "step": 5360 |
| }, |
| { |
| "epoch": 0.0007527804513710837, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.0002999997222623046, |
| "loss": 0.5169, |
| "step": 5370 |
| }, |
| { |
| "epoch": 0.0007541822771650709, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.00029999972098944833, |
| "loss": 0.4289, |
| "step": 5380 |
| }, |
| { |
| "epoch": 0.000755584102959058, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.00029999971971368195, |
| "loss": 0.4885, |
| "step": 5390 |
| }, |
| { |
| "epoch": 0.0007569859287530451, |
| "grad_norm": 0.3203125, |
| "learning_rate": 0.0002999997184350055, |
| "loss": 0.4621, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.0007583877545470322, |
| "grad_norm": 0.314453125, |
| "learning_rate": 0.0002999997171534191, |
| "loss": 0.3875, |
| "step": 5410 |
| }, |
| { |
| "epoch": 0.0007597895803410194, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.0002999997158689226, |
| "loss": 0.3891, |
| "step": 5420 |
| }, |
| { |
| "epoch": 0.0007611914061350065, |
| "grad_norm": 0.408203125, |
| "learning_rate": 0.00029999971458151604, |
| "loss": 0.4734, |
| "step": 5430 |
| }, |
| { |
| "epoch": 0.0007625932319289936, |
| "grad_norm": 1.046875, |
| "learning_rate": 0.00029999971329119945, |
| "loss": 0.5127, |
| "step": 5440 |
| }, |
| { |
| "epoch": 0.0007639950577229808, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.00029999971199797285, |
| "loss": 0.5229, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.0007653968835169678, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.00029999971070183615, |
| "loss": 0.5307, |
| "step": 5460 |
| }, |
| { |
| "epoch": 0.000766798709310955, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.0002999997094027894, |
| "loss": 0.4848, |
| "step": 5470 |
| }, |
| { |
| "epoch": 0.000768200535104942, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.00029999970810083265, |
| "loss": 0.4901, |
| "step": 5480 |
| }, |
| { |
| "epoch": 0.0007696023608989292, |
| "grad_norm": 0.369140625, |
| "learning_rate": 0.00029999970679596584, |
| "loss": 0.4517, |
| "step": 5490 |
| }, |
| { |
| "epoch": 0.0007710041866929163, |
| "grad_norm": 0.263671875, |
| "learning_rate": 0.000299999705488189, |
| "loss": 0.4509, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.0007710041866929163, |
| "eval_loss": 0.18971732258796692, |
| "eval_runtime": 112.0563, |
| "eval_samples_per_second": 203.853, |
| "eval_steps_per_second": 1.597, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.0007724060124869034, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.000299999704177502, |
| "loss": 0.4598, |
| "step": 5510 |
| }, |
| { |
| "epoch": 0.0007738078382808906, |
| "grad_norm": 0.365234375, |
| "learning_rate": 0.00029999970286390506, |
| "loss": 0.4504, |
| "step": 5520 |
| }, |
| { |
| "epoch": 0.0007752096640748777, |
| "grad_norm": 0.81640625, |
| "learning_rate": 0.0002999997015473981, |
| "loss": 0.4226, |
| "step": 5530 |
| }, |
| { |
| "epoch": 0.0007766114898688648, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.000299999700227981, |
| "loss": 0.4202, |
| "step": 5540 |
| }, |
| { |
| "epoch": 0.0007780133156628519, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999996989056539, |
| "loss": 0.4346, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.0007794151414568391, |
| "grad_norm": 0.392578125, |
| "learning_rate": 0.00029999969758041673, |
| "loss": 0.3287, |
| "step": 5560 |
| }, |
| { |
| "epoch": 0.0007808169672508261, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.00029999969625226956, |
| "loss": 0.5162, |
| "step": 5570 |
| }, |
| { |
| "epoch": 0.0007822187930448133, |
| "grad_norm": 0.70703125, |
| "learning_rate": 0.0002999996949212123, |
| "loss": 0.4915, |
| "step": 5580 |
| }, |
| { |
| "epoch": 0.0007836206188388004, |
| "grad_norm": 0.28125, |
| "learning_rate": 0.000299999693587245, |
| "loss": 0.4213, |
| "step": 5590 |
| }, |
| { |
| "epoch": 0.0007850224446327875, |
| "grad_norm": 0.94140625, |
| "learning_rate": 0.00029999969225036766, |
| "loss": 0.3892, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.0007864242704267747, |
| "grad_norm": 0.921875, |
| "learning_rate": 0.0002999996909105803, |
| "loss": 0.4821, |
| "step": 5610 |
| }, |
| { |
| "epoch": 0.0007878260962207618, |
| "grad_norm": 0.423828125, |
| "learning_rate": 0.00029999968956788284, |
| "loss": 0.3946, |
| "step": 5620 |
| }, |
| { |
| "epoch": 0.0007892279220147489, |
| "grad_norm": 0.349609375, |
| "learning_rate": 0.00029999968822227535, |
| "loss": 0.4474, |
| "step": 5630 |
| }, |
| { |
| "epoch": 0.000790629747808736, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.00029999968687375786, |
| "loss": 0.4803, |
| "step": 5640 |
| }, |
| { |
| "epoch": 0.0007920315736027232, |
| "grad_norm": 0.515625, |
| "learning_rate": 0.00029999968552233027, |
| "loss": 0.4669, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.0007934333993967102, |
| "grad_norm": 0.287109375, |
| "learning_rate": 0.0002999996841679926, |
| "loss": 0.3951, |
| "step": 5660 |
| }, |
| { |
| "epoch": 0.0007948352251906974, |
| "grad_norm": 0.275390625, |
| "learning_rate": 0.00029999968281074497, |
| "loss": 0.3465, |
| "step": 5670 |
| }, |
| { |
| "epoch": 0.0007962370509846845, |
| "grad_norm": 0.37109375, |
| "learning_rate": 0.00029999968145058727, |
| "loss": 0.4077, |
| "step": 5680 |
| }, |
| { |
| "epoch": 0.0007976388767786716, |
| "grad_norm": 0.37890625, |
| "learning_rate": 0.0002999996800875195, |
| "loss": 0.4023, |
| "step": 5690 |
| }, |
| { |
| "epoch": 0.0007990407025726587, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.00029999967872154166, |
| "loss": 0.4003, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.0008004425283666459, |
| "grad_norm": 0.8046875, |
| "learning_rate": 0.00029999967735265385, |
| "loss": 0.494, |
| "step": 5710 |
| }, |
| { |
| "epoch": 0.000801844354160633, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.00029999967598085594, |
| "loss": 0.4937, |
| "step": 5720 |
| }, |
| { |
| "epoch": 0.0008032461799546201, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.000299999674606148, |
| "loss": 0.4274, |
| "step": 5730 |
| }, |
| { |
| "epoch": 0.0008046480057486072, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.00029999967322853, |
| "loss": 0.4245, |
| "step": 5740 |
| }, |
| { |
| "epoch": 0.0008060498315425943, |
| "grad_norm": 0.8125, |
| "learning_rate": 0.00029999967184800194, |
| "loss": 0.5633, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.0008074516573365815, |
| "grad_norm": 0.70703125, |
| "learning_rate": 0.00029999967046456387, |
| "loss": 0.4766, |
| "step": 5760 |
| }, |
| { |
| "epoch": 0.0008088534831305685, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00029999966907821575, |
| "loss": 0.4312, |
| "step": 5770 |
| }, |
| { |
| "epoch": 0.0008102553089245557, |
| "grad_norm": 0.45703125, |
| "learning_rate": 0.0002999996676889576, |
| "loss": 0.3684, |
| "step": 5780 |
| }, |
| { |
| "epoch": 0.0008116571347185428, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.0002999996662967893, |
| "loss": 0.3765, |
| "step": 5790 |
| }, |
| { |
| "epoch": 0.0008130589605125299, |
| "grad_norm": 0.46484375, |
| "learning_rate": 0.00029999966490171107, |
| "loss": 0.4691, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.0008144607863065171, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.00029999966350372273, |
| "loss": 0.4451, |
| "step": 5810 |
| }, |
| { |
| "epoch": 0.0008158626121005042, |
| "grad_norm": 0.302734375, |
| "learning_rate": 0.00029999966210282435, |
| "loss": 0.4453, |
| "step": 5820 |
| }, |
| { |
| "epoch": 0.0008172644378944913, |
| "grad_norm": 0.423828125, |
| "learning_rate": 0.00029999966069901596, |
| "loss": 0.3543, |
| "step": 5830 |
| }, |
| { |
| "epoch": 0.0008186662636884784, |
| "grad_norm": 0.86328125, |
| "learning_rate": 0.0002999996592922975, |
| "loss": 0.4092, |
| "step": 5840 |
| }, |
| { |
| "epoch": 0.0008200680894824656, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.00029999965788266904, |
| "loss": 0.4338, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.0008214699152764526, |
| "grad_norm": 0.287109375, |
| "learning_rate": 0.00029999965647013044, |
| "loss": 0.4996, |
| "step": 5860 |
| }, |
| { |
| "epoch": 0.0008228717410704398, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.00029999965505468184, |
| "loss": 0.4116, |
| "step": 5870 |
| }, |
| { |
| "epoch": 0.0008242735668644269, |
| "grad_norm": 0.267578125, |
| "learning_rate": 0.00029999965363632325, |
| "loss": 0.4168, |
| "step": 5880 |
| }, |
| { |
| "epoch": 0.000825675392658414, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.00029999965221505455, |
| "loss": 0.4213, |
| "step": 5890 |
| }, |
| { |
| "epoch": 0.0008270772184524012, |
| "grad_norm": 0.9296875, |
| "learning_rate": 0.0002999996507908758, |
| "loss": 0.4801, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.0008284790442463883, |
| "grad_norm": 1.0703125, |
| "learning_rate": 0.000299999649363787, |
| "loss": 0.4743, |
| "step": 5910 |
| }, |
| { |
| "epoch": 0.0008298808700403754, |
| "grad_norm": 0.22265625, |
| "learning_rate": 0.0002999996479337882, |
| "loss": 0.2799, |
| "step": 5920 |
| }, |
| { |
| "epoch": 0.0008312826958343625, |
| "grad_norm": 0.4609375, |
| "learning_rate": 0.0002999996465008793, |
| "loss": 0.5026, |
| "step": 5930 |
| }, |
| { |
| "epoch": 0.0008326845216283497, |
| "grad_norm": 1.03125, |
| "learning_rate": 0.0002999996450650604, |
| "loss": 0.4738, |
| "step": 5940 |
| }, |
| { |
| "epoch": 0.0008340863474223367, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.00029999964362633144, |
| "loss": 0.4826, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.0008354881732163239, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.0002999996421846924, |
| "loss": 0.4934, |
| "step": 5960 |
| }, |
| { |
| "epoch": 0.000836889999010311, |
| "grad_norm": 0.38671875, |
| "learning_rate": 0.00029999964074014336, |
| "loss": 0.4993, |
| "step": 5970 |
| }, |
| { |
| "epoch": 0.0008382918248042981, |
| "grad_norm": 0.625, |
| "learning_rate": 0.0002999996392926843, |
| "loss": 0.4764, |
| "step": 5980 |
| }, |
| { |
| "epoch": 0.0008396936505982852, |
| "grad_norm": 0.58984375, |
| "learning_rate": 0.0002999996378423151, |
| "loss": 0.4652, |
| "step": 5990 |
| }, |
| { |
| "epoch": 0.0008410954763922723, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.00029999963638903594, |
| "loss": 0.465, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.0008410954763922723, |
| "eval_loss": 0.18518111109733582, |
| "eval_runtime": 112.0637, |
| "eval_samples_per_second": 203.839, |
| "eval_steps_per_second": 1.597, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.0008424973021862595, |
| "grad_norm": 1.0, |
| "learning_rate": 0.0002999996349328467, |
| "loss": 0.5146, |
| "step": 6010 |
| }, |
| { |
| "epoch": 0.0008438991279802466, |
| "grad_norm": 0.369140625, |
| "learning_rate": 0.00029999963347374743, |
| "loss": 0.4153, |
| "step": 6020 |
| }, |
| { |
| "epoch": 0.0008453009537742337, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.0002999996320117381, |
| "loss": 0.4145, |
| "step": 6030 |
| }, |
| { |
| "epoch": 0.0008467027795682208, |
| "grad_norm": 0.298828125, |
| "learning_rate": 0.0002999996305468187, |
| "loss": 0.4206, |
| "step": 6040 |
| }, |
| { |
| "epoch": 0.000848104605362208, |
| "grad_norm": 0.6875, |
| "learning_rate": 0.00029999962907898927, |
| "loss": 0.4394, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.000849506431156195, |
| "grad_norm": 0.416015625, |
| "learning_rate": 0.0002999996276082498, |
| "loss": 0.3766, |
| "step": 6060 |
| }, |
| { |
| "epoch": 0.0008509082569501822, |
| "grad_norm": 0.42578125, |
| "learning_rate": 0.0002999996261346003, |
| "loss": 0.4398, |
| "step": 6070 |
| }, |
| { |
| "epoch": 0.0008523100827441693, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.00029999962465804075, |
| "loss": 0.4544, |
| "step": 6080 |
| }, |
| { |
| "epoch": 0.0008537119085381564, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.00029999962317857115, |
| "loss": 0.4677, |
| "step": 6090 |
| }, |
| { |
| "epoch": 0.0008551137343321436, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.0002999996216961915, |
| "loss": 0.4984, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.0008565155601261307, |
| "grad_norm": 0.53515625, |
| "learning_rate": 0.0002999996202109018, |
| "loss": 0.4111, |
| "step": 6110 |
| }, |
| { |
| "epoch": 0.0008579173859201178, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.00029999961872270205, |
| "loss": 0.4226, |
| "step": 6120 |
| }, |
| { |
| "epoch": 0.0008593192117141049, |
| "grad_norm": 0.70703125, |
| "learning_rate": 0.0002999996172315923, |
| "loss": 0.491, |
| "step": 6130 |
| }, |
| { |
| "epoch": 0.0008607210375080921, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00029999961573757243, |
| "loss": 0.4407, |
| "step": 6140 |
| }, |
| { |
| "epoch": 0.0008621228633020791, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.0002999996142406426, |
| "loss": 0.4048, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.0008635246890960663, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.00029999961274080266, |
| "loss": 0.3733, |
| "step": 6160 |
| }, |
| { |
| "epoch": 0.0008649265148900534, |
| "grad_norm": 0.58984375, |
| "learning_rate": 0.0002999996112380527, |
| "loss": 0.4596, |
| "step": 6170 |
| }, |
| { |
| "epoch": 0.0008663283406840405, |
| "grad_norm": 0.57421875, |
| "learning_rate": 0.0002999996097323927, |
| "loss": 0.4819, |
| "step": 6180 |
| }, |
| { |
| "epoch": 0.0008677301664780277, |
| "grad_norm": 0.30859375, |
| "learning_rate": 0.0002999996082238226, |
| "loss": 0.4029, |
| "step": 6190 |
| }, |
| { |
| "epoch": 0.0008691319922720148, |
| "grad_norm": 0.65234375, |
| "learning_rate": 0.00029999960671234254, |
| "loss": 0.5001, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.0008705338180660019, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.00029999960519795237, |
| "loss": 0.4356, |
| "step": 6210 |
| }, |
| { |
| "epoch": 0.000871935643859989, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.0002999996036806522, |
| "loss": 0.4399, |
| "step": 6220 |
| }, |
| { |
| "epoch": 0.0008733374696539762, |
| "grad_norm": 0.76171875, |
| "learning_rate": 0.00029999960216044196, |
| "loss": 0.5213, |
| "step": 6230 |
| }, |
| { |
| "epoch": 0.0008747392954479632, |
| "grad_norm": 1.1640625, |
| "learning_rate": 0.00029999960063732163, |
| "loss": 0.4467, |
| "step": 6240 |
| }, |
| { |
| "epoch": 0.0008761411212419504, |
| "grad_norm": 1.1015625, |
| "learning_rate": 0.00029999959911129135, |
| "loss": 0.5033, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.0008775429470359374, |
| "grad_norm": 0.330078125, |
| "learning_rate": 0.00029999959758235097, |
| "loss": 0.3733, |
| "step": 6260 |
| }, |
| { |
| "epoch": 0.0008789447728299246, |
| "grad_norm": 0.19921875, |
| "learning_rate": 0.0002999995960505005, |
| "loss": 0.3974, |
| "step": 6270 |
| }, |
| { |
| "epoch": 0.0008803465986239117, |
| "grad_norm": 0.33984375, |
| "learning_rate": 0.0002999995945157401, |
| "loss": 0.4224, |
| "step": 6280 |
| }, |
| { |
| "epoch": 0.0008817484244178988, |
| "grad_norm": 0.53515625, |
| "learning_rate": 0.00029999959297806955, |
| "loss": 0.4442, |
| "step": 6290 |
| }, |
| { |
| "epoch": 0.000883150250211886, |
| "grad_norm": 0.46484375, |
| "learning_rate": 0.000299999591437489, |
| "loss": 0.4996, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.0008845520760058731, |
| "grad_norm": 0.9375, |
| "learning_rate": 0.0002999995898939984, |
| "loss": 0.4069, |
| "step": 6310 |
| }, |
| { |
| "epoch": 0.0008859539017998602, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.00029999958834759776, |
| "loss": 0.4179, |
| "step": 6320 |
| }, |
| { |
| "epoch": 0.0008873557275938473, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.0002999995867982871, |
| "loss": 0.5006, |
| "step": 6330 |
| }, |
| { |
| "epoch": 0.0008887575533878345, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.00029999958524606636, |
| "loss": 0.46, |
| "step": 6340 |
| }, |
| { |
| "epoch": 0.0008901593791818215, |
| "grad_norm": 0.232421875, |
| "learning_rate": 0.0002999995836909356, |
| "loss": 0.4165, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.0008915612049758087, |
| "grad_norm": 0.98046875, |
| "learning_rate": 0.00029999958213289474, |
| "loss": 0.3791, |
| "step": 6360 |
| }, |
| { |
| "epoch": 0.0008929630307697958, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.0002999995805719439, |
| "loss": 0.3416, |
| "step": 6370 |
| }, |
| { |
| "epoch": 0.0008943648565637829, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.000299999579008083, |
| "loss": 0.4323, |
| "step": 6380 |
| }, |
| { |
| "epoch": 0.0008957666823577701, |
| "grad_norm": 0.4375, |
| "learning_rate": 0.000299999577441312, |
| "loss": 0.4569, |
| "step": 6390 |
| }, |
| { |
| "epoch": 0.0008971685081517572, |
| "grad_norm": 0.546875, |
| "learning_rate": 0.000299999575871631, |
| "loss": 0.4431, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.0008985703339457443, |
| "grad_norm": 1.1171875, |
| "learning_rate": 0.00029999957429904, |
| "loss": 0.5003, |
| "step": 6410 |
| }, |
| { |
| "epoch": 0.0008999721597397314, |
| "grad_norm": 0.85546875, |
| "learning_rate": 0.0002999995727235389, |
| "loss": 0.4411, |
| "step": 6420 |
| }, |
| { |
| "epoch": 0.0009013739855337186, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999995711451278, |
| "loss": 0.4253, |
| "step": 6430 |
| }, |
| { |
| "epoch": 0.0009027758113277056, |
| "grad_norm": 0.98828125, |
| "learning_rate": 0.0002999995695638066, |
| "loss": 0.4617, |
| "step": 6440 |
| }, |
| { |
| "epoch": 0.0009041776371216928, |
| "grad_norm": 0.4375, |
| "learning_rate": 0.0002999995679795754, |
| "loss": 0.4213, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.0009055794629156799, |
| "grad_norm": 1.3828125, |
| "learning_rate": 0.0002999995663924341, |
| "loss": 0.471, |
| "step": 6460 |
| }, |
| { |
| "epoch": 0.000906981288709667, |
| "grad_norm": 0.6015625, |
| "learning_rate": 0.00029999956480238277, |
| "loss": 0.4532, |
| "step": 6470 |
| }, |
| { |
| "epoch": 0.0009083831145036542, |
| "grad_norm": 0.9609375, |
| "learning_rate": 0.00029999956320942144, |
| "loss": 0.4206, |
| "step": 6480 |
| }, |
| { |
| "epoch": 0.0009097849402976413, |
| "grad_norm": 0.94921875, |
| "learning_rate": 0.00029999956161355006, |
| "loss": 0.5678, |
| "step": 6490 |
| }, |
| { |
| "epoch": 0.0009111867660916284, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.00029999956001476857, |
| "loss": 0.3739, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.0009111867660916284, |
| "eval_loss": 0.18671706318855286, |
| "eval_runtime": 113.805, |
| "eval_samples_per_second": 200.721, |
| "eval_steps_per_second": 1.573, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.0009125885918856155, |
| "grad_norm": 0.3515625, |
| "learning_rate": 0.00029999955841307713, |
| "loss": 0.4547, |
| "step": 6510 |
| }, |
| { |
| "epoch": 0.0009139904176796027, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.0002999995568084756, |
| "loss": 0.4481, |
| "step": 6520 |
| }, |
| { |
| "epoch": 0.0009153922434735897, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.000299999555200964, |
| "loss": 0.4814, |
| "step": 6530 |
| }, |
| { |
| "epoch": 0.0009167940692675769, |
| "grad_norm": 0.40625, |
| "learning_rate": 0.0002999995535905424, |
| "loss": 0.4352, |
| "step": 6540 |
| }, |
| { |
| "epoch": 0.0009181958950615639, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.0002999995519772107, |
| "loss": 0.4045, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.0009195977208555511, |
| "grad_norm": 0.6875, |
| "learning_rate": 0.000299999550360969, |
| "loss": 0.4183, |
| "step": 6560 |
| }, |
| { |
| "epoch": 0.0009209995466495382, |
| "grad_norm": 0.58984375, |
| "learning_rate": 0.00029999954874181724, |
| "loss": 0.3972, |
| "step": 6570 |
| }, |
| { |
| "epoch": 0.0009224013724435253, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.0002999995471197555, |
| "loss": 0.488, |
| "step": 6580 |
| }, |
| { |
| "epoch": 0.0009238031982375125, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.00029999954549478363, |
| "loss": 0.4826, |
| "step": 6590 |
| }, |
| { |
| "epoch": 0.0009252050240314996, |
| "grad_norm": 0.3671875, |
| "learning_rate": 0.0002999995438669017, |
| "loss": 0.3605, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.0009266068498254867, |
| "grad_norm": 0.255859375, |
| "learning_rate": 0.00029999954223610976, |
| "loss": 0.3911, |
| "step": 6610 |
| }, |
| { |
| "epoch": 0.0009280086756194738, |
| "grad_norm": 0.4140625, |
| "learning_rate": 0.0002999995406024078, |
| "loss": 0.5245, |
| "step": 6620 |
| }, |
| { |
| "epoch": 0.000929410501413461, |
| "grad_norm": 0.390625, |
| "learning_rate": 0.00029999953896579583, |
| "loss": 0.3988, |
| "step": 6630 |
| }, |
| { |
| "epoch": 0.000930812327207448, |
| "grad_norm": 0.50390625, |
| "learning_rate": 0.00029999953732627377, |
| "loss": 0.522, |
| "step": 6640 |
| }, |
| { |
| "epoch": 0.0009322141530014352, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.00029999953568384164, |
| "loss": 0.462, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.0009336159787954223, |
| "grad_norm": 0.6015625, |
| "learning_rate": 0.00029999953403849947, |
| "loss": 0.4395, |
| "step": 6660 |
| }, |
| { |
| "epoch": 0.0009350178045894094, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999995323902473, |
| "loss": 0.5009, |
| "step": 6670 |
| }, |
| { |
| "epoch": 0.0009364196303833966, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.00029999953073908507, |
| "loss": 0.4336, |
| "step": 6680 |
| }, |
| { |
| "epoch": 0.0009378214561773837, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999995290850128, |
| "loss": 0.4126, |
| "step": 6690 |
| }, |
| { |
| "epoch": 0.0009392232819713708, |
| "grad_norm": 0.359375, |
| "learning_rate": 0.00029999952742803046, |
| "loss": 0.4957, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.0009406251077653579, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.0002999995257681381, |
| "loss": 0.4758, |
| "step": 6710 |
| }, |
| { |
| "epoch": 0.0009420269335593451, |
| "grad_norm": 0.314453125, |
| "learning_rate": 0.0002999995241053357, |
| "loss": 0.3698, |
| "step": 6720 |
| }, |
| { |
| "epoch": 0.0009434287593533321, |
| "grad_norm": 0.462890625, |
| "learning_rate": 0.0002999995224396232, |
| "loss": 0.4371, |
| "step": 6730 |
| }, |
| { |
| "epoch": 0.0009448305851473193, |
| "grad_norm": 0.390625, |
| "learning_rate": 0.00029999952077100077, |
| "loss": 0.4538, |
| "step": 6740 |
| }, |
| { |
| "epoch": 0.0009462324109413064, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999995190994682, |
| "loss": 0.4969, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.0009476342367352935, |
| "grad_norm": 1.2109375, |
| "learning_rate": 0.0002999995174250256, |
| "loss": 0.4484, |
| "step": 6760 |
| }, |
| { |
| "epoch": 0.0009490360625292807, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.000299999515747673, |
| "loss": 0.4748, |
| "step": 6770 |
| }, |
| { |
| "epoch": 0.0009504378883232677, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.00029999951406741034, |
| "loss": 0.5366, |
| "step": 6780 |
| }, |
| { |
| "epoch": 0.0009518397141172549, |
| "grad_norm": 0.26171875, |
| "learning_rate": 0.0002999995123842376, |
| "loss": 0.4718, |
| "step": 6790 |
| }, |
| { |
| "epoch": 0.000953241539911242, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.00029999951069815484, |
| "loss": 0.3908, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.0009546433657052291, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.00029999950900916203, |
| "loss": 0.4005, |
| "step": 6810 |
| }, |
| { |
| "epoch": 0.0009560451914992162, |
| "grad_norm": 0.515625, |
| "learning_rate": 0.0002999995073172592, |
| "loss": 0.4265, |
| "step": 6820 |
| }, |
| { |
| "epoch": 0.0009574470172932034, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.0002999995056224463, |
| "loss": 0.4673, |
| "step": 6830 |
| }, |
| { |
| "epoch": 0.0009588488430871904, |
| "grad_norm": 0.28515625, |
| "learning_rate": 0.00029999950392472336, |
| "loss": 0.3959, |
| "step": 6840 |
| }, |
| { |
| "epoch": 0.0009602506688811776, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.0002999995022240904, |
| "loss": 0.5393, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.0009616524946751647, |
| "grad_norm": 0.828125, |
| "learning_rate": 0.0002999995005205474, |
| "loss": 0.4281, |
| "step": 6860 |
| }, |
| { |
| "epoch": 0.0009630543204691518, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.0002999994988140943, |
| "loss": 0.378, |
| "step": 6870 |
| }, |
| { |
| "epoch": 0.000964456146263139, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.0002999994971047312, |
| "loss": 0.445, |
| "step": 6880 |
| }, |
| { |
| "epoch": 0.0009658579720571261, |
| "grad_norm": 0.57421875, |
| "learning_rate": 0.0002999994953924581, |
| "loss": 0.4868, |
| "step": 6890 |
| }, |
| { |
| "epoch": 0.0009672597978511132, |
| "grad_norm": 0.93359375, |
| "learning_rate": 0.0002999994936772749, |
| "loss": 0.479, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.0009686616236451003, |
| "grad_norm": 0.921875, |
| "learning_rate": 0.0002999994919591817, |
| "loss": 0.5115, |
| "step": 6910 |
| }, |
| { |
| "epoch": 0.0009700634494390875, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.0002999994902381784, |
| "loss": 0.4955, |
| "step": 6920 |
| }, |
| { |
| "epoch": 0.0009714652752330745, |
| "grad_norm": 1.34375, |
| "learning_rate": 0.0002999994885142651, |
| "loss": 0.5283, |
| "step": 6930 |
| }, |
| { |
| "epoch": 0.0009728671010270617, |
| "grad_norm": 0.458984375, |
| "learning_rate": 0.00029999948678744175, |
| "loss": 0.5511, |
| "step": 6940 |
| }, |
| { |
| "epoch": 0.0009742689268210488, |
| "grad_norm": 0.7109375, |
| "learning_rate": 0.0002999994850577083, |
| "loss": 0.4193, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.0009756707526150359, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.0002999994833250649, |
| "loss": 0.4947, |
| "step": 6960 |
| }, |
| { |
| "epoch": 0.000977072578409023, |
| "grad_norm": 0.4140625, |
| "learning_rate": 0.0002999994815895114, |
| "loss": 0.3735, |
| "step": 6970 |
| }, |
| { |
| "epoch": 0.0009784744042030103, |
| "grad_norm": 0.34765625, |
| "learning_rate": 0.00029999947985104785, |
| "loss": 0.4227, |
| "step": 6980 |
| }, |
| { |
| "epoch": 0.0009798762299969973, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.00029999947810967426, |
| "loss": 0.3318, |
| "step": 6990 |
| }, |
| { |
| "epoch": 0.0009812780557909844, |
| "grad_norm": 0.80078125, |
| "learning_rate": 0.0002999994763653906, |
| "loss": 0.4498, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.0009812780557909844, |
| "eval_loss": 0.1828116476535797, |
| "eval_runtime": 112.0643, |
| "eval_samples_per_second": 203.838, |
| "eval_steps_per_second": 1.597, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.0009826798815849714, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.00029999947461819697, |
| "loss": 0.4011, |
| "step": 7010 |
| }, |
| { |
| "epoch": 0.0009840817073789587, |
| "grad_norm": 1.2109375, |
| "learning_rate": 0.00029999947286809327, |
| "loss": 0.5584, |
| "step": 7020 |
| }, |
| { |
| "epoch": 0.0009854835331729458, |
| "grad_norm": 0.4296875, |
| "learning_rate": 0.0002999994711150795, |
| "loss": 0.4164, |
| "step": 7030 |
| }, |
| { |
| "epoch": 0.0009868853589669328, |
| "grad_norm": 0.58984375, |
| "learning_rate": 0.0002999994693591557, |
| "loss": 0.4351, |
| "step": 7040 |
| }, |
| { |
| "epoch": 0.00098828718476092, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.0002999994676003219, |
| "loss": 0.3533, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.0009896890105549072, |
| "grad_norm": 0.423828125, |
| "learning_rate": 0.00029999946583857806, |
| "loss": 0.4176, |
| "step": 7060 |
| }, |
| { |
| "epoch": 0.0009910908363488942, |
| "grad_norm": 0.412109375, |
| "learning_rate": 0.0002999994640739241, |
| "loss": 0.4999, |
| "step": 7070 |
| }, |
| { |
| "epoch": 0.0009924926621428813, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.00029999946230636014, |
| "loss": 0.4877, |
| "step": 7080 |
| }, |
| { |
| "epoch": 0.0009938944879368686, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.0002999994605358862, |
| "loss": 0.473, |
| "step": 7090 |
| }, |
| { |
| "epoch": 0.0009952963137308556, |
| "grad_norm": 1.0, |
| "learning_rate": 0.0002999994587625022, |
| "loss": 0.4986, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.0009966981395248427, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.00029999945698620806, |
| "loss": 0.3922, |
| "step": 7110 |
| }, |
| { |
| "epoch": 0.0009980999653188298, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.00029999945520700394, |
| "loss": 0.4771, |
| "step": 7120 |
| }, |
| { |
| "epoch": 0.000999501791112817, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.0002999994534248897, |
| "loss": 0.441, |
| "step": 7130 |
| }, |
| { |
| "epoch": 0.001000903616906804, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.00029999945163986555, |
| "loss": 0.5059, |
| "step": 7140 |
| }, |
| { |
| "epoch": 0.0010023054427007912, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.00029999944985193127, |
| "loss": 0.465, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.0010037072684947784, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.000299999448061087, |
| "loss": 0.4113, |
| "step": 7160 |
| }, |
| { |
| "epoch": 0.0010051090942887655, |
| "grad_norm": 0.953125, |
| "learning_rate": 0.00029999944626733267, |
| "loss": 0.5342, |
| "step": 7170 |
| }, |
| { |
| "epoch": 0.0010065109200827526, |
| "grad_norm": 0.56640625, |
| "learning_rate": 0.0002999994444706683, |
| "loss": 0.531, |
| "step": 7180 |
| }, |
| { |
| "epoch": 0.0010079127458767396, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.00029999944267109386, |
| "loss": 0.4291, |
| "step": 7190 |
| }, |
| { |
| "epoch": 0.001009314571670727, |
| "grad_norm": 0.91796875, |
| "learning_rate": 0.00029999944086860937, |
| "loss": 0.56, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.001010716397464714, |
| "grad_norm": 1.0859375, |
| "learning_rate": 0.0002999994390632149, |
| "loss": 0.4209, |
| "step": 7210 |
| }, |
| { |
| "epoch": 0.001012118223258701, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.0002999994372549103, |
| "loss": 0.5099, |
| "step": 7220 |
| }, |
| { |
| "epoch": 0.001013520049052688, |
| "grad_norm": 1.0625, |
| "learning_rate": 0.00029999943544369576, |
| "loss": 0.567, |
| "step": 7230 |
| }, |
| { |
| "epoch": 0.0010149218748466754, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.0002999994336295711, |
| "loss": 0.474, |
| "step": 7240 |
| }, |
| { |
| "epoch": 0.0010163237006406624, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.0002999994318125364, |
| "loss": 0.4313, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.0010177255264346495, |
| "grad_norm": 0.30078125, |
| "learning_rate": 0.00029999942999259173, |
| "loss": 0.4166, |
| "step": 7260 |
| }, |
| { |
| "epoch": 0.0010191273522286368, |
| "grad_norm": 0.9296875, |
| "learning_rate": 0.000299999428169737, |
| "loss": 0.3713, |
| "step": 7270 |
| }, |
| { |
| "epoch": 0.0010205291780226238, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.0002999994263439722, |
| "loss": 0.4229, |
| "step": 7280 |
| }, |
| { |
| "epoch": 0.0010219310038166109, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.00029999942451529733, |
| "loss": 0.4572, |
| "step": 7290 |
| }, |
| { |
| "epoch": 0.001023332829610598, |
| "grad_norm": 0.6015625, |
| "learning_rate": 0.0002999994226837125, |
| "loss": 0.4676, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.0010247346554045852, |
| "grad_norm": 0.85546875, |
| "learning_rate": 0.0002999994208492175, |
| "loss": 0.4612, |
| "step": 7310 |
| }, |
| { |
| "epoch": 0.0010261364811985723, |
| "grad_norm": 0.95703125, |
| "learning_rate": 0.00029999941901181256, |
| "loss": 0.4747, |
| "step": 7320 |
| }, |
| { |
| "epoch": 0.0010275383069925593, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.00029999941717149755, |
| "loss": 0.4067, |
| "step": 7330 |
| }, |
| { |
| "epoch": 0.0010289401327865464, |
| "grad_norm": 0.5078125, |
| "learning_rate": 0.0002999994153282725, |
| "loss": 0.433, |
| "step": 7340 |
| }, |
| { |
| "epoch": 0.0010303419585805337, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.00029999941348213743, |
| "loss": 0.4342, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.0010317437843745207, |
| "grad_norm": 0.52734375, |
| "learning_rate": 0.0002999994116330923, |
| "loss": 0.5533, |
| "step": 7360 |
| }, |
| { |
| "epoch": 0.0010331456101685078, |
| "grad_norm": 1.140625, |
| "learning_rate": 0.00029999940978113715, |
| "loss": 0.5318, |
| "step": 7370 |
| }, |
| { |
| "epoch": 0.001034547435962495, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.00029999940792627193, |
| "loss": 0.4702, |
| "step": 7380 |
| }, |
| { |
| "epoch": 0.0010359492617564821, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.00029999940606849666, |
| "loss": 0.4446, |
| "step": 7390 |
| }, |
| { |
| "epoch": 0.0010373510875504692, |
| "grad_norm": 0.58984375, |
| "learning_rate": 0.00029999940420781133, |
| "loss": 0.4316, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.0010387529133444563, |
| "grad_norm": 0.498046875, |
| "learning_rate": 0.000299999402344216, |
| "loss": 0.378, |
| "step": 7410 |
| }, |
| { |
| "epoch": 0.0010401547391384435, |
| "grad_norm": 0.56640625, |
| "learning_rate": 0.00029999940047771063, |
| "loss": 0.4685, |
| "step": 7420 |
| }, |
| { |
| "epoch": 0.0010415565649324306, |
| "grad_norm": 0.2255859375, |
| "learning_rate": 0.0002999993986082952, |
| "loss": 0.4783, |
| "step": 7430 |
| }, |
| { |
| "epoch": 0.0010429583907264177, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.00029999939673596977, |
| "loss": 0.4002, |
| "step": 7440 |
| }, |
| { |
| "epoch": 0.001044360216520405, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.0002999993948607343, |
| "loss": 0.473, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.001045762042314392, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.0002999993929825887, |
| "loss": 0.4573, |
| "step": 7460 |
| }, |
| { |
| "epoch": 0.001047163868108379, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999993911015331, |
| "loss": 0.4299, |
| "step": 7470 |
| }, |
| { |
| "epoch": 0.0010485656939023661, |
| "grad_norm": 0.546875, |
| "learning_rate": 0.00029999938921756753, |
| "loss": 0.4994, |
| "step": 7480 |
| }, |
| { |
| "epoch": 0.0010499675196963534, |
| "grad_norm": 0.443359375, |
| "learning_rate": 0.00029999938733069184, |
| "loss": 0.4764, |
| "step": 7490 |
| }, |
| { |
| "epoch": 0.0010513693454903405, |
| "grad_norm": 0.1767578125, |
| "learning_rate": 0.00029999938544090615, |
| "loss": 0.3844, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.0010513693454903405, |
| "eval_loss": 0.18380461633205414, |
| "eval_runtime": 114.1658, |
| "eval_samples_per_second": 200.086, |
| "eval_steps_per_second": 1.568, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.0010527711712843275, |
| "grad_norm": 0.4921875, |
| "learning_rate": 0.0002999993835482104, |
| "loss": 0.4668, |
| "step": 7510 |
| }, |
| { |
| "epoch": 0.0010541729970783146, |
| "grad_norm": 0.41796875, |
| "learning_rate": 0.0002999993816526046, |
| "loss": 0.4184, |
| "step": 7520 |
| }, |
| { |
| "epoch": 0.0010555748228723019, |
| "grad_norm": 0.9296875, |
| "learning_rate": 0.00029999937975408876, |
| "loss": 0.4809, |
| "step": 7530 |
| }, |
| { |
| "epoch": 0.001056976648666289, |
| "grad_norm": 0.80078125, |
| "learning_rate": 0.0002999993778526629, |
| "loss": 0.5232, |
| "step": 7540 |
| }, |
| { |
| "epoch": 0.001058378474460276, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.00029999937594832695, |
| "loss": 0.4522, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.0010597803002542633, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.00029999937404108105, |
| "loss": 0.4973, |
| "step": 7560 |
| }, |
| { |
| "epoch": 0.0010611821260482503, |
| "grad_norm": 0.435546875, |
| "learning_rate": 0.00029999937213092505, |
| "loss": 0.4475, |
| "step": 7570 |
| }, |
| { |
| "epoch": 0.0010625839518422374, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.000299999370217859, |
| "loss": 0.5236, |
| "step": 7580 |
| }, |
| { |
| "epoch": 0.0010639857776362244, |
| "grad_norm": 0.9921875, |
| "learning_rate": 0.000299999368301883, |
| "loss": 0.4271, |
| "step": 7590 |
| }, |
| { |
| "epoch": 0.0010653876034302117, |
| "grad_norm": 0.93359375, |
| "learning_rate": 0.0002999993663829968, |
| "loss": 0.4317, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.0010667894292241988, |
| "grad_norm": 0.4375, |
| "learning_rate": 0.00029999936446120065, |
| "loss": 0.3837, |
| "step": 7610 |
| }, |
| { |
| "epoch": 0.0010681912550181858, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.00029999936253649444, |
| "loss": 0.4802, |
| "step": 7620 |
| }, |
| { |
| "epoch": 0.001069593080812173, |
| "grad_norm": 0.24609375, |
| "learning_rate": 0.0002999993606088783, |
| "loss": 0.4218, |
| "step": 7630 |
| }, |
| { |
| "epoch": 0.0010709949066061602, |
| "grad_norm": 0.419921875, |
| "learning_rate": 0.000299999358678352, |
| "loss": 0.3988, |
| "step": 7640 |
| }, |
| { |
| "epoch": 0.0010723967324001472, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.00029999935674491563, |
| "loss": 0.4898, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.0010737985581941343, |
| "grad_norm": 0.734375, |
| "learning_rate": 0.0002999993548085693, |
| "loss": 0.5254, |
| "step": 7660 |
| }, |
| { |
| "epoch": 0.0010752003839881216, |
| "grad_norm": 0.365234375, |
| "learning_rate": 0.0002999993528693129, |
| "loss": 0.3633, |
| "step": 7670 |
| }, |
| { |
| "epoch": 0.0010766022097821086, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.00029999935092714646, |
| "loss": 0.4217, |
| "step": 7680 |
| }, |
| { |
| "epoch": 0.0010780040355760957, |
| "grad_norm": 0.45703125, |
| "learning_rate": 0.00029999934898207003, |
| "loss": 0.4005, |
| "step": 7690 |
| }, |
| { |
| "epoch": 0.0010794058613700828, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.00029999934703408345, |
| "loss": 0.5546, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.00108080768716407, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.0002999993450831869, |
| "loss": 0.5362, |
| "step": 7710 |
| }, |
| { |
| "epoch": 0.001082209512958057, |
| "grad_norm": 0.53125, |
| "learning_rate": 0.00029999934312938034, |
| "loss": 0.4965, |
| "step": 7720 |
| }, |
| { |
| "epoch": 0.0010836113387520442, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.0002999993411726637, |
| "loss": 0.4596, |
| "step": 7730 |
| }, |
| { |
| "epoch": 0.0010850131645460314, |
| "grad_norm": 0.486328125, |
| "learning_rate": 0.000299999339213037, |
| "loss": 0.4036, |
| "step": 7740 |
| }, |
| { |
| "epoch": 0.0010864149903400185, |
| "grad_norm": 0.310546875, |
| "learning_rate": 0.00029999933725050027, |
| "loss": 0.3324, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.0010878168161340056, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.00029999933528505353, |
| "loss": 0.3971, |
| "step": 7760 |
| }, |
| { |
| "epoch": 0.0010892186419279926, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.00029999933331669674, |
| "loss": 0.4291, |
| "step": 7770 |
| }, |
| { |
| "epoch": 0.00109062046772198, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999993313454299, |
| "loss": 0.5129, |
| "step": 7780 |
| }, |
| { |
| "epoch": 0.001092022293515967, |
| "grad_norm": 0.828125, |
| "learning_rate": 0.00029999932937125305, |
| "loss": 0.4775, |
| "step": 7790 |
| }, |
| { |
| "epoch": 0.001093424119309954, |
| "grad_norm": 0.41796875, |
| "learning_rate": 0.0002999993273941661, |
| "loss": 0.3819, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.001094825945103941, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.00029999932541416915, |
| "loss": 0.421, |
| "step": 7810 |
| }, |
| { |
| "epoch": 0.0010962277708979284, |
| "grad_norm": 0.255859375, |
| "learning_rate": 0.00029999932343126214, |
| "loss": 0.3671, |
| "step": 7820 |
| }, |
| { |
| "epoch": 0.0010976295966919154, |
| "grad_norm": 0.375, |
| "learning_rate": 0.0002999993214454451, |
| "loss": 0.3949, |
| "step": 7830 |
| }, |
| { |
| "epoch": 0.0010990314224859025, |
| "grad_norm": 0.3125, |
| "learning_rate": 0.00029999931945671804, |
| "loss": 0.4246, |
| "step": 7840 |
| }, |
| { |
| "epoch": 0.0011004332482798898, |
| "grad_norm": 0.353515625, |
| "learning_rate": 0.00029999931746508093, |
| "loss": 0.3755, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.0011018350740738768, |
| "grad_norm": 0.328125, |
| "learning_rate": 0.0002999993154705338, |
| "loss": 0.3755, |
| "step": 7860 |
| }, |
| { |
| "epoch": 0.0011032368998678639, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.0002999993134730766, |
| "loss": 0.382, |
| "step": 7870 |
| }, |
| { |
| "epoch": 0.001104638725661851, |
| "grad_norm": 0.515625, |
| "learning_rate": 0.00029999931147270934, |
| "loss": 0.4672, |
| "step": 7880 |
| }, |
| { |
| "epoch": 0.0011060405514558382, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.0002999993094694321, |
| "loss": 0.3952, |
| "step": 7890 |
| }, |
| { |
| "epoch": 0.0011074423772498253, |
| "grad_norm": 0.95703125, |
| "learning_rate": 0.00029999930746324476, |
| "loss": 0.4959, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.0011088442030438123, |
| "grad_norm": 0.28125, |
| "learning_rate": 0.0002999993054541474, |
| "loss": 0.3773, |
| "step": 7910 |
| }, |
| { |
| "epoch": 0.0011102460288377994, |
| "grad_norm": 0.46875, |
| "learning_rate": 0.00029999930344214, |
| "loss": 0.4186, |
| "step": 7920 |
| }, |
| { |
| "epoch": 0.0011116478546317867, |
| "grad_norm": 0.6953125, |
| "learning_rate": 0.0002999993014272226, |
| "loss": 0.4753, |
| "step": 7930 |
| }, |
| { |
| "epoch": 0.0011130496804257737, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.00029999929940939513, |
| "loss": 0.5497, |
| "step": 7940 |
| }, |
| { |
| "epoch": 0.0011144515062197608, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.0002999992973886576, |
| "loss": 0.4491, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.001115853332013748, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.0002999992953650101, |
| "loss": 0.3838, |
| "step": 7960 |
| }, |
| { |
| "epoch": 0.0011172551578077351, |
| "grad_norm": 0.5546875, |
| "learning_rate": 0.0002999992933384525, |
| "loss": 0.4518, |
| "step": 7970 |
| }, |
| { |
| "epoch": 0.0011186569836017222, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.0002999992913089848, |
| "loss": 0.4592, |
| "step": 7980 |
| }, |
| { |
| "epoch": 0.0011200588093957093, |
| "grad_norm": 0.333984375, |
| "learning_rate": 0.0002999992892766072, |
| "loss": 0.438, |
| "step": 7990 |
| }, |
| { |
| "epoch": 0.0011214606351896965, |
| "grad_norm": 0.2314453125, |
| "learning_rate": 0.0002999992872413195, |
| "loss": 0.4245, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.0011214606351896965, |
| "eval_loss": 0.18062791228294373, |
| "eval_runtime": 112.1678, |
| "eval_samples_per_second": 203.65, |
| "eval_steps_per_second": 1.596, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.0011228624609836836, |
| "grad_norm": 0.85546875, |
| "learning_rate": 0.00029999928520312177, |
| "loss": 0.3692, |
| "step": 8010 |
| }, |
| { |
| "epoch": 0.0011242642867776707, |
| "grad_norm": 0.189453125, |
| "learning_rate": 0.0002999992831620139, |
| "loss": 0.4953, |
| "step": 8020 |
| }, |
| { |
| "epoch": 0.001125666112571658, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.00029999928111799614, |
| "loss": 0.4422, |
| "step": 8030 |
| }, |
| { |
| "epoch": 0.001127067938365645, |
| "grad_norm": 0.443359375, |
| "learning_rate": 0.0002999992790710683, |
| "loss": 0.428, |
| "step": 8040 |
| }, |
| { |
| "epoch": 0.001128469764159632, |
| "grad_norm": 0.96875, |
| "learning_rate": 0.0002999992770212304, |
| "loss": 0.4894, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.0011298715899536191, |
| "grad_norm": 0.421875, |
| "learning_rate": 0.00029999927496848246, |
| "loss": 0.4164, |
| "step": 8060 |
| }, |
| { |
| "epoch": 0.0011312734157476064, |
| "grad_norm": 0.384765625, |
| "learning_rate": 0.00029999927291282446, |
| "loss": 0.4902, |
| "step": 8070 |
| }, |
| { |
| "epoch": 0.0011326752415415935, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.00029999927085425646, |
| "loss": 0.4481, |
| "step": 8080 |
| }, |
| { |
| "epoch": 0.0011340770673355805, |
| "grad_norm": 1.140625, |
| "learning_rate": 0.0002999992687927784, |
| "loss": 0.3717, |
| "step": 8090 |
| }, |
| { |
| "epoch": 0.0011354788931295676, |
| "grad_norm": 0.345703125, |
| "learning_rate": 0.0002999992667283903, |
| "loss": 0.4978, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.0011368807189235549, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.00029999926466109215, |
| "loss": 0.4372, |
| "step": 8110 |
| }, |
| { |
| "epoch": 0.001138282544717542, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.000299999262590884, |
| "loss": 0.4048, |
| "step": 8120 |
| }, |
| { |
| "epoch": 0.001139684370511529, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.0002999992605177658, |
| "loss": 0.4937, |
| "step": 8130 |
| }, |
| { |
| "epoch": 0.0011410861963055163, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.0002999992584417376, |
| "loss": 0.4527, |
| "step": 8140 |
| }, |
| { |
| "epoch": 0.0011424880220995033, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999992563627993, |
| "loss": 0.502, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.0011438898478934904, |
| "grad_norm": 0.50390625, |
| "learning_rate": 0.0002999992542809509, |
| "loss": 0.4664, |
| "step": 8160 |
| }, |
| { |
| "epoch": 0.0011452916736874774, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.0002999992521961926, |
| "loss": 0.5214, |
| "step": 8170 |
| }, |
| { |
| "epoch": 0.0011466934994814647, |
| "grad_norm": 0.5546875, |
| "learning_rate": 0.0002999992501085242, |
| "loss": 0.4809, |
| "step": 8180 |
| }, |
| { |
| "epoch": 0.0011480953252754518, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.00029999924801794577, |
| "loss": 0.4875, |
| "step": 8190 |
| }, |
| { |
| "epoch": 0.0011494971510694388, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.0002999992459244573, |
| "loss": 0.4272, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.001150898976863426, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.0002999992438280588, |
| "loss": 0.5558, |
| "step": 8210 |
| }, |
| { |
| "epoch": 0.0011523008026574132, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.0002999992417287502, |
| "loss": 0.3699, |
| "step": 8220 |
| }, |
| { |
| "epoch": 0.0011537026284514002, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.00029999923962653163, |
| "loss": 0.4134, |
| "step": 8230 |
| }, |
| { |
| "epoch": 0.0011551044542453873, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.000299999237521403, |
| "loss": 0.3791, |
| "step": 8240 |
| }, |
| { |
| "epoch": 0.0011565062800393746, |
| "grad_norm": 0.57421875, |
| "learning_rate": 0.0002999992354133643, |
| "loss": 0.3503, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.0011579081058333616, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.00029999923330241565, |
| "loss": 0.4115, |
| "step": 8260 |
| }, |
| { |
| "epoch": 0.0011593099316273487, |
| "grad_norm": 0.240234375, |
| "learning_rate": 0.00029999923118855686, |
| "loss": 0.3909, |
| "step": 8270 |
| }, |
| { |
| "epoch": 0.0011607117574213358, |
| "grad_norm": 0.177734375, |
| "learning_rate": 0.0002999992290717881, |
| "loss": 0.4405, |
| "step": 8280 |
| }, |
| { |
| "epoch": 0.001162113583215323, |
| "grad_norm": 1.046875, |
| "learning_rate": 0.0002999992269521093, |
| "loss": 0.4961, |
| "step": 8290 |
| }, |
| { |
| "epoch": 0.00116351540900931, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.0002999992248295204, |
| "loss": 0.3559, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.0011649172348032972, |
| "grad_norm": 0.494140625, |
| "learning_rate": 0.0002999992227040215, |
| "loss": 0.4841, |
| "step": 8310 |
| }, |
| { |
| "epoch": 0.0011663190605972844, |
| "grad_norm": 0.34375, |
| "learning_rate": 0.0002999992205756126, |
| "loss": 0.4471, |
| "step": 8320 |
| }, |
| { |
| "epoch": 0.0011677208863912715, |
| "grad_norm": 0.671875, |
| "learning_rate": 0.00029999921844429364, |
| "loss": 0.4465, |
| "step": 8330 |
| }, |
| { |
| "epoch": 0.0011691227121852586, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.00029999921631006465, |
| "loss": 0.4335, |
| "step": 8340 |
| }, |
| { |
| "epoch": 0.0011705245379792456, |
| "grad_norm": 0.478515625, |
| "learning_rate": 0.0002999992141729256, |
| "loss": 0.4393, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.0011719263637732329, |
| "grad_norm": 1.0859375, |
| "learning_rate": 0.0002999992120328765, |
| "loss": 0.451, |
| "step": 8360 |
| }, |
| { |
| "epoch": 0.00117332818956722, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.0002999992098899174, |
| "loss": 0.432, |
| "step": 8370 |
| }, |
| { |
| "epoch": 0.001174730015361207, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999992077440482, |
| "loss": 0.4606, |
| "step": 8380 |
| }, |
| { |
| "epoch": 0.001176131841155194, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.00029999920559526907, |
| "loss": 0.5024, |
| "step": 8390 |
| }, |
| { |
| "epoch": 0.0011775336669491814, |
| "grad_norm": 0.2314453125, |
| "learning_rate": 0.0002999992034435798, |
| "loss": 0.3843, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.0011789354927431684, |
| "grad_norm": 0.361328125, |
| "learning_rate": 0.0002999992012889805, |
| "loss": 0.4505, |
| "step": 8410 |
| }, |
| { |
| "epoch": 0.0011803373185371555, |
| "grad_norm": 1.0546875, |
| "learning_rate": 0.0002999991991314712, |
| "loss": 0.5677, |
| "step": 8420 |
| }, |
| { |
| "epoch": 0.0011817391443311428, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.0002999991969710519, |
| "loss": 0.4781, |
| "step": 8430 |
| }, |
| { |
| "epoch": 0.0011831409701251298, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.00029999919480772254, |
| "loss": 0.4487, |
| "step": 8440 |
| }, |
| { |
| "epoch": 0.0011845427959191169, |
| "grad_norm": 0.478515625, |
| "learning_rate": 0.00029999919264148307, |
| "loss": 0.4665, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.001185944621713104, |
| "grad_norm": 0.984375, |
| "learning_rate": 0.00029999919047233366, |
| "loss": 0.5278, |
| "step": 8460 |
| }, |
| { |
| "epoch": 0.0011873464475070912, |
| "grad_norm": 0.953125, |
| "learning_rate": 0.00029999918830027415, |
| "loss": 0.4261, |
| "step": 8470 |
| }, |
| { |
| "epoch": 0.0011887482733010783, |
| "grad_norm": 0.26171875, |
| "learning_rate": 0.0002999991861253046, |
| "loss": 0.4043, |
| "step": 8480 |
| }, |
| { |
| "epoch": 0.0011901500990950653, |
| "grad_norm": 0.91796875, |
| "learning_rate": 0.00029999918394742507, |
| "loss": 0.4822, |
| "step": 8490 |
| }, |
| { |
| "epoch": 0.0011915519248890524, |
| "grad_norm": 0.75, |
| "learning_rate": 0.00029999918176663545, |
| "loss": 0.5023, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.0011915519248890524, |
| "eval_loss": 0.18091367185115814, |
| "eval_runtime": 112.0263, |
| "eval_samples_per_second": 203.907, |
| "eval_steps_per_second": 1.598, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.0011929537506830397, |
| "grad_norm": 0.2490234375, |
| "learning_rate": 0.00029999917958293583, |
| "loss": 0.4, |
| "step": 8510 |
| }, |
| { |
| "epoch": 0.0011943555764770267, |
| "grad_norm": 0.32421875, |
| "learning_rate": 0.0002999991773963261, |
| "loss": 0.3551, |
| "step": 8520 |
| }, |
| { |
| "epoch": 0.0011957574022710138, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.00029999917520680643, |
| "loss": 0.3701, |
| "step": 8530 |
| }, |
| { |
| "epoch": 0.001197159228065001, |
| "grad_norm": 0.31640625, |
| "learning_rate": 0.00029999917301437665, |
| "loss": 0.4104, |
| "step": 8540 |
| }, |
| { |
| "epoch": 0.0011985610538589881, |
| "grad_norm": 0.2578125, |
| "learning_rate": 0.0002999991708190369, |
| "loss": 0.4813, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.0011999628796529752, |
| "grad_norm": 0.310546875, |
| "learning_rate": 0.00029999916862078705, |
| "loss": 0.3847, |
| "step": 8560 |
| }, |
| { |
| "epoch": 0.0012013647054469623, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.0002999991664196272, |
| "loss": 0.4201, |
| "step": 8570 |
| }, |
| { |
| "epoch": 0.0012027665312409495, |
| "grad_norm": 0.59375, |
| "learning_rate": 0.00029999916421555734, |
| "loss": 0.4336, |
| "step": 8580 |
| }, |
| { |
| "epoch": 0.0012041683570349366, |
| "grad_norm": 0.451171875, |
| "learning_rate": 0.0002999991620085774, |
| "loss": 0.5009, |
| "step": 8590 |
| }, |
| { |
| "epoch": 0.0012055701828289237, |
| "grad_norm": 0.443359375, |
| "learning_rate": 0.00029999915979868737, |
| "loss": 0.4635, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.001206972008622911, |
| "grad_norm": 0.28515625, |
| "learning_rate": 0.0002999991575858874, |
| "loss": 0.3898, |
| "step": 8610 |
| }, |
| { |
| "epoch": 0.001208373834416898, |
| "grad_norm": 0.26953125, |
| "learning_rate": 0.00029999915537017735, |
| "loss": 0.35, |
| "step": 8620 |
| }, |
| { |
| "epoch": 0.001209775660210885, |
| "grad_norm": 0.46484375, |
| "learning_rate": 0.00029999915315155726, |
| "loss": 0.4431, |
| "step": 8630 |
| }, |
| { |
| "epoch": 0.0012111774860048721, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.00029999915093002717, |
| "loss": 0.4535, |
| "step": 8640 |
| }, |
| { |
| "epoch": 0.0012125793117988594, |
| "grad_norm": 0.455078125, |
| "learning_rate": 0.00029999914870558703, |
| "loss": 0.3979, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.0012139811375928464, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.00029999914647823683, |
| "loss": 0.4376, |
| "step": 8660 |
| }, |
| { |
| "epoch": 0.0012153829633868335, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.0002999991442479766, |
| "loss": 0.482, |
| "step": 8670 |
| }, |
| { |
| "epoch": 0.0012167847891808206, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.00029999914201480634, |
| "loss": 0.4178, |
| "step": 8680 |
| }, |
| { |
| "epoch": 0.0012181866149748078, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.00029999913977872604, |
| "loss": 0.4327, |
| "step": 8690 |
| }, |
| { |
| "epoch": 0.001219588440768795, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.0002999991375397357, |
| "loss": 0.4699, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.001220990266562782, |
| "grad_norm": 1.0078125, |
| "learning_rate": 0.00029999913529783534, |
| "loss": 0.5145, |
| "step": 8710 |
| }, |
| { |
| "epoch": 0.0012223920923567692, |
| "grad_norm": 0.5546875, |
| "learning_rate": 0.00029999913305302494, |
| "loss": 0.4242, |
| "step": 8720 |
| }, |
| { |
| "epoch": 0.0012237939181507563, |
| "grad_norm": 0.421875, |
| "learning_rate": 0.0002999991308053045, |
| "loss": 0.3744, |
| "step": 8730 |
| }, |
| { |
| "epoch": 0.0012251957439447434, |
| "grad_norm": 0.4375, |
| "learning_rate": 0.000299999128554674, |
| "loss": 0.4844, |
| "step": 8740 |
| }, |
| { |
| "epoch": 0.0012265975697387304, |
| "grad_norm": 0.57421875, |
| "learning_rate": 0.0002999991263011335, |
| "loss": 0.4545, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.0012279993955327177, |
| "grad_norm": 0.99609375, |
| "learning_rate": 0.00029999912404468296, |
| "loss": 0.4804, |
| "step": 8760 |
| }, |
| { |
| "epoch": 0.0012294012213267048, |
| "grad_norm": 0.875, |
| "learning_rate": 0.0002999991217853224, |
| "loss": 0.47, |
| "step": 8770 |
| }, |
| { |
| "epoch": 0.0012308030471206918, |
| "grad_norm": 0.99609375, |
| "learning_rate": 0.00029999911952305173, |
| "loss": 0.4163, |
| "step": 8780 |
| }, |
| { |
| "epoch": 0.0012322048729146789, |
| "grad_norm": 0.65234375, |
| "learning_rate": 0.0002999991172578711, |
| "loss": 0.5008, |
| "step": 8790 |
| }, |
| { |
| "epoch": 0.0012336066987086662, |
| "grad_norm": 1.0546875, |
| "learning_rate": 0.0002999991149897804, |
| "loss": 0.5043, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.0012350085245026532, |
| "grad_norm": 1.15625, |
| "learning_rate": 0.0002999991127187797, |
| "loss": 0.4966, |
| "step": 8810 |
| }, |
| { |
| "epoch": 0.0012364103502966403, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.0002999991104448689, |
| "loss": 0.4113, |
| "step": 8820 |
| }, |
| { |
| "epoch": 0.0012378121760906276, |
| "grad_norm": 0.76171875, |
| "learning_rate": 0.0002999991081680481, |
| "loss": 0.419, |
| "step": 8830 |
| }, |
| { |
| "epoch": 0.0012392140018846146, |
| "grad_norm": 0.408203125, |
| "learning_rate": 0.0002999991058883173, |
| "loss": 0.423, |
| "step": 8840 |
| }, |
| { |
| "epoch": 0.0012406158276786017, |
| "grad_norm": 0.33203125, |
| "learning_rate": 0.0002999991036056764, |
| "loss": 0.4298, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.0012420176534725887, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.00029999910132012553, |
| "loss": 0.4679, |
| "step": 8860 |
| }, |
| { |
| "epoch": 0.001243419479266576, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.00029999909903166455, |
| "loss": 0.4492, |
| "step": 8870 |
| }, |
| { |
| "epoch": 0.001244821305060563, |
| "grad_norm": 0.490234375, |
| "learning_rate": 0.00029999909674029363, |
| "loss": 0.5003, |
| "step": 8880 |
| }, |
| { |
| "epoch": 0.0012462231308545501, |
| "grad_norm": 0.29296875, |
| "learning_rate": 0.0002999990944460126, |
| "loss": 0.3791, |
| "step": 8890 |
| }, |
| { |
| "epoch": 0.0012476249566485374, |
| "grad_norm": 0.34765625, |
| "learning_rate": 0.0002999990921488215, |
| "loss": 0.4881, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.0012490267824425245, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.0002999990898487205, |
| "loss": 0.4312, |
| "step": 8910 |
| }, |
| { |
| "epoch": 0.0012504286082365115, |
| "grad_norm": 0.318359375, |
| "learning_rate": 0.00029999908754570936, |
| "loss": 0.4883, |
| "step": 8920 |
| }, |
| { |
| "epoch": 0.0012518304340304986, |
| "grad_norm": 0.42578125, |
| "learning_rate": 0.0002999990852397882, |
| "loss": 0.3525, |
| "step": 8930 |
| }, |
| { |
| "epoch": 0.0012532322598244859, |
| "grad_norm": 0.53125, |
| "learning_rate": 0.000299999082930957, |
| "loss": 0.5207, |
| "step": 8940 |
| }, |
| { |
| "epoch": 0.001254634085618473, |
| "grad_norm": 0.76171875, |
| "learning_rate": 0.0002999990806192158, |
| "loss": 0.3815, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.00125603591141246, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.00029999907830456456, |
| "loss": 0.4554, |
| "step": 8960 |
| }, |
| { |
| "epoch": 0.001257437737206447, |
| "grad_norm": 0.384765625, |
| "learning_rate": 0.00029999907598700327, |
| "loss": 0.4522, |
| "step": 8970 |
| }, |
| { |
| "epoch": 0.0012588395630004343, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999990736665319, |
| "loss": 0.4533, |
| "step": 8980 |
| }, |
| { |
| "epoch": 0.0012602413887944214, |
| "grad_norm": 0.462890625, |
| "learning_rate": 0.00029999907134315053, |
| "loss": 0.5037, |
| "step": 8990 |
| }, |
| { |
| "epoch": 0.0012616432145884085, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.00029999906901685913, |
| "loss": 0.4848, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.0012616432145884085, |
| "eval_loss": 0.17965355515480042, |
| "eval_runtime": 113.876, |
| "eval_samples_per_second": 200.595, |
| "eval_steps_per_second": 1.572, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.0012630450403823957, |
| "grad_norm": 0.47265625, |
| "learning_rate": 0.00029999906668765774, |
| "loss": 0.4496, |
| "step": 9010 |
| }, |
| { |
| "epoch": 0.0012644468661763828, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.00029999906435554624, |
| "loss": 0.4478, |
| "step": 9020 |
| }, |
| { |
| "epoch": 0.0012658486919703699, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.00029999906202052474, |
| "loss": 0.4299, |
| "step": 9030 |
| }, |
| { |
| "epoch": 0.001267250517764357, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.00029999905968259324, |
| "loss": 0.5, |
| "step": 9040 |
| }, |
| { |
| "epoch": 0.0012686523435583442, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.00029999905734175164, |
| "loss": 0.4341, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.0012700541693523313, |
| "grad_norm": 0.87890625, |
| "learning_rate": 0.00029999905499800004, |
| "loss": 0.5121, |
| "step": 9060 |
| }, |
| { |
| "epoch": 0.0012714559951463183, |
| "grad_norm": 0.75, |
| "learning_rate": 0.00029999905265133843, |
| "loss": 0.423, |
| "step": 9070 |
| }, |
| { |
| "epoch": 0.0012728578209403054, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.0002999990503017668, |
| "loss": 0.5157, |
| "step": 9080 |
| }, |
| { |
| "epoch": 0.0012742596467342927, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.000299999047949285, |
| "loss": 0.4593, |
| "step": 9090 |
| }, |
| { |
| "epoch": 0.0012756614725282797, |
| "grad_norm": 0.8046875, |
| "learning_rate": 0.0002999990455938933, |
| "loss": 0.4342, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.0012770632983222668, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.00029999904323559155, |
| "loss": 0.4865, |
| "step": 9110 |
| }, |
| { |
| "epoch": 0.001278465124116254, |
| "grad_norm": 0.255859375, |
| "learning_rate": 0.00029999904087437974, |
| "loss": 0.448, |
| "step": 9120 |
| }, |
| { |
| "epoch": 0.0012798669499102411, |
| "grad_norm": 1.03125, |
| "learning_rate": 0.0002999990385102579, |
| "loss": 0.4956, |
| "step": 9130 |
| }, |
| { |
| "epoch": 0.0012812687757042282, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.00029999903614322596, |
| "loss": 0.4379, |
| "step": 9140 |
| }, |
| { |
| "epoch": 0.0012826706014982152, |
| "grad_norm": 0.392578125, |
| "learning_rate": 0.0002999990337732841, |
| "loss": 0.5302, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.0012840724272922025, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.0002999990314004322, |
| "loss": 0.5031, |
| "step": 9160 |
| }, |
| { |
| "epoch": 0.0012854742530861896, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.00029999902902467016, |
| "loss": 0.3132, |
| "step": 9170 |
| }, |
| { |
| "epoch": 0.0012868760788801766, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.00029999902664599814, |
| "loss": 0.3933, |
| "step": 9180 |
| }, |
| { |
| "epoch": 0.001288277904674164, |
| "grad_norm": 0.296875, |
| "learning_rate": 0.0002999990242644161, |
| "loss": 0.4043, |
| "step": 9190 |
| }, |
| { |
| "epoch": 0.001289679730468151, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.00029999902187992405, |
| "loss": 0.5478, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.001291081556262138, |
| "grad_norm": 0.26171875, |
| "learning_rate": 0.0002999990194925219, |
| "loss": 0.495, |
| "step": 9210 |
| }, |
| { |
| "epoch": 0.001292483382056125, |
| "grad_norm": 0.345703125, |
| "learning_rate": 0.00029999901710220974, |
| "loss": 0.5168, |
| "step": 9220 |
| }, |
| { |
| "epoch": 0.0012938852078501124, |
| "grad_norm": 0.310546875, |
| "learning_rate": 0.00029999901470898757, |
| "loss": 0.3826, |
| "step": 9230 |
| }, |
| { |
| "epoch": 0.0012952870336440994, |
| "grad_norm": 0.470703125, |
| "learning_rate": 0.0002999990123128554, |
| "loss": 0.4363, |
| "step": 9240 |
| }, |
| { |
| "epoch": 0.0012966888594380865, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.0002999990099138131, |
| "loss": 0.4175, |
| "step": 9250 |
| }, |
| { |
| "epoch": 0.0012980906852320736, |
| "grad_norm": 0.34765625, |
| "learning_rate": 0.00029999900751186083, |
| "loss": 0.3774, |
| "step": 9260 |
| }, |
| { |
| "epoch": 0.0012994925110260608, |
| "grad_norm": 0.9140625, |
| "learning_rate": 0.00029999900510699855, |
| "loss": 0.4917, |
| "step": 9270 |
| }, |
| { |
| "epoch": 0.001300894336820048, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.00029999900269922616, |
| "loss": 0.5462, |
| "step": 9280 |
| }, |
| { |
| "epoch": 0.001302296162614035, |
| "grad_norm": 0.35546875, |
| "learning_rate": 0.0002999990002885438, |
| "loss": 0.344, |
| "step": 9290 |
| }, |
| { |
| "epoch": 0.0013036979884080222, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.00029999899787495134, |
| "loss": 0.499, |
| "step": 9300 |
| }, |
| { |
| "epoch": 0.0013050998142020093, |
| "grad_norm": 1.0703125, |
| "learning_rate": 0.0002999989954584489, |
| "loss": 0.4163, |
| "step": 9310 |
| }, |
| { |
| "epoch": 0.0013065016399959964, |
| "grad_norm": 0.296875, |
| "learning_rate": 0.0002999989930390364, |
| "loss": 0.4647, |
| "step": 9320 |
| }, |
| { |
| "epoch": 0.0013079034657899834, |
| "grad_norm": 0.44921875, |
| "learning_rate": 0.00029999899061671387, |
| "loss": 0.4569, |
| "step": 9330 |
| }, |
| { |
| "epoch": 0.0013093052915839707, |
| "grad_norm": 0.150390625, |
| "learning_rate": 0.00029999898819148133, |
| "loss": 0.4494, |
| "step": 9340 |
| }, |
| { |
| "epoch": 0.0013107071173779578, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.0002999989857633388, |
| "loss": 0.4393, |
| "step": 9350 |
| }, |
| { |
| "epoch": 0.0013121089431719448, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.00029999898333228615, |
| "loss": 0.5231, |
| "step": 9360 |
| }, |
| { |
| "epoch": 0.0013135107689659319, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.0002999989808983235, |
| "loss": 0.3771, |
| "step": 9370 |
| }, |
| { |
| "epoch": 0.0013149125947599192, |
| "grad_norm": 1.1171875, |
| "learning_rate": 0.0002999989784614508, |
| "loss": 0.5389, |
| "step": 9380 |
| }, |
| { |
| "epoch": 0.0013163144205539062, |
| "grad_norm": 0.6484375, |
| "learning_rate": 0.0002999989760216681, |
| "loss": 0.4195, |
| "step": 9390 |
| }, |
| { |
| "epoch": 0.0013177162463478933, |
| "grad_norm": 0.283203125, |
| "learning_rate": 0.00029999897357897536, |
| "loss": 0.4314, |
| "step": 9400 |
| }, |
| { |
| "epoch": 0.0013191180721418806, |
| "grad_norm": 1.15625, |
| "learning_rate": 0.00029999897113337255, |
| "loss": 0.4837, |
| "step": 9410 |
| }, |
| { |
| "epoch": 0.0013205198979358676, |
| "grad_norm": 0.4609375, |
| "learning_rate": 0.00029999896868485975, |
| "loss": 0.4618, |
| "step": 9420 |
| }, |
| { |
| "epoch": 0.0013219217237298547, |
| "grad_norm": 1.171875, |
| "learning_rate": 0.0002999989662334369, |
| "loss": 0.4941, |
| "step": 9430 |
| }, |
| { |
| "epoch": 0.0013233235495238417, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.000299998963779104, |
| "loss": 0.4681, |
| "step": 9440 |
| }, |
| { |
| "epoch": 0.001324725375317829, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.0002999989613218611, |
| "loss": 0.4324, |
| "step": 9450 |
| }, |
| { |
| "epoch": 0.001326127201111816, |
| "grad_norm": 0.85546875, |
| "learning_rate": 0.0002999989588617082, |
| "loss": 0.4494, |
| "step": 9460 |
| }, |
| { |
| "epoch": 0.0013275290269058031, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.00029999895639864517, |
| "loss": 0.4027, |
| "step": 9470 |
| }, |
| { |
| "epoch": 0.0013289308526997904, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.00029999895393267216, |
| "loss": 0.4776, |
| "step": 9480 |
| }, |
| { |
| "epoch": 0.0013303326784937775, |
| "grad_norm": 0.51171875, |
| "learning_rate": 0.0002999989514637891, |
| "loss": 0.4024, |
| "step": 9490 |
| }, |
| { |
| "epoch": 0.0013317345042877645, |
| "grad_norm": 0.91015625, |
| "learning_rate": 0.0002999989489919961, |
| "loss": 0.405, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.0013317345042877645, |
| "eval_loss": 0.1779126524925232, |
| "eval_runtime": 112.2872, |
| "eval_samples_per_second": 203.434, |
| "eval_steps_per_second": 1.594, |
| "step": 9500 |
| }, |
| { |
| "epoch": 0.0013331363300817516, |
| "grad_norm": 0.380859375, |
| "learning_rate": 0.000299998946517293, |
| "loss": 0.4613, |
| "step": 9510 |
| }, |
| { |
| "epoch": 0.0013345381558757389, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.0002999989440396798, |
| "loss": 0.4131, |
| "step": 9520 |
| }, |
| { |
| "epoch": 0.001335939981669726, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.00029999894155915665, |
| "loss": 0.4716, |
| "step": 9530 |
| }, |
| { |
| "epoch": 0.001337341807463713, |
| "grad_norm": 0.234375, |
| "learning_rate": 0.00029999893907572343, |
| "loss": 0.3789, |
| "step": 9540 |
| }, |
| { |
| "epoch": 0.0013387436332577, |
| "grad_norm": 0.5546875, |
| "learning_rate": 0.0002999989365893802, |
| "loss": 0.4503, |
| "step": 9550 |
| }, |
| { |
| "epoch": 0.0013401454590516873, |
| "grad_norm": 0.44921875, |
| "learning_rate": 0.00029999893410012694, |
| "loss": 0.3946, |
| "step": 9560 |
| }, |
| { |
| "epoch": 0.0013415472848456744, |
| "grad_norm": 0.3359375, |
| "learning_rate": 0.0002999989316079636, |
| "loss": 0.3684, |
| "step": 9570 |
| }, |
| { |
| "epoch": 0.0013429491106396615, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.0002999989291128903, |
| "loss": 0.4323, |
| "step": 9580 |
| }, |
| { |
| "epoch": 0.0013443509364336487, |
| "grad_norm": 0.3359375, |
| "learning_rate": 0.0002999989266149069, |
| "loss": 0.403, |
| "step": 9590 |
| }, |
| { |
| "epoch": 0.0013457527622276358, |
| "grad_norm": 1.40625, |
| "learning_rate": 0.00029999892411401355, |
| "loss": 0.4677, |
| "step": 9600 |
| }, |
| { |
| "epoch": 0.0013471545880216229, |
| "grad_norm": 0.5390625, |
| "learning_rate": 0.00029999892161021013, |
| "loss": 0.4505, |
| "step": 9610 |
| }, |
| { |
| "epoch": 0.00134855641381561, |
| "grad_norm": 0.625, |
| "learning_rate": 0.00029999891910349665, |
| "loss": 0.3674, |
| "step": 9620 |
| }, |
| { |
| "epoch": 0.0013499582396095972, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.00029999891659387317, |
| "loss": 0.4835, |
| "step": 9630 |
| }, |
| { |
| "epoch": 0.0013513600654035843, |
| "grad_norm": 0.431640625, |
| "learning_rate": 0.00029999891408133964, |
| "loss": 0.4168, |
| "step": 9640 |
| }, |
| { |
| "epoch": 0.0013527618911975713, |
| "grad_norm": 0.45703125, |
| "learning_rate": 0.00029999891156589606, |
| "loss": 0.4533, |
| "step": 9650 |
| }, |
| { |
| "epoch": 0.0013541637169915584, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.0002999989090475425, |
| "loss": 0.4661, |
| "step": 9660 |
| }, |
| { |
| "epoch": 0.0013555655427855457, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.00029999890652627884, |
| "loss": 0.5244, |
| "step": 9670 |
| }, |
| { |
| "epoch": 0.0013569673685795327, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.0002999989040021052, |
| "loss": 0.4458, |
| "step": 9680 |
| }, |
| { |
| "epoch": 0.0013583691943735198, |
| "grad_norm": 0.65234375, |
| "learning_rate": 0.0002999989014750215, |
| "loss": 0.3922, |
| "step": 9690 |
| }, |
| { |
| "epoch": 0.001359771020167507, |
| "grad_norm": 0.404296875, |
| "learning_rate": 0.0002999988989450278, |
| "loss": 0.4443, |
| "step": 9700 |
| }, |
| { |
| "epoch": 0.0013611728459614941, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.00029999889641212405, |
| "loss": 0.3954, |
| "step": 9710 |
| }, |
| { |
| "epoch": 0.0013625746717554812, |
| "grad_norm": 0.32421875, |
| "learning_rate": 0.00029999889387631026, |
| "loss": 0.4527, |
| "step": 9720 |
| }, |
| { |
| "epoch": 0.0013639764975494682, |
| "grad_norm": 0.2294921875, |
| "learning_rate": 0.0002999988913375865, |
| "loss": 0.3806, |
| "step": 9730 |
| }, |
| { |
| "epoch": 0.0013653783233434555, |
| "grad_norm": 0.5390625, |
| "learning_rate": 0.0002999988887959527, |
| "loss": 0.46, |
| "step": 9740 |
| }, |
| { |
| "epoch": 0.0013667801491374426, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.00029999888625140874, |
| "loss": 0.4645, |
| "step": 9750 |
| }, |
| { |
| "epoch": 0.0013681819749314296, |
| "grad_norm": 0.5, |
| "learning_rate": 0.0002999988837039549, |
| "loss": 0.3977, |
| "step": 9760 |
| }, |
| { |
| "epoch": 0.001369583800725417, |
| "grad_norm": 0.70703125, |
| "learning_rate": 0.0002999988811535909, |
| "loss": 0.4475, |
| "step": 9770 |
| }, |
| { |
| "epoch": 0.001370985626519404, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.00029999887860031695, |
| "loss": 0.4696, |
| "step": 9780 |
| }, |
| { |
| "epoch": 0.001372387452313391, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.000299998876044133, |
| "loss": 0.4126, |
| "step": 9790 |
| }, |
| { |
| "epoch": 0.001373789278107378, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.0002999988734850389, |
| "loss": 0.4177, |
| "step": 9800 |
| }, |
| { |
| "epoch": 0.0013751911039013654, |
| "grad_norm": 0.70703125, |
| "learning_rate": 0.00029999887092303486, |
| "loss": 0.4654, |
| "step": 9810 |
| }, |
| { |
| "epoch": 0.0013765929296953524, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999988683581208, |
| "loss": 0.4431, |
| "step": 9820 |
| }, |
| { |
| "epoch": 0.0013779947554893395, |
| "grad_norm": 0.80859375, |
| "learning_rate": 0.00029999886579029666, |
| "loss": 0.4776, |
| "step": 9830 |
| }, |
| { |
| "epoch": 0.0013793965812833266, |
| "grad_norm": 0.6015625, |
| "learning_rate": 0.0002999988632195625, |
| "loss": 0.507, |
| "step": 9840 |
| }, |
| { |
| "epoch": 0.0013807984070773138, |
| "grad_norm": 0.359375, |
| "learning_rate": 0.00029999886064591835, |
| "loss": 0.496, |
| "step": 9850 |
| }, |
| { |
| "epoch": 0.001382200232871301, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.00029999885806936415, |
| "loss": 0.5075, |
| "step": 9860 |
| }, |
| { |
| "epoch": 0.001383602058665288, |
| "grad_norm": 0.328125, |
| "learning_rate": 0.0002999988554898999, |
| "loss": 0.3726, |
| "step": 9870 |
| }, |
| { |
| "epoch": 0.0013850038844592752, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999885290752564, |
| "loss": 0.4425, |
| "step": 9880 |
| }, |
| { |
| "epoch": 0.0013864057102532623, |
| "grad_norm": 0.435546875, |
| "learning_rate": 0.0002999988503222413, |
| "loss": 0.4121, |
| "step": 9890 |
| }, |
| { |
| "epoch": 0.0013878075360472494, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.00029999884773404697, |
| "loss": 0.3684, |
| "step": 9900 |
| }, |
| { |
| "epoch": 0.0013892093618412364, |
| "grad_norm": 0.49609375, |
| "learning_rate": 0.0002999988451429426, |
| "loss": 0.3725, |
| "step": 9910 |
| }, |
| { |
| "epoch": 0.0013906111876352237, |
| "grad_norm": 0.419921875, |
| "learning_rate": 0.00029999884254892825, |
| "loss": 0.469, |
| "step": 9920 |
| }, |
| { |
| "epoch": 0.0013920130134292108, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.00029999883995200383, |
| "loss": 0.4056, |
| "step": 9930 |
| }, |
| { |
| "epoch": 0.0013934148392231978, |
| "grad_norm": 0.3828125, |
| "learning_rate": 0.00029999883735216937, |
| "loss": 0.415, |
| "step": 9940 |
| }, |
| { |
| "epoch": 0.0013948166650171849, |
| "grad_norm": 0.5078125, |
| "learning_rate": 0.00029999883474942485, |
| "loss": 0.4442, |
| "step": 9950 |
| }, |
| { |
| "epoch": 0.0013962184908111722, |
| "grad_norm": 0.1845703125, |
| "learning_rate": 0.00029999883214377034, |
| "loss": 0.4451, |
| "step": 9960 |
| }, |
| { |
| "epoch": 0.0013976203166051592, |
| "grad_norm": 0.25390625, |
| "learning_rate": 0.0002999988295352058, |
| "loss": 0.4303, |
| "step": 9970 |
| }, |
| { |
| "epoch": 0.0013990221423991463, |
| "grad_norm": 0.99609375, |
| "learning_rate": 0.00029999882692373125, |
| "loss": 0.4909, |
| "step": 9980 |
| }, |
| { |
| "epoch": 0.0014004239681931336, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.00029999882430934663, |
| "loss": 0.3695, |
| "step": 9990 |
| }, |
| { |
| "epoch": 0.0014018257939871206, |
| "grad_norm": 0.9921875, |
| "learning_rate": 0.000299998821692052, |
| "loss": 0.489, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.0014018257939871206, |
| "eval_loss": 0.17730337381362915, |
| "eval_runtime": 112.2399, |
| "eval_samples_per_second": 203.519, |
| "eval_steps_per_second": 1.595, |
| "step": 10000 |
| }, |
| { |
| "epoch": 0.0014032276197811077, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.0002999988190718473, |
| "loss": 0.4894, |
| "step": 10010 |
| }, |
| { |
| "epoch": 0.0014046294455750947, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999988164487327, |
| "loss": 0.427, |
| "step": 10020 |
| }, |
| { |
| "epoch": 0.001406031271369082, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999988138227079, |
| "loss": 0.404, |
| "step": 10030 |
| }, |
| { |
| "epoch": 0.001407433097163069, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.0002999988111937732, |
| "loss": 0.3942, |
| "step": 10040 |
| }, |
| { |
| "epoch": 0.0014088349229570561, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.00029999880856192835, |
| "loss": 0.3519, |
| "step": 10050 |
| }, |
| { |
| "epoch": 0.0014102367487510434, |
| "grad_norm": 0.259765625, |
| "learning_rate": 0.00029999880592717357, |
| "loss": 0.4191, |
| "step": 10060 |
| }, |
| { |
| "epoch": 0.0014116385745450305, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.0002999988032895087, |
| "loss": 0.4562, |
| "step": 10070 |
| }, |
| { |
| "epoch": 0.0014130404003390175, |
| "grad_norm": 0.9765625, |
| "learning_rate": 0.00029999880064893387, |
| "loss": 0.4004, |
| "step": 10080 |
| }, |
| { |
| "epoch": 0.0014144422261330046, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.00029999879800544894, |
| "loss": 0.4163, |
| "step": 10090 |
| }, |
| { |
| "epoch": 0.0014158440519269919, |
| "grad_norm": 0.30859375, |
| "learning_rate": 0.000299998795359054, |
| "loss": 0.292, |
| "step": 10100 |
| }, |
| { |
| "epoch": 0.001417245877720979, |
| "grad_norm": 0.263671875, |
| "learning_rate": 0.000299998792709749, |
| "loss": 0.3318, |
| "step": 10110 |
| }, |
| { |
| "epoch": 0.001418647703514966, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.00029999879005753404, |
| "loss": 0.4539, |
| "step": 10120 |
| }, |
| { |
| "epoch": 0.001420049529308953, |
| "grad_norm": 0.361328125, |
| "learning_rate": 0.00029999878740240906, |
| "loss": 0.3881, |
| "step": 10130 |
| }, |
| { |
| "epoch": 0.0014214513551029403, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.00029999878474437397, |
| "loss": 0.4689, |
| "step": 10140 |
| }, |
| { |
| "epoch": 0.0014228531808969274, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00029999878208342894, |
| "loss": 0.4459, |
| "step": 10150 |
| }, |
| { |
| "epoch": 0.0014242550066909145, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.0002999987794195738, |
| "loss": 0.3846, |
| "step": 10160 |
| }, |
| { |
| "epoch": 0.0014256568324849017, |
| "grad_norm": 0.8203125, |
| "learning_rate": 0.0002999987767528086, |
| "loss": 0.4947, |
| "step": 10170 |
| }, |
| { |
| "epoch": 0.0014270586582788888, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.00029999877408313347, |
| "loss": 0.5135, |
| "step": 10180 |
| }, |
| { |
| "epoch": 0.0014284604840728759, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.0002999987714105483, |
| "loss": 0.4745, |
| "step": 10190 |
| }, |
| { |
| "epoch": 0.001429862309866863, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.00029999876873505304, |
| "loss": 0.4396, |
| "step": 10200 |
| }, |
| { |
| "epoch": 0.0014312641356608502, |
| "grad_norm": 1.0, |
| "learning_rate": 0.0002999987660566478, |
| "loss": 0.4613, |
| "step": 10210 |
| }, |
| { |
| "epoch": 0.0014326659614548373, |
| "grad_norm": 1.015625, |
| "learning_rate": 0.0002999987633753325, |
| "loss": 0.4413, |
| "step": 10220 |
| }, |
| { |
| "epoch": 0.0014340677872488243, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999987606911072, |
| "loss": 0.3807, |
| "step": 10230 |
| }, |
| { |
| "epoch": 0.0014354696130428114, |
| "grad_norm": 0.275390625, |
| "learning_rate": 0.00029999875800397187, |
| "loss": 0.3948, |
| "step": 10240 |
| }, |
| { |
| "epoch": 0.0014368714388367987, |
| "grad_norm": 0.2197265625, |
| "learning_rate": 0.00029999875531392647, |
| "loss": 0.354, |
| "step": 10250 |
| }, |
| { |
| "epoch": 0.0014382732646307857, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.00029999875262097113, |
| "loss": 0.4706, |
| "step": 10260 |
| }, |
| { |
| "epoch": 0.0014396750904247728, |
| "grad_norm": 0.39453125, |
| "learning_rate": 0.0002999987499251057, |
| "loss": 0.3163, |
| "step": 10270 |
| }, |
| { |
| "epoch": 0.00144107691621876, |
| "grad_norm": 0.65234375, |
| "learning_rate": 0.00029999874722633023, |
| "loss": 0.3933, |
| "step": 10280 |
| }, |
| { |
| "epoch": 0.0014424787420127471, |
| "grad_norm": 0.431640625, |
| "learning_rate": 0.0002999987445246448, |
| "loss": 0.3804, |
| "step": 10290 |
| }, |
| { |
| "epoch": 0.0014438805678067342, |
| "grad_norm": 0.70703125, |
| "learning_rate": 0.00029999874182004923, |
| "loss": 0.4831, |
| "step": 10300 |
| }, |
| { |
| "epoch": 0.0014452823936007212, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.0002999987391125437, |
| "loss": 0.3895, |
| "step": 10310 |
| }, |
| { |
| "epoch": 0.0014466842193947085, |
| "grad_norm": 0.298828125, |
| "learning_rate": 0.00029999873640212813, |
| "loss": 0.4482, |
| "step": 10320 |
| }, |
| { |
| "epoch": 0.0014480860451886956, |
| "grad_norm": 0.26953125, |
| "learning_rate": 0.0002999987336888025, |
| "loss": 0.4603, |
| "step": 10330 |
| }, |
| { |
| "epoch": 0.0014494878709826826, |
| "grad_norm": 0.44140625, |
| "learning_rate": 0.0002999987309725669, |
| "loss": 0.4359, |
| "step": 10340 |
| }, |
| { |
| "epoch": 0.00145088969677667, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.00029999872825342127, |
| "loss": 0.4599, |
| "step": 10350 |
| }, |
| { |
| "epoch": 0.001452291522570657, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00029999872553136556, |
| "loss": 0.3849, |
| "step": 10360 |
| }, |
| { |
| "epoch": 0.001453693348364644, |
| "grad_norm": 0.2578125, |
| "learning_rate": 0.00029999872280639985, |
| "loss": 0.3681, |
| "step": 10370 |
| }, |
| { |
| "epoch": 0.001455095174158631, |
| "grad_norm": 0.431640625, |
| "learning_rate": 0.0002999987200785241, |
| "loss": 0.3418, |
| "step": 10380 |
| }, |
| { |
| "epoch": 0.0014564969999526184, |
| "grad_norm": 0.5078125, |
| "learning_rate": 0.0002999987173477384, |
| "loss": 0.4339, |
| "step": 10390 |
| }, |
| { |
| "epoch": 0.0014578988257466054, |
| "grad_norm": 0.435546875, |
| "learning_rate": 0.0002999987146140426, |
| "loss": 0.3566, |
| "step": 10400 |
| }, |
| { |
| "epoch": 0.0014593006515405925, |
| "grad_norm": 1.234375, |
| "learning_rate": 0.00029999871187743677, |
| "loss": 0.4247, |
| "step": 10410 |
| }, |
| { |
| "epoch": 0.0014607024773345796, |
| "grad_norm": 0.96875, |
| "learning_rate": 0.00029999870913792085, |
| "loss": 0.5284, |
| "step": 10420 |
| }, |
| { |
| "epoch": 0.0014621043031285668, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.00029999870639549505, |
| "loss": 0.4373, |
| "step": 10430 |
| }, |
| { |
| "epoch": 0.0014635061289225539, |
| "grad_norm": 0.48828125, |
| "learning_rate": 0.0002999987036501591, |
| "loss": 0.3596, |
| "step": 10440 |
| }, |
| { |
| "epoch": 0.001464907954716541, |
| "grad_norm": 0.236328125, |
| "learning_rate": 0.0002999987009019132, |
| "loss": 0.4211, |
| "step": 10450 |
| }, |
| { |
| "epoch": 0.0014663097805105282, |
| "grad_norm": 0.546875, |
| "learning_rate": 0.00029999869815075725, |
| "loss": 0.487, |
| "step": 10460 |
| }, |
| { |
| "epoch": 0.0014677116063045153, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.00029999869539669124, |
| "loss": 0.4045, |
| "step": 10470 |
| }, |
| { |
| "epoch": 0.0014691134320985024, |
| "grad_norm": 0.380859375, |
| "learning_rate": 0.0002999986926397152, |
| "loss": 0.4048, |
| "step": 10480 |
| }, |
| { |
| "epoch": 0.0014705152578924894, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.00029999868987982915, |
| "loss": 0.4103, |
| "step": 10490 |
| }, |
| { |
| "epoch": 0.0014719170836864767, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.0002999986871170331, |
| "loss": 0.3853, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.0014719170836864767, |
| "eval_loss": 0.1777743548154831, |
| "eval_runtime": 113.9971, |
| "eval_samples_per_second": 200.382, |
| "eval_steps_per_second": 1.57, |
| "step": 10500 |
| }, |
| { |
| "epoch": 0.0014733189094804637, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.000299998684351327, |
| "loss": 0.4171, |
| "step": 10510 |
| }, |
| { |
| "epoch": 0.0014747207352744508, |
| "grad_norm": 0.361328125, |
| "learning_rate": 0.0002999986815827109, |
| "loss": 0.4019, |
| "step": 10520 |
| }, |
| { |
| "epoch": 0.0014761225610684379, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999986788111848, |
| "loss": 0.5203, |
| "step": 10530 |
| }, |
| { |
| "epoch": 0.0014775243868624251, |
| "grad_norm": 0.640625, |
| "learning_rate": 0.00029999867603674855, |
| "loss": 0.4741, |
| "step": 10540 |
| }, |
| { |
| "epoch": 0.0014789262126564122, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.0002999986732594024, |
| "loss": 0.4593, |
| "step": 10550 |
| }, |
| { |
| "epoch": 0.0014803280384503993, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.0002999986704791461, |
| "loss": 0.45, |
| "step": 10560 |
| }, |
| { |
| "epoch": 0.0014817298642443865, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.0002999986676959799, |
| "loss": 0.4173, |
| "step": 10570 |
| }, |
| { |
| "epoch": 0.0014831316900383736, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.00029999866490990355, |
| "loss": 0.4825, |
| "step": 10580 |
| }, |
| { |
| "epoch": 0.0014845335158323607, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.0002999986621209173, |
| "loss": 0.4325, |
| "step": 10590 |
| }, |
| { |
| "epoch": 0.0014859353416263477, |
| "grad_norm": 0.65234375, |
| "learning_rate": 0.00029999865932902096, |
| "loss": 0.3705, |
| "step": 10600 |
| }, |
| { |
| "epoch": 0.001487337167420335, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.0002999986565342146, |
| "loss": 0.4873, |
| "step": 10610 |
| }, |
| { |
| "epoch": 0.001488738993214322, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.00029999865373649815, |
| "loss": 0.4666, |
| "step": 10620 |
| }, |
| { |
| "epoch": 0.0014901408190083091, |
| "grad_norm": 0.2177734375, |
| "learning_rate": 0.00029999865093587177, |
| "loss": 0.3025, |
| "step": 10630 |
| }, |
| { |
| "epoch": 0.0014915426448022964, |
| "grad_norm": 0.49609375, |
| "learning_rate": 0.0002999986481323353, |
| "loss": 0.4764, |
| "step": 10640 |
| }, |
| { |
| "epoch": 0.0014929444705962835, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.0002999986453258888, |
| "loss": 0.4676, |
| "step": 10650 |
| }, |
| { |
| "epoch": 0.0014943462963902705, |
| "grad_norm": 0.26171875, |
| "learning_rate": 0.0002999986425165323, |
| "loss": 0.4282, |
| "step": 10660 |
| }, |
| { |
| "epoch": 0.0014957481221842576, |
| "grad_norm": 0.3046875, |
| "learning_rate": 0.0002999986397042658, |
| "loss": 0.3724, |
| "step": 10670 |
| }, |
| { |
| "epoch": 0.0014971499479782449, |
| "grad_norm": 0.423828125, |
| "learning_rate": 0.0002999986368890892, |
| "loss": 0.3169, |
| "step": 10680 |
| }, |
| { |
| "epoch": 0.001498551773772232, |
| "grad_norm": 0.33984375, |
| "learning_rate": 0.0002999986340710027, |
| "loss": 0.3254, |
| "step": 10690 |
| }, |
| { |
| "epoch": 0.001499953599566219, |
| "grad_norm": 0.2197265625, |
| "learning_rate": 0.00029999863125000604, |
| "loss": 0.4022, |
| "step": 10700 |
| }, |
| { |
| "epoch": 0.001501355425360206, |
| "grad_norm": 0.81640625, |
| "learning_rate": 0.00029999862842609946, |
| "loss": 0.5181, |
| "step": 10710 |
| }, |
| { |
| "epoch": 0.0015027572511541933, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.00029999862559928277, |
| "loss": 0.5158, |
| "step": 10720 |
| }, |
| { |
| "epoch": 0.0015041590769481804, |
| "grad_norm": 0.81640625, |
| "learning_rate": 0.0002999986227695561, |
| "loss": 0.3475, |
| "step": 10730 |
| }, |
| { |
| "epoch": 0.0015055609027421674, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.0002999986199369194, |
| "loss": 0.4349, |
| "step": 10740 |
| }, |
| { |
| "epoch": 0.0015069627285361547, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.0002999986171013726, |
| "loss": 0.4544, |
| "step": 10750 |
| }, |
| { |
| "epoch": 0.0015083645543301418, |
| "grad_norm": 1.2109375, |
| "learning_rate": 0.0002999986142629159, |
| "loss": 0.5036, |
| "step": 10760 |
| }, |
| { |
| "epoch": 0.0015097663801241288, |
| "grad_norm": 0.5, |
| "learning_rate": 0.0002999986114215491, |
| "loss": 0.4059, |
| "step": 10770 |
| }, |
| { |
| "epoch": 0.001511168205918116, |
| "grad_norm": 0.341796875, |
| "learning_rate": 0.0002999986085772723, |
| "loss": 0.3344, |
| "step": 10780 |
| }, |
| { |
| "epoch": 0.0015125700317121032, |
| "grad_norm": 0.75, |
| "learning_rate": 0.00029999860573008544, |
| "loss": 0.486, |
| "step": 10790 |
| }, |
| { |
| "epoch": 0.0015139718575060902, |
| "grad_norm": 0.25, |
| "learning_rate": 0.00029999860287998855, |
| "loss": 0.347, |
| "step": 10800 |
| }, |
| { |
| "epoch": 0.0015153736833000773, |
| "grad_norm": 0.4453125, |
| "learning_rate": 0.0002999986000269817, |
| "loss": 0.3681, |
| "step": 10810 |
| }, |
| { |
| "epoch": 0.0015167755090940644, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.00029999859717106476, |
| "loss": 0.4981, |
| "step": 10820 |
| }, |
| { |
| "epoch": 0.0015181773348880516, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.0002999985943122379, |
| "loss": 0.4436, |
| "step": 10830 |
| }, |
| { |
| "epoch": 0.0015195791606820387, |
| "grad_norm": 1.0859375, |
| "learning_rate": 0.0002999985914505009, |
| "loss": 0.4291, |
| "step": 10840 |
| }, |
| { |
| "epoch": 0.0015209809864760258, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.0002999985885858539, |
| "loss": 0.4227, |
| "step": 10850 |
| }, |
| { |
| "epoch": 0.001522382812270013, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.00029999858571829683, |
| "loss": 0.4109, |
| "step": 10860 |
| }, |
| { |
| "epoch": 0.001523784638064, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.00029999858284782984, |
| "loss": 0.487, |
| "step": 10870 |
| }, |
| { |
| "epoch": 0.0015251864638579872, |
| "grad_norm": 0.6875, |
| "learning_rate": 0.00029999857997445274, |
| "loss": 0.4839, |
| "step": 10880 |
| }, |
| { |
| "epoch": 0.0015265882896519742, |
| "grad_norm": 0.7109375, |
| "learning_rate": 0.00029999857709816564, |
| "loss": 0.4406, |
| "step": 10890 |
| }, |
| { |
| "epoch": 0.0015279901154459615, |
| "grad_norm": 0.52734375, |
| "learning_rate": 0.00029999857421896854, |
| "loss": 0.4348, |
| "step": 10900 |
| }, |
| { |
| "epoch": 0.0015293919412399486, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.0002999985713368614, |
| "loss": 0.4313, |
| "step": 10910 |
| }, |
| { |
| "epoch": 0.0015307937670339356, |
| "grad_norm": 0.92578125, |
| "learning_rate": 0.00029999856845184424, |
| "loss": 0.4236, |
| "step": 10920 |
| }, |
| { |
| "epoch": 0.001532195592827923, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.00029999856556391704, |
| "loss": 0.4851, |
| "step": 10930 |
| }, |
| { |
| "epoch": 0.00153359741862191, |
| "grad_norm": 0.95703125, |
| "learning_rate": 0.00029999856267307984, |
| "loss": 0.3592, |
| "step": 10940 |
| }, |
| { |
| "epoch": 0.001534999244415897, |
| "grad_norm": 0.462890625, |
| "learning_rate": 0.00029999855977933253, |
| "loss": 0.4506, |
| "step": 10950 |
| }, |
| { |
| "epoch": 0.001536401070209884, |
| "grad_norm": 0.4375, |
| "learning_rate": 0.0002999985568826753, |
| "loss": 0.467, |
| "step": 10960 |
| }, |
| { |
| "epoch": 0.0015378028960038714, |
| "grad_norm": 0.9765625, |
| "learning_rate": 0.00029999855398310803, |
| "loss": 0.4379, |
| "step": 10970 |
| }, |
| { |
| "epoch": 0.0015392047217978584, |
| "grad_norm": 0.326171875, |
| "learning_rate": 0.00029999855108063067, |
| "loss": 0.3838, |
| "step": 10980 |
| }, |
| { |
| "epoch": 0.0015406065475918455, |
| "grad_norm": 0.8203125, |
| "learning_rate": 0.0002999985481752433, |
| "loss": 0.4903, |
| "step": 10990 |
| }, |
| { |
| "epoch": 0.0015420083733858325, |
| "grad_norm": 0.8515625, |
| "learning_rate": 0.00029999854526694596, |
| "loss": 0.4298, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.0015420083733858325, |
| "eval_loss": 0.1755325049161911, |
| "eval_runtime": 112.2725, |
| "eval_samples_per_second": 203.46, |
| "eval_steps_per_second": 1.594, |
| "step": 11000 |
| }, |
| { |
| "epoch": 0.0015434101991798198, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.00029999854235573855, |
| "loss": 0.3867, |
| "step": 11010 |
| }, |
| { |
| "epoch": 0.0015448120249738069, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.00029999853944162114, |
| "loss": 0.4283, |
| "step": 11020 |
| }, |
| { |
| "epoch": 0.001546213850767794, |
| "grad_norm": 0.89453125, |
| "learning_rate": 0.00029999853652459374, |
| "loss": 0.4642, |
| "step": 11030 |
| }, |
| { |
| "epoch": 0.0015476156765617812, |
| "grad_norm": 0.83984375, |
| "learning_rate": 0.0002999985336046562, |
| "loss": 0.3961, |
| "step": 11040 |
| }, |
| { |
| "epoch": 0.0015490175023557683, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.00029999853068180877, |
| "loss": 0.4539, |
| "step": 11050 |
| }, |
| { |
| "epoch": 0.0015504193281497553, |
| "grad_norm": 0.984375, |
| "learning_rate": 0.0002999985277560512, |
| "loss": 0.45, |
| "step": 11060 |
| }, |
| { |
| "epoch": 0.0015518211539437424, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999985248273837, |
| "loss": 0.432, |
| "step": 11070 |
| }, |
| { |
| "epoch": 0.0015532229797377297, |
| "grad_norm": 1.0546875, |
| "learning_rate": 0.00029999852189580614, |
| "loss": 0.499, |
| "step": 11080 |
| }, |
| { |
| "epoch": 0.0015546248055317167, |
| "grad_norm": 0.4296875, |
| "learning_rate": 0.0002999985189613185, |
| "loss": 0.3849, |
| "step": 11090 |
| }, |
| { |
| "epoch": 0.0015560266313257038, |
| "grad_norm": 0.5390625, |
| "learning_rate": 0.00029999851602392096, |
| "loss": 0.3608, |
| "step": 11100 |
| }, |
| { |
| "epoch": 0.0015574284571196909, |
| "grad_norm": 0.3828125, |
| "learning_rate": 0.0002999985130836133, |
| "loss": 0.4467, |
| "step": 11110 |
| }, |
| { |
| "epoch": 0.0015588302829136781, |
| "grad_norm": 0.87109375, |
| "learning_rate": 0.00029999851014039563, |
| "loss": 0.4526, |
| "step": 11120 |
| }, |
| { |
| "epoch": 0.0015602321087076652, |
| "grad_norm": 0.97265625, |
| "learning_rate": 0.00029999850719426797, |
| "loss": 0.4376, |
| "step": 11130 |
| }, |
| { |
| "epoch": 0.0015616339345016523, |
| "grad_norm": 0.91796875, |
| "learning_rate": 0.00029999850424523026, |
| "loss": 0.3923, |
| "step": 11140 |
| }, |
| { |
| "epoch": 0.0015630357602956395, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.0002999985012932825, |
| "loss": 0.4162, |
| "step": 11150 |
| }, |
| { |
| "epoch": 0.0015644375860896266, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.0002999984983384248, |
| "loss": 0.4105, |
| "step": 11160 |
| }, |
| { |
| "epoch": 0.0015658394118836137, |
| "grad_norm": 0.87109375, |
| "learning_rate": 0.000299998495380657, |
| "loss": 0.4134, |
| "step": 11170 |
| }, |
| { |
| "epoch": 0.0015672412376776007, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.0002999984924199792, |
| "loss": 0.4347, |
| "step": 11180 |
| }, |
| { |
| "epoch": 0.001568643063471588, |
| "grad_norm": 0.8046875, |
| "learning_rate": 0.0002999984894563914, |
| "loss": 0.4853, |
| "step": 11190 |
| }, |
| { |
| "epoch": 0.001570044889265575, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.0002999984864898935, |
| "loss": 0.4014, |
| "step": 11200 |
| }, |
| { |
| "epoch": 0.0015714467150595621, |
| "grad_norm": 0.95703125, |
| "learning_rate": 0.00029999848352048564, |
| "loss": 0.4483, |
| "step": 11210 |
| }, |
| { |
| "epoch": 0.0015728485408535494, |
| "grad_norm": 0.376953125, |
| "learning_rate": 0.00029999848054816777, |
| "loss": 0.4587, |
| "step": 11220 |
| }, |
| { |
| "epoch": 0.0015742503666475365, |
| "grad_norm": 0.5390625, |
| "learning_rate": 0.0002999984775729398, |
| "loss": 0.3971, |
| "step": 11230 |
| }, |
| { |
| "epoch": 0.0015756521924415235, |
| "grad_norm": 0.9375, |
| "learning_rate": 0.0002999984745948019, |
| "loss": 0.5006, |
| "step": 11240 |
| }, |
| { |
| "epoch": 0.0015770540182355106, |
| "grad_norm": 0.2470703125, |
| "learning_rate": 0.00029999847161375396, |
| "loss": 0.3465, |
| "step": 11250 |
| }, |
| { |
| "epoch": 0.0015784558440294979, |
| "grad_norm": 0.65234375, |
| "learning_rate": 0.00029999846862979594, |
| "loss": 0.4071, |
| "step": 11260 |
| }, |
| { |
| "epoch": 0.001579857669823485, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.0002999984656429279, |
| "loss": 0.4812, |
| "step": 11270 |
| }, |
| { |
| "epoch": 0.001581259495617472, |
| "grad_norm": 0.55859375, |
| "learning_rate": 0.0002999984626531499, |
| "loss": 0.4278, |
| "step": 11280 |
| }, |
| { |
| "epoch": 0.001582661321411459, |
| "grad_norm": 0.96484375, |
| "learning_rate": 0.0002999984596604618, |
| "loss": 0.4164, |
| "step": 11290 |
| }, |
| { |
| "epoch": 0.0015840631472054463, |
| "grad_norm": 0.8046875, |
| "learning_rate": 0.00029999845666486374, |
| "loss": 0.4443, |
| "step": 11300 |
| }, |
| { |
| "epoch": 0.0015854649729994334, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999984536663556, |
| "loss": 0.5397, |
| "step": 11310 |
| }, |
| { |
| "epoch": 0.0015868667987934204, |
| "grad_norm": 1.6015625, |
| "learning_rate": 0.00029999845066493755, |
| "loss": 0.4188, |
| "step": 11320 |
| }, |
| { |
| "epoch": 0.0015882686245874077, |
| "grad_norm": 0.27734375, |
| "learning_rate": 0.00029999844766060937, |
| "loss": 0.397, |
| "step": 11330 |
| }, |
| { |
| "epoch": 0.0015896704503813948, |
| "grad_norm": 0.984375, |
| "learning_rate": 0.0002999984446533712, |
| "loss": 0.3977, |
| "step": 11340 |
| }, |
| { |
| "epoch": 0.0015910722761753818, |
| "grad_norm": 0.83984375, |
| "learning_rate": 0.000299998441643223, |
| "loss": 0.4544, |
| "step": 11350 |
| }, |
| { |
| "epoch": 0.001592474101969369, |
| "grad_norm": 0.96484375, |
| "learning_rate": 0.0002999984386301648, |
| "loss": 0.4528, |
| "step": 11360 |
| }, |
| { |
| "epoch": 0.0015938759277633562, |
| "grad_norm": 0.4921875, |
| "learning_rate": 0.00029999843561419657, |
| "loss": 0.3638, |
| "step": 11370 |
| }, |
| { |
| "epoch": 0.0015952777535573432, |
| "grad_norm": 0.640625, |
| "learning_rate": 0.0002999984325953183, |
| "loss": 0.4077, |
| "step": 11380 |
| }, |
| { |
| "epoch": 0.0015966795793513303, |
| "grad_norm": 0.5546875, |
| "learning_rate": 0.00029999842957353, |
| "loss": 0.4546, |
| "step": 11390 |
| }, |
| { |
| "epoch": 0.0015980814051453174, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.0002999984265488317, |
| "loss": 0.4067, |
| "step": 11400 |
| }, |
| { |
| "epoch": 0.0015994832309393046, |
| "grad_norm": 0.19140625, |
| "learning_rate": 0.0002999984235212234, |
| "loss": 0.4523, |
| "step": 11410 |
| }, |
| { |
| "epoch": 0.0016008850567332917, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.000299998420490705, |
| "loss": 0.4503, |
| "step": 11420 |
| }, |
| { |
| "epoch": 0.0016022868825272788, |
| "grad_norm": 0.953125, |
| "learning_rate": 0.00029999841745727664, |
| "loss": 0.3976, |
| "step": 11430 |
| }, |
| { |
| "epoch": 0.001603688708321266, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999841442093826, |
| "loss": 0.445, |
| "step": 11440 |
| }, |
| { |
| "epoch": 0.001605090534115253, |
| "grad_norm": 0.1728515625, |
| "learning_rate": 0.00029999841138168983, |
| "loss": 0.3924, |
| "step": 11450 |
| }, |
| { |
| "epoch": 0.0016064923599092402, |
| "grad_norm": 0.7890625, |
| "learning_rate": 0.0002999984083395314, |
| "loss": 0.5265, |
| "step": 11460 |
| }, |
| { |
| "epoch": 0.0016078941857032272, |
| "grad_norm": 0.439453125, |
| "learning_rate": 0.00029999840529446297, |
| "loss": 0.4448, |
| "step": 11470 |
| }, |
| { |
| "epoch": 0.0016092960114972145, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.00029999840224648444, |
| "loss": 0.3815, |
| "step": 11480 |
| }, |
| { |
| "epoch": 0.0016106978372912016, |
| "grad_norm": 0.384765625, |
| "learning_rate": 0.00029999839919559596, |
| "loss": 0.4106, |
| "step": 11490 |
| }, |
| { |
| "epoch": 0.0016120996630851886, |
| "grad_norm": 0.3125, |
| "learning_rate": 0.00029999839614179737, |
| "loss": 0.4216, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.0016120996630851886, |
| "eval_loss": 0.175471693277359, |
| "eval_runtime": 112.4366, |
| "eval_samples_per_second": 203.163, |
| "eval_steps_per_second": 1.592, |
| "step": 11500 |
| }, |
| { |
| "epoch": 0.001613501488879176, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999983930850889, |
| "loss": 0.3826, |
| "step": 11510 |
| }, |
| { |
| "epoch": 0.001614903314673163, |
| "grad_norm": 0.4609375, |
| "learning_rate": 0.0002999983900254703, |
| "loss": 0.4388, |
| "step": 11520 |
| }, |
| { |
| "epoch": 0.00161630514046715, |
| "grad_norm": 0.81640625, |
| "learning_rate": 0.0002999983869629417, |
| "loss": 0.4358, |
| "step": 11530 |
| }, |
| { |
| "epoch": 0.001617706966261137, |
| "grad_norm": 0.3984375, |
| "learning_rate": 0.0002999983838975031, |
| "loss": 0.3945, |
| "step": 11540 |
| }, |
| { |
| "epoch": 0.0016191087920551244, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.00029999838082915445, |
| "loss": 0.4987, |
| "step": 11550 |
| }, |
| { |
| "epoch": 0.0016205106178491114, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.0002999983777578958, |
| "loss": 0.4542, |
| "step": 11560 |
| }, |
| { |
| "epoch": 0.0016219124436430985, |
| "grad_norm": 0.25, |
| "learning_rate": 0.00029999837468372713, |
| "loss": 0.4642, |
| "step": 11570 |
| }, |
| { |
| "epoch": 0.0016233142694370855, |
| "grad_norm": 0.2431640625, |
| "learning_rate": 0.0002999983716066484, |
| "loss": 0.4516, |
| "step": 11580 |
| }, |
| { |
| "epoch": 0.0016247160952310728, |
| "grad_norm": 0.5, |
| "learning_rate": 0.0002999983685266597, |
| "loss": 0.425, |
| "step": 11590 |
| }, |
| { |
| "epoch": 0.0016261179210250599, |
| "grad_norm": 1.109375, |
| "learning_rate": 0.00029999836544376097, |
| "loss": 0.488, |
| "step": 11600 |
| }, |
| { |
| "epoch": 0.001627519746819047, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.0002999983623579522, |
| "loss": 0.3917, |
| "step": 11610 |
| }, |
| { |
| "epoch": 0.0016289215726130342, |
| "grad_norm": 0.75, |
| "learning_rate": 0.00029999835926923345, |
| "loss": 0.3768, |
| "step": 11620 |
| }, |
| { |
| "epoch": 0.0016303233984070213, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.00029999835617760466, |
| "loss": 0.3811, |
| "step": 11630 |
| }, |
| { |
| "epoch": 0.0016317252242010083, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.0002999983530830658, |
| "loss": 0.4377, |
| "step": 11640 |
| }, |
| { |
| "epoch": 0.0016331270499949954, |
| "grad_norm": 1.125, |
| "learning_rate": 0.00029999834998561693, |
| "loss": 0.4541, |
| "step": 11650 |
| }, |
| { |
| "epoch": 0.0016345288757889827, |
| "grad_norm": 1.0546875, |
| "learning_rate": 0.0002999983468852581, |
| "loss": 0.4223, |
| "step": 11660 |
| }, |
| { |
| "epoch": 0.0016359307015829697, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.0002999983437819892, |
| "loss": 0.5063, |
| "step": 11670 |
| }, |
| { |
| "epoch": 0.0016373325273769568, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.0002999983406758103, |
| "loss": 0.4157, |
| "step": 11680 |
| }, |
| { |
| "epoch": 0.0016387343531709439, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.00029999833756672137, |
| "loss": 0.4603, |
| "step": 11690 |
| }, |
| { |
| "epoch": 0.0016401361789649311, |
| "grad_norm": 0.53515625, |
| "learning_rate": 0.00029999833445472243, |
| "loss": 0.4326, |
| "step": 11700 |
| }, |
| { |
| "epoch": 0.0016415380047589182, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999833133981344, |
| "loss": 0.355, |
| "step": 11710 |
| }, |
| { |
| "epoch": 0.0016429398305529053, |
| "grad_norm": 0.31640625, |
| "learning_rate": 0.0002999983282219945, |
| "loss": 0.3694, |
| "step": 11720 |
| }, |
| { |
| "epoch": 0.0016443416563468925, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.00029999832510126546, |
| "loss": 0.4451, |
| "step": 11730 |
| }, |
| { |
| "epoch": 0.0016457434821408796, |
| "grad_norm": 1.046875, |
| "learning_rate": 0.0002999983219776264, |
| "loss": 0.5196, |
| "step": 11740 |
| }, |
| { |
| "epoch": 0.0016471453079348667, |
| "grad_norm": 0.486328125, |
| "learning_rate": 0.0002999983188510774, |
| "loss": 0.4154, |
| "step": 11750 |
| }, |
| { |
| "epoch": 0.0016485471337288537, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.0002999983157216183, |
| "loss": 0.4227, |
| "step": 11760 |
| }, |
| { |
| "epoch": 0.001649948959522841, |
| "grad_norm": 0.25, |
| "learning_rate": 0.0002999983125892492, |
| "loss": 0.4026, |
| "step": 11770 |
| }, |
| { |
| "epoch": 0.001651350785316828, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.0002999983094539701, |
| "loss": 0.5898, |
| "step": 11780 |
| }, |
| { |
| "epoch": 0.0016527526111108151, |
| "grad_norm": 0.734375, |
| "learning_rate": 0.00029999830631578095, |
| "loss": 0.3777, |
| "step": 11790 |
| }, |
| { |
| "epoch": 0.0016541544369048024, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.0002999983031746818, |
| "loss": 0.3929, |
| "step": 11800 |
| }, |
| { |
| "epoch": 0.0016555562626987895, |
| "grad_norm": 0.21875, |
| "learning_rate": 0.00029999830003067267, |
| "loss": 0.3358, |
| "step": 11810 |
| }, |
| { |
| "epoch": 0.0016569580884927765, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.00029999829688375347, |
| "loss": 0.4753, |
| "step": 11820 |
| }, |
| { |
| "epoch": 0.0016583599142867636, |
| "grad_norm": 0.71875, |
| "learning_rate": 0.0002999982937339242, |
| "loss": 0.4277, |
| "step": 11830 |
| }, |
| { |
| "epoch": 0.0016597617400807509, |
| "grad_norm": 0.201171875, |
| "learning_rate": 0.00029999829058118503, |
| "loss": 0.3802, |
| "step": 11840 |
| }, |
| { |
| "epoch": 0.001661163565874738, |
| "grad_norm": 0.478515625, |
| "learning_rate": 0.0002999982874255358, |
| "loss": 0.4961, |
| "step": 11850 |
| }, |
| { |
| "epoch": 0.001662565391668725, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.00029999828426697655, |
| "loss": 0.4635, |
| "step": 11860 |
| }, |
| { |
| "epoch": 0.001663967217462712, |
| "grad_norm": 0.40625, |
| "learning_rate": 0.0002999982811055072, |
| "loss": 0.4261, |
| "step": 11870 |
| }, |
| { |
| "epoch": 0.0016653690432566993, |
| "grad_norm": 0.46875, |
| "learning_rate": 0.0002999982779411279, |
| "loss": 0.4546, |
| "step": 11880 |
| }, |
| { |
| "epoch": 0.0016667708690506864, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999827477383855, |
| "loss": 0.3615, |
| "step": 11890 |
| }, |
| { |
| "epoch": 0.0016681726948446734, |
| "grad_norm": 0.82421875, |
| "learning_rate": 0.0002999982716036392, |
| "loss": 0.5243, |
| "step": 11900 |
| }, |
| { |
| "epoch": 0.0016695745206386607, |
| "grad_norm": 0.216796875, |
| "learning_rate": 0.00029999826843052986, |
| "loss": 0.4004, |
| "step": 11910 |
| }, |
| { |
| "epoch": 0.0016709763464326478, |
| "grad_norm": 0.419921875, |
| "learning_rate": 0.00029999826525451046, |
| "loss": 0.4181, |
| "step": 11920 |
| }, |
| { |
| "epoch": 0.0016723781722266348, |
| "grad_norm": 0.515625, |
| "learning_rate": 0.00029999826207558107, |
| "loss": 0.4921, |
| "step": 11930 |
| }, |
| { |
| "epoch": 0.001673779998020622, |
| "grad_norm": 0.9921875, |
| "learning_rate": 0.0002999982588937416, |
| "loss": 0.3902, |
| "step": 11940 |
| }, |
| { |
| "epoch": 0.0016751818238146092, |
| "grad_norm": 0.43359375, |
| "learning_rate": 0.0002999982557089922, |
| "loss": 0.4505, |
| "step": 11950 |
| }, |
| { |
| "epoch": 0.0016765836496085962, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.0002999982525213327, |
| "loss": 0.4233, |
| "step": 11960 |
| }, |
| { |
| "epoch": 0.0016779854754025833, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.0002999982493307633, |
| "loss": 0.442, |
| "step": 11970 |
| }, |
| { |
| "epoch": 0.0016793873011965704, |
| "grad_norm": 0.44921875, |
| "learning_rate": 0.0002999982461372838, |
| "loss": 0.3414, |
| "step": 11980 |
| }, |
| { |
| "epoch": 0.0016807891269905576, |
| "grad_norm": 0.3984375, |
| "learning_rate": 0.00029999824294089423, |
| "loss": 0.4397, |
| "step": 11990 |
| }, |
| { |
| "epoch": 0.0016821909527845447, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.0002999982397415947, |
| "loss": 0.4557, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.0016821909527845447, |
| "eval_loss": 0.1744946837425232, |
| "eval_runtime": 114.0463, |
| "eval_samples_per_second": 200.296, |
| "eval_steps_per_second": 1.57, |
| "step": 12000 |
| }, |
| { |
| "epoch": 0.0016835927785785318, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999982365393852, |
| "loss": 0.4089, |
| "step": 12010 |
| }, |
| { |
| "epoch": 0.001684994604372519, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.0002999982333342656, |
| "loss": 0.441, |
| "step": 12020 |
| }, |
| { |
| "epoch": 0.001686396430166506, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.000299998230126236, |
| "loss": 0.3261, |
| "step": 12030 |
| }, |
| { |
| "epoch": 0.0016877982559604932, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.0002999982269152964, |
| "loss": 0.4775, |
| "step": 12040 |
| }, |
| { |
| "epoch": 0.0016892000817544802, |
| "grad_norm": 0.80078125, |
| "learning_rate": 0.00029999822370144674, |
| "loss": 0.4376, |
| "step": 12050 |
| }, |
| { |
| "epoch": 0.0016906019075484675, |
| "grad_norm": 0.494140625, |
| "learning_rate": 0.0002999982204846871, |
| "loss": 0.4219, |
| "step": 12060 |
| }, |
| { |
| "epoch": 0.0016920037333424546, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.00029999821726501744, |
| "loss": 0.3873, |
| "step": 12070 |
| }, |
| { |
| "epoch": 0.0016934055591364416, |
| "grad_norm": 0.181640625, |
| "learning_rate": 0.00029999821404243774, |
| "loss": 0.4289, |
| "step": 12080 |
| }, |
| { |
| "epoch": 0.001694807384930429, |
| "grad_norm": 0.26171875, |
| "learning_rate": 0.00029999821081694804, |
| "loss": 0.4635, |
| "step": 12090 |
| }, |
| { |
| "epoch": 0.001696209210724416, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.0002999982075885483, |
| "loss": 0.4446, |
| "step": 12100 |
| }, |
| { |
| "epoch": 0.001697611036518403, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.0002999982043572386, |
| "loss": 0.3795, |
| "step": 12110 |
| }, |
| { |
| "epoch": 0.00169901286231239, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.00029999820112301884, |
| "loss": 0.4547, |
| "step": 12120 |
| }, |
| { |
| "epoch": 0.0017004146881063774, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.00029999819788588904, |
| "loss": 0.4917, |
| "step": 12130 |
| }, |
| { |
| "epoch": 0.0017018165139003644, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.00029999819464584924, |
| "loss": 0.3711, |
| "step": 12140 |
| }, |
| { |
| "epoch": 0.0017032183396943515, |
| "grad_norm": 0.28515625, |
| "learning_rate": 0.00029999819140289944, |
| "loss": 0.4076, |
| "step": 12150 |
| }, |
| { |
| "epoch": 0.0017046201654883385, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.00029999818815703964, |
| "loss": 0.4338, |
| "step": 12160 |
| }, |
| { |
| "epoch": 0.0017060219912823258, |
| "grad_norm": 0.404296875, |
| "learning_rate": 0.0002999981849082698, |
| "loss": 0.3822, |
| "step": 12170 |
| }, |
| { |
| "epoch": 0.0017074238170763129, |
| "grad_norm": 0.734375, |
| "learning_rate": 0.00029999818165658994, |
| "loss": 0.4759, |
| "step": 12180 |
| }, |
| { |
| "epoch": 0.0017088256428703, |
| "grad_norm": 0.353515625, |
| "learning_rate": 0.0002999981784020001, |
| "loss": 0.3642, |
| "step": 12190 |
| }, |
| { |
| "epoch": 0.0017102274686642872, |
| "grad_norm": 0.6484375, |
| "learning_rate": 0.00029999817514450014, |
| "loss": 0.461, |
| "step": 12200 |
| }, |
| { |
| "epoch": 0.0017116292944582743, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.00029999817188409024, |
| "loss": 0.4225, |
| "step": 12210 |
| }, |
| { |
| "epoch": 0.0017130311202522613, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.0002999981686207703, |
| "loss": 0.4653, |
| "step": 12220 |
| }, |
| { |
| "epoch": 0.0017144329460462484, |
| "grad_norm": 0.396484375, |
| "learning_rate": 0.00029999816535454033, |
| "loss": 0.4366, |
| "step": 12230 |
| }, |
| { |
| "epoch": 0.0017158347718402357, |
| "grad_norm": 0.6953125, |
| "learning_rate": 0.0002999981620854004, |
| "loss": 0.3372, |
| "step": 12240 |
| }, |
| { |
| "epoch": 0.0017172365976342227, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.0002999981588133504, |
| "loss": 0.3731, |
| "step": 12250 |
| }, |
| { |
| "epoch": 0.0017186384234282098, |
| "grad_norm": 0.484375, |
| "learning_rate": 0.0002999981555383904, |
| "loss": 0.4849, |
| "step": 12260 |
| }, |
| { |
| "epoch": 0.0017200402492221969, |
| "grad_norm": 0.796875, |
| "learning_rate": 0.0002999981522605204, |
| "loss": 0.4366, |
| "step": 12270 |
| }, |
| { |
| "epoch": 0.0017214420750161841, |
| "grad_norm": 0.9375, |
| "learning_rate": 0.0002999981489797403, |
| "loss": 0.4627, |
| "step": 12280 |
| }, |
| { |
| "epoch": 0.0017228439008101712, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.0002999981456960503, |
| "loss": 0.4922, |
| "step": 12290 |
| }, |
| { |
| "epoch": 0.0017242457266041583, |
| "grad_norm": 0.640625, |
| "learning_rate": 0.0002999981424094502, |
| "loss": 0.5141, |
| "step": 12300 |
| }, |
| { |
| "epoch": 0.0017256475523981455, |
| "grad_norm": 0.58984375, |
| "learning_rate": 0.0002999981391199401, |
| "loss": 0.4268, |
| "step": 12310 |
| }, |
| { |
| "epoch": 0.0017270493781921326, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00029999813582752007, |
| "loss": 0.4026, |
| "step": 12320 |
| }, |
| { |
| "epoch": 0.0017284512039861197, |
| "grad_norm": 0.61328125, |
| "learning_rate": 0.0002999981325321899, |
| "loss": 0.3239, |
| "step": 12330 |
| }, |
| { |
| "epoch": 0.0017298530297801067, |
| "grad_norm": 0.244140625, |
| "learning_rate": 0.00029999812923394976, |
| "loss": 0.3925, |
| "step": 12340 |
| }, |
| { |
| "epoch": 0.001731254855574094, |
| "grad_norm": 0.4765625, |
| "learning_rate": 0.0002999981259327996, |
| "loss": 0.3677, |
| "step": 12350 |
| }, |
| { |
| "epoch": 0.001732656681368081, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.00029999812262873946, |
| "loss": 0.4404, |
| "step": 12360 |
| }, |
| { |
| "epoch": 0.0017340585071620681, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.0002999981193217693, |
| "loss": 0.4245, |
| "step": 12370 |
| }, |
| { |
| "epoch": 0.0017354603329560554, |
| "grad_norm": 0.361328125, |
| "learning_rate": 0.00029999811601188906, |
| "loss": 0.4636, |
| "step": 12380 |
| }, |
| { |
| "epoch": 0.0017368621587500424, |
| "grad_norm": 0.30859375, |
| "learning_rate": 0.00029999811269909886, |
| "loss": 0.3831, |
| "step": 12390 |
| }, |
| { |
| "epoch": 0.0017382639845440295, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.00029999810938339866, |
| "loss": 0.4938, |
| "step": 12400 |
| }, |
| { |
| "epoch": 0.0017396658103380166, |
| "grad_norm": 1.0, |
| "learning_rate": 0.0002999981060647884, |
| "loss": 0.4036, |
| "step": 12410 |
| }, |
| { |
| "epoch": 0.0017410676361320038, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.0002999981027432681, |
| "loss": 0.4091, |
| "step": 12420 |
| }, |
| { |
| "epoch": 0.001742469461925991, |
| "grad_norm": 0.443359375, |
| "learning_rate": 0.00029999809941883785, |
| "loss": 0.4134, |
| "step": 12430 |
| }, |
| { |
| "epoch": 0.001743871287719978, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.0002999980960914975, |
| "loss": 0.4591, |
| "step": 12440 |
| }, |
| { |
| "epoch": 0.001745273113513965, |
| "grad_norm": 0.228515625, |
| "learning_rate": 0.00029999809276124725, |
| "loss": 0.4345, |
| "step": 12450 |
| }, |
| { |
| "epoch": 0.0017466749393079523, |
| "grad_norm": 0.71484375, |
| "learning_rate": 0.0002999980894280869, |
| "loss": 0.3948, |
| "step": 12460 |
| }, |
| { |
| "epoch": 0.0017480767651019394, |
| "grad_norm": 0.3515625, |
| "learning_rate": 0.00029999808609201654, |
| "loss": 0.4106, |
| "step": 12470 |
| }, |
| { |
| "epoch": 0.0017494785908959264, |
| "grad_norm": 0.24609375, |
| "learning_rate": 0.0002999980827530362, |
| "loss": 0.4008, |
| "step": 12480 |
| }, |
| { |
| "epoch": 0.0017508804166899137, |
| "grad_norm": 0.23046875, |
| "learning_rate": 0.00029999807941114584, |
| "loss": 0.3922, |
| "step": 12490 |
| }, |
| { |
| "epoch": 0.0017522822424839008, |
| "grad_norm": 0.189453125, |
| "learning_rate": 0.00029999807606634544, |
| "loss": 0.4446, |
| "step": 12500 |
| }, |
| { |
| "epoch": 0.0017522822424839008, |
| "eval_loss": 0.17290730774402618, |
| "eval_runtime": 112.4498, |
| "eval_samples_per_second": 203.139, |
| "eval_steps_per_second": 1.592, |
| "step": 12500 |
| }, |
| { |
| "epoch": 0.0017536840682778878, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.000299998072718635, |
| "loss": 0.4245, |
| "step": 12510 |
| }, |
| { |
| "epoch": 0.0017550858940718749, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.0002999980693680146, |
| "loss": 0.3517, |
| "step": 12520 |
| }, |
| { |
| "epoch": 0.0017564877198658622, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.0002999980660144842, |
| "loss": 0.4871, |
| "step": 12530 |
| }, |
| { |
| "epoch": 0.0017578895456598492, |
| "grad_norm": 0.298828125, |
| "learning_rate": 0.00029999806265804373, |
| "loss": 0.4103, |
| "step": 12540 |
| }, |
| { |
| "epoch": 0.0017592913714538363, |
| "grad_norm": 0.50390625, |
| "learning_rate": 0.00029999805929869323, |
| "loss": 0.4575, |
| "step": 12550 |
| }, |
| { |
| "epoch": 0.0017606931972478234, |
| "grad_norm": 0.8984375, |
| "learning_rate": 0.00029999805593643273, |
| "loss": 0.5051, |
| "step": 12560 |
| }, |
| { |
| "epoch": 0.0017620950230418106, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999980525712622, |
| "loss": 0.4247, |
| "step": 12570 |
| }, |
| { |
| "epoch": 0.0017634968488357977, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.0002999980492031818, |
| "loss": 0.3596, |
| "step": 12580 |
| }, |
| { |
| "epoch": 0.0017648986746297847, |
| "grad_norm": 0.82421875, |
| "learning_rate": 0.00029999804583219123, |
| "loss": 0.4019, |
| "step": 12590 |
| }, |
| { |
| "epoch": 0.001766300500423772, |
| "grad_norm": 0.375, |
| "learning_rate": 0.0002999980424582906, |
| "loss": 0.3925, |
| "step": 12600 |
| }, |
| { |
| "epoch": 0.001767702326217759, |
| "grad_norm": 0.37109375, |
| "learning_rate": 0.00029999803908148013, |
| "loss": 0.492, |
| "step": 12610 |
| }, |
| { |
| "epoch": 0.0017691041520117461, |
| "grad_norm": 0.88671875, |
| "learning_rate": 0.0002999980357017595, |
| "loss": 0.4479, |
| "step": 12620 |
| }, |
| { |
| "epoch": 0.0017705059778057332, |
| "grad_norm": 0.462890625, |
| "learning_rate": 0.0002999980323191289, |
| "loss": 0.4261, |
| "step": 12630 |
| }, |
| { |
| "epoch": 0.0017719078035997205, |
| "grad_norm": 0.9375, |
| "learning_rate": 0.0002999980289335884, |
| "loss": 0.5056, |
| "step": 12640 |
| }, |
| { |
| "epoch": 0.0017733096293937075, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.0002999980255451377, |
| "loss": 0.4311, |
| "step": 12650 |
| }, |
| { |
| "epoch": 0.0017747114551876946, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.00029999802215377713, |
| "loss": 0.5066, |
| "step": 12660 |
| }, |
| { |
| "epoch": 0.0017761132809816819, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.0002999980187595064, |
| "loss": 0.4447, |
| "step": 12670 |
| }, |
| { |
| "epoch": 0.001777515106775669, |
| "grad_norm": 0.85546875, |
| "learning_rate": 0.0002999980153623258, |
| "loss": 0.4514, |
| "step": 12680 |
| }, |
| { |
| "epoch": 0.001778916932569656, |
| "grad_norm": 0.4296875, |
| "learning_rate": 0.0002999980119622351, |
| "loss": 0.3377, |
| "step": 12690 |
| }, |
| { |
| "epoch": 0.001780318758363643, |
| "grad_norm": 0.8203125, |
| "learning_rate": 0.0002999980085592344, |
| "loss": 0.4139, |
| "step": 12700 |
| }, |
| { |
| "epoch": 0.0017817205841576303, |
| "grad_norm": 0.2578125, |
| "learning_rate": 0.00029999800515332373, |
| "loss": 0.3494, |
| "step": 12710 |
| }, |
| { |
| "epoch": 0.0017831224099516174, |
| "grad_norm": 0.4375, |
| "learning_rate": 0.00029999800174450303, |
| "loss": 0.3762, |
| "step": 12720 |
| }, |
| { |
| "epoch": 0.0017845242357456045, |
| "grad_norm": 1.328125, |
| "learning_rate": 0.0002999979983327723, |
| "loss": 0.4438, |
| "step": 12730 |
| }, |
| { |
| "epoch": 0.0017859260615395915, |
| "grad_norm": 0.859375, |
| "learning_rate": 0.00029999799491813153, |
| "loss": 0.359, |
| "step": 12740 |
| }, |
| { |
| "epoch": 0.0017873278873335788, |
| "grad_norm": 0.9375, |
| "learning_rate": 0.0002999979915005808, |
| "loss": 0.5009, |
| "step": 12750 |
| }, |
| { |
| "epoch": 0.0017887297131275659, |
| "grad_norm": 1.015625, |
| "learning_rate": 0.00029999798808012004, |
| "loss": 0.4575, |
| "step": 12760 |
| }, |
| { |
| "epoch": 0.001790131538921553, |
| "grad_norm": 0.5859375, |
| "learning_rate": 0.00029999798465674924, |
| "loss": 0.4969, |
| "step": 12770 |
| }, |
| { |
| "epoch": 0.0017915333647155402, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.00029999798123046844, |
| "loss": 0.5035, |
| "step": 12780 |
| }, |
| { |
| "epoch": 0.0017929351905095273, |
| "grad_norm": 0.47265625, |
| "learning_rate": 0.00029999797780127764, |
| "loss": 0.3995, |
| "step": 12790 |
| }, |
| { |
| "epoch": 0.0017943370163035143, |
| "grad_norm": 0.8203125, |
| "learning_rate": 0.00029999797436917674, |
| "loss": 0.5084, |
| "step": 12800 |
| }, |
| { |
| "epoch": 0.0017957388420975014, |
| "grad_norm": 0.408203125, |
| "learning_rate": 0.00029999797093416595, |
| "loss": 0.4455, |
| "step": 12810 |
| }, |
| { |
| "epoch": 0.0017971406678914887, |
| "grad_norm": 0.87890625, |
| "learning_rate": 0.00029999796749624505, |
| "loss": 0.4266, |
| "step": 12820 |
| }, |
| { |
| "epoch": 0.0017985424936854757, |
| "grad_norm": 0.470703125, |
| "learning_rate": 0.0002999979640554142, |
| "loss": 0.4251, |
| "step": 12830 |
| }, |
| { |
| "epoch": 0.0017999443194794628, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.00029999796061167336, |
| "loss": 0.4692, |
| "step": 12840 |
| }, |
| { |
| "epoch": 0.0018013461452734498, |
| "grad_norm": 0.734375, |
| "learning_rate": 0.00029999795716502246, |
| "loss": 0.5269, |
| "step": 12850 |
| }, |
| { |
| "epoch": 0.0018027479710674371, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.00029999795371546157, |
| "loss": 0.436, |
| "step": 12860 |
| }, |
| { |
| "epoch": 0.0018041497968614242, |
| "grad_norm": 0.353515625, |
| "learning_rate": 0.0002999979502629906, |
| "loss": 0.4383, |
| "step": 12870 |
| }, |
| { |
| "epoch": 0.0018055516226554112, |
| "grad_norm": 0.6953125, |
| "learning_rate": 0.00029999794680760967, |
| "loss": 0.4605, |
| "step": 12880 |
| }, |
| { |
| "epoch": 0.0018069534484493985, |
| "grad_norm": 0.63671875, |
| "learning_rate": 0.0002999979433493187, |
| "loss": 0.4385, |
| "step": 12890 |
| }, |
| { |
| "epoch": 0.0018083552742433856, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.0002999979398881178, |
| "loss": 0.4601, |
| "step": 12900 |
| }, |
| { |
| "epoch": 0.0018097571000373726, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.00029999793642400684, |
| "loss": 0.4269, |
| "step": 12910 |
| }, |
| { |
| "epoch": 0.0018111589258313597, |
| "grad_norm": 0.5, |
| "learning_rate": 0.00029999793295698584, |
| "loss": 0.3948, |
| "step": 12920 |
| }, |
| { |
| "epoch": 0.001812560751625347, |
| "grad_norm": 0.56640625, |
| "learning_rate": 0.0002999979294870548, |
| "loss": 0.412, |
| "step": 12930 |
| }, |
| { |
| "epoch": 0.001813962577419334, |
| "grad_norm": 0.6484375, |
| "learning_rate": 0.0002999979260142138, |
| "loss": 0.4961, |
| "step": 12940 |
| }, |
| { |
| "epoch": 0.001815364403213321, |
| "grad_norm": 0.33984375, |
| "learning_rate": 0.00029999792253846276, |
| "loss": 0.4089, |
| "step": 12950 |
| }, |
| { |
| "epoch": 0.0018167662290073084, |
| "grad_norm": 0.91015625, |
| "learning_rate": 0.00029999791905980177, |
| "loss": 0.388, |
| "step": 12960 |
| }, |
| { |
| "epoch": 0.0018181680548012954, |
| "grad_norm": 0.267578125, |
| "learning_rate": 0.0002999979155782307, |
| "loss": 0.4039, |
| "step": 12970 |
| }, |
| { |
| "epoch": 0.0018195698805952825, |
| "grad_norm": 0.318359375, |
| "learning_rate": 0.00029999791209374963, |
| "loss": 0.4272, |
| "step": 12980 |
| }, |
| { |
| "epoch": 0.0018209717063892696, |
| "grad_norm": 0.7734375, |
| "learning_rate": 0.0002999979086063586, |
| "loss": 0.4296, |
| "step": 12990 |
| }, |
| { |
| "epoch": 0.0018223735321832568, |
| "grad_norm": 0.5078125, |
| "learning_rate": 0.0002999979051160575, |
| "loss": 0.4481, |
| "step": 13000 |
| }, |
| { |
| "epoch": 0.0018223735321832568, |
| "eval_loss": 0.17369048297405243, |
| "eval_runtime": 112.6443, |
| "eval_samples_per_second": 202.789, |
| "eval_steps_per_second": 1.589, |
| "step": 13000 |
| }, |
| { |
| "epoch": 0.001823775357977244, |
| "grad_norm": 1.125, |
| "learning_rate": 0.0002999979016228464, |
| "loss": 0.4938, |
| "step": 13010 |
| }, |
| { |
| "epoch": 0.001825177183771231, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.00029999789812672526, |
| "loss": 0.3418, |
| "step": 13020 |
| }, |
| { |
| "epoch": 0.001826579009565218, |
| "grad_norm": 0.56640625, |
| "learning_rate": 0.00029999789462769417, |
| "loss": 0.4171, |
| "step": 13030 |
| }, |
| { |
| "epoch": 0.0018279808353592053, |
| "grad_norm": 0.6953125, |
| "learning_rate": 0.00029999789112575303, |
| "loss": 0.4075, |
| "step": 13040 |
| }, |
| { |
| "epoch": 0.0018293826611531924, |
| "grad_norm": 0.57421875, |
| "learning_rate": 0.0002999978876209019, |
| "loss": 0.3656, |
| "step": 13050 |
| }, |
| { |
| "epoch": 0.0018307844869471794, |
| "grad_norm": 0.76171875, |
| "learning_rate": 0.0002999978841131407, |
| "loss": 0.4442, |
| "step": 13060 |
| }, |
| { |
| "epoch": 0.0018321863127411667, |
| "grad_norm": 0.333984375, |
| "learning_rate": 0.00029999788060246956, |
| "loss": 0.349, |
| "step": 13070 |
| }, |
| { |
| "epoch": 0.0018335881385351538, |
| "grad_norm": 0.328125, |
| "learning_rate": 0.0002999978770888884, |
| "loss": 0.3392, |
| "step": 13080 |
| }, |
| { |
| "epoch": 0.0018349899643291408, |
| "grad_norm": 0.169921875, |
| "learning_rate": 0.0002999978735723972, |
| "loss": 0.3594, |
| "step": 13090 |
| }, |
| { |
| "epoch": 0.0018363917901231279, |
| "grad_norm": 0.79296875, |
| "learning_rate": 0.000299997870052996, |
| "loss": 0.4715, |
| "step": 13100 |
| }, |
| { |
| "epoch": 0.0018377936159171152, |
| "grad_norm": 0.150390625, |
| "learning_rate": 0.0002999978665306848, |
| "loss": 0.4352, |
| "step": 13110 |
| }, |
| { |
| "epoch": 0.0018391954417111022, |
| "grad_norm": 0.2099609375, |
| "learning_rate": 0.00029999786300546357, |
| "loss": 0.3783, |
| "step": 13120 |
| }, |
| { |
| "epoch": 0.0018405972675050893, |
| "grad_norm": 0.5625, |
| "learning_rate": 0.00029999785947733233, |
| "loss": 0.4245, |
| "step": 13130 |
| }, |
| { |
| "epoch": 0.0018419990932990763, |
| "grad_norm": 0.7578125, |
| "learning_rate": 0.00029999785594629104, |
| "loss": 0.3549, |
| "step": 13140 |
| }, |
| { |
| "epoch": 0.0018434009190930636, |
| "grad_norm": 0.859375, |
| "learning_rate": 0.0002999978524123398, |
| "loss": 0.4832, |
| "step": 13150 |
| }, |
| { |
| "epoch": 0.0018448027448870507, |
| "grad_norm": 0.423828125, |
| "learning_rate": 0.0002999978488754785, |
| "loss": 0.3666, |
| "step": 13160 |
| }, |
| { |
| "epoch": 0.0018462045706810377, |
| "grad_norm": 0.50390625, |
| "learning_rate": 0.00029999784533570723, |
| "loss": 0.3677, |
| "step": 13170 |
| }, |
| { |
| "epoch": 0.001847606396475025, |
| "grad_norm": 0.4453125, |
| "learning_rate": 0.000299997841793026, |
| "loss": 0.465, |
| "step": 13180 |
| }, |
| { |
| "epoch": 0.001849008222269012, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.00029999783824743466, |
| "loss": 0.388, |
| "step": 13190 |
| }, |
| { |
| "epoch": 0.0018504100480629991, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999978346989334, |
| "loss": 0.3845, |
| "step": 13200 |
| }, |
| { |
| "epoch": 0.0018518118738569862, |
| "grad_norm": 1.015625, |
| "learning_rate": 0.00029999783114752204, |
| "loss": 0.3664, |
| "step": 13210 |
| }, |
| { |
| "epoch": 0.0018532136996509735, |
| "grad_norm": 0.87109375, |
| "learning_rate": 0.0002999978275932007, |
| "loss": 0.5365, |
| "step": 13220 |
| }, |
| { |
| "epoch": 0.0018546155254449605, |
| "grad_norm": 0.25, |
| "learning_rate": 0.0002999978240359693, |
| "loss": 0.4392, |
| "step": 13230 |
| }, |
| { |
| "epoch": 0.0018560173512389476, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.000299997820475828, |
| "loss": 0.3636, |
| "step": 13240 |
| }, |
| { |
| "epoch": 0.0018574191770329349, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.00029999781691277666, |
| "loss": 0.4864, |
| "step": 13250 |
| }, |
| { |
| "epoch": 0.001858821002826922, |
| "grad_norm": 0.275390625, |
| "learning_rate": 0.0002999978133468152, |
| "loss": 0.4072, |
| "step": 13260 |
| }, |
| { |
| "epoch": 0.001860222828620909, |
| "grad_norm": 0.55859375, |
| "learning_rate": 0.00029999780977794385, |
| "loss": 0.4495, |
| "step": 13270 |
| }, |
| { |
| "epoch": 0.001861624654414896, |
| "grad_norm": 0.2890625, |
| "learning_rate": 0.00029999780620616247, |
| "loss": 0.3696, |
| "step": 13280 |
| }, |
| { |
| "epoch": 0.0018630264802088833, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.00029999780263147103, |
| "loss": 0.4009, |
| "step": 13290 |
| }, |
| { |
| "epoch": 0.0018644283060028704, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.0002999977990538696, |
| "loss": 0.3759, |
| "step": 13300 |
| }, |
| { |
| "epoch": 0.0018658301317968575, |
| "grad_norm": 0.166015625, |
| "learning_rate": 0.0002999977954733582, |
| "loss": 0.3558, |
| "step": 13310 |
| }, |
| { |
| "epoch": 0.0018672319575908445, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999977918899368, |
| "loss": 0.3603, |
| "step": 13320 |
| }, |
| { |
| "epoch": 0.0018686337833848318, |
| "grad_norm": 0.6875, |
| "learning_rate": 0.0002999977883036053, |
| "loss": 0.4357, |
| "step": 13330 |
| }, |
| { |
| "epoch": 0.0018700356091788189, |
| "grad_norm": 1.0234375, |
| "learning_rate": 0.00029999778471436384, |
| "loss": 0.4183, |
| "step": 13340 |
| }, |
| { |
| "epoch": 0.001871437434972806, |
| "grad_norm": 1.0703125, |
| "learning_rate": 0.0002999977811222124, |
| "loss": 0.4305, |
| "step": 13350 |
| }, |
| { |
| "epoch": 0.0018728392607667932, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.0002999977775271509, |
| "loss": 0.4278, |
| "step": 13360 |
| }, |
| { |
| "epoch": 0.0018742410865607803, |
| "grad_norm": 1.1640625, |
| "learning_rate": 0.0002999977739291794, |
| "loss": 0.4415, |
| "step": 13370 |
| }, |
| { |
| "epoch": 0.0018756429123547673, |
| "grad_norm": 0.84765625, |
| "learning_rate": 0.00029999777032829793, |
| "loss": 0.4681, |
| "step": 13380 |
| }, |
| { |
| "epoch": 0.0018770447381487544, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.00029999776672450646, |
| "loss": 0.4305, |
| "step": 13390 |
| }, |
| { |
| "epoch": 0.0018784465639427417, |
| "grad_norm": 1.0859375, |
| "learning_rate": 0.0002999977631178049, |
| "loss": 0.4568, |
| "step": 13400 |
| }, |
| { |
| "epoch": 0.0018798483897367287, |
| "grad_norm": 0.74609375, |
| "learning_rate": 0.00029999775950819335, |
| "loss": 0.4561, |
| "step": 13410 |
| }, |
| { |
| "epoch": 0.0018812502155307158, |
| "grad_norm": 0.7421875, |
| "learning_rate": 0.00029999775589567183, |
| "loss": 0.4982, |
| "step": 13420 |
| }, |
| { |
| "epoch": 0.0018826520413247028, |
| "grad_norm": 0.3828125, |
| "learning_rate": 0.0002999977522802403, |
| "loss": 0.4583, |
| "step": 13430 |
| }, |
| { |
| "epoch": 0.0018840538671186901, |
| "grad_norm": 1.015625, |
| "learning_rate": 0.0002999977486618988, |
| "loss": 0.5271, |
| "step": 13440 |
| }, |
| { |
| "epoch": 0.0018854556929126772, |
| "grad_norm": 0.69921875, |
| "learning_rate": 0.0002999977450406472, |
| "loss": 0.4311, |
| "step": 13450 |
| }, |
| { |
| "epoch": 0.0018868575187066642, |
| "grad_norm": 0.6640625, |
| "learning_rate": 0.00029999774141648564, |
| "loss": 0.4316, |
| "step": 13460 |
| }, |
| { |
| "epoch": 0.0018882593445006515, |
| "grad_norm": 0.58203125, |
| "learning_rate": 0.000299997737789414, |
| "loss": 0.3692, |
| "step": 13470 |
| }, |
| { |
| "epoch": 0.0018896611702946386, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.00029999773415943245, |
| "loss": 0.4599, |
| "step": 13480 |
| }, |
| { |
| "epoch": 0.0018910629960886256, |
| "grad_norm": 0.8828125, |
| "learning_rate": 0.0002999977305265408, |
| "loss": 0.5223, |
| "step": 13490 |
| }, |
| { |
| "epoch": 0.0018924648218826127, |
| "grad_norm": 1.1875, |
| "learning_rate": 0.0002999977268907392, |
| "loss": 0.4116, |
| "step": 13500 |
| }, |
| { |
| "epoch": 0.0018924648218826127, |
| "eval_loss": 0.17309853434562683, |
| "eval_runtime": 114.1817, |
| "eval_samples_per_second": 200.058, |
| "eval_steps_per_second": 1.568, |
| "step": 13500 |
| }, |
| { |
| "epoch": 0.0018938666476766, |
| "grad_norm": 0.94921875, |
| "learning_rate": 0.00029999772325202764, |
| "loss": 0.519, |
| "step": 13510 |
| }, |
| { |
| "epoch": 0.001895268473470587, |
| "grad_norm": 0.6484375, |
| "learning_rate": 0.00029999771961040597, |
| "loss": 0.4413, |
| "step": 13520 |
| }, |
| { |
| "epoch": 0.001896670299264574, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.00029999771596587435, |
| "loss": 0.4443, |
| "step": 13530 |
| }, |
| { |
| "epoch": 0.0018980721250585614, |
| "grad_norm": 0.81640625, |
| "learning_rate": 0.00029999771231843273, |
| "loss": 0.486, |
| "step": 13540 |
| }, |
| { |
| "epoch": 0.0018994739508525484, |
| "grad_norm": 0.5234375, |
| "learning_rate": 0.000299997708668081, |
| "loss": 0.3188, |
| "step": 13550 |
| }, |
| { |
| "epoch": 0.0019008757766465355, |
| "grad_norm": 0.294921875, |
| "learning_rate": 0.0002999977050148194, |
| "loss": 0.3781, |
| "step": 13560 |
| }, |
| { |
| "epoch": 0.0019022776024405226, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.00029999770135864773, |
| "loss": 0.4946, |
| "step": 13570 |
| }, |
| { |
| "epoch": 0.0019036794282345098, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.00029999769769956607, |
| "loss": 0.3202, |
| "step": 13580 |
| }, |
| { |
| "epoch": 0.001905081254028497, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999769403757435, |
| "loss": 0.4782, |
| "step": 13590 |
| }, |
| { |
| "epoch": 0.001906483079822484, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.0002999976903726727, |
| "loss": 0.4282, |
| "step": 13600 |
| }, |
| { |
| "epoch": 0.001907884905616471, |
| "grad_norm": 0.54296875, |
| "learning_rate": 0.000299997686704861, |
| "loss": 0.4034, |
| "step": 13610 |
| }, |
| { |
| "epoch": 0.0019092867314104583, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.00029999768303413927, |
| "loss": 0.4568, |
| "step": 13620 |
| }, |
| { |
| "epoch": 0.0019106885572044454, |
| "grad_norm": 0.291015625, |
| "learning_rate": 0.00029999767936050755, |
| "loss": 0.3956, |
| "step": 13630 |
| }, |
| { |
| "epoch": 0.0019120903829984324, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.00029999767568396584, |
| "loss": 0.4545, |
| "step": 13640 |
| }, |
| { |
| "epoch": 0.0019134922087924197, |
| "grad_norm": 0.30078125, |
| "learning_rate": 0.00029999767200451414, |
| "loss": 0.489, |
| "step": 13650 |
| }, |
| { |
| "epoch": 0.0019148940345864068, |
| "grad_norm": 1.28125, |
| "learning_rate": 0.0002999976683221524, |
| "loss": 0.536, |
| "step": 13660 |
| }, |
| { |
| "epoch": 0.0019162958603803938, |
| "grad_norm": 0.2578125, |
| "learning_rate": 0.0002999976646368806, |
| "loss": 0.3801, |
| "step": 13670 |
| }, |
| { |
| "epoch": 0.0019176976861743809, |
| "grad_norm": 0.52734375, |
| "learning_rate": 0.00029999766094869886, |
| "loss": 0.4321, |
| "step": 13680 |
| }, |
| { |
| "epoch": 0.0019190995119683682, |
| "grad_norm": 0.25390625, |
| "learning_rate": 0.0002999976572576071, |
| "loss": 0.3094, |
| "step": 13690 |
| }, |
| { |
| "epoch": 0.0019205013377623552, |
| "grad_norm": 0.39453125, |
| "learning_rate": 0.00029999765356360534, |
| "loss": 0.3829, |
| "step": 13700 |
| }, |
| { |
| "epoch": 0.0019219031635563423, |
| "grad_norm": 0.423828125, |
| "learning_rate": 0.00029999764986669353, |
| "loss": 0.4466, |
| "step": 13710 |
| }, |
| { |
| "epoch": 0.0019233049893503293, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.0002999976461668718, |
| "loss": 0.4148, |
| "step": 13720 |
| }, |
| { |
| "epoch": 0.0019247068151443166, |
| "grad_norm": 0.83984375, |
| "learning_rate": 0.00029999764246414, |
| "loss": 0.4261, |
| "step": 13730 |
| }, |
| { |
| "epoch": 0.0019261086409383037, |
| "grad_norm": 0.48828125, |
| "learning_rate": 0.00029999763875849817, |
| "loss": 0.4347, |
| "step": 13740 |
| }, |
| { |
| "epoch": 0.0019275104667322907, |
| "grad_norm": 0.515625, |
| "learning_rate": 0.00029999763504994637, |
| "loss": 0.4352, |
| "step": 13750 |
| }, |
| { |
| "epoch": 0.001928912292526278, |
| "grad_norm": 0.5703125, |
| "learning_rate": 0.00029999763133848456, |
| "loss": 0.428, |
| "step": 13760 |
| }, |
| { |
| "epoch": 0.001930314118320265, |
| "grad_norm": 0.45703125, |
| "learning_rate": 0.00029999762762411276, |
| "loss": 0.337, |
| "step": 13770 |
| }, |
| { |
| "epoch": 0.0019317159441142521, |
| "grad_norm": 0.60546875, |
| "learning_rate": 0.00029999762390683096, |
| "loss": 0.3705, |
| "step": 13780 |
| }, |
| { |
| "epoch": 0.0019331177699082392, |
| "grad_norm": 0.486328125, |
| "learning_rate": 0.0002999976201866391, |
| "loss": 0.4369, |
| "step": 13790 |
| }, |
| { |
| "epoch": 0.0019345195957022265, |
| "grad_norm": 0.609375, |
| "learning_rate": 0.00029999761646353726, |
| "loss": 0.4915, |
| "step": 13800 |
| }, |
| { |
| "epoch": 0.0019359214214962135, |
| "grad_norm": 0.984375, |
| "learning_rate": 0.0002999976127375254, |
| "loss": 0.4433, |
| "step": 13810 |
| }, |
| { |
| "epoch": 0.0019373232472902006, |
| "grad_norm": 0.443359375, |
| "learning_rate": 0.00029999760900860356, |
| "loss": 0.465, |
| "step": 13820 |
| }, |
| { |
| "epoch": 0.0019387250730841879, |
| "grad_norm": 0.9140625, |
| "learning_rate": 0.0002999976052767717, |
| "loss": 0.4211, |
| "step": 13830 |
| }, |
| { |
| "epoch": 0.001940126898878175, |
| "grad_norm": 0.357421875, |
| "learning_rate": 0.00029999760154202987, |
| "loss": 0.413, |
| "step": 13840 |
| }, |
| { |
| "epoch": 0.001941528724672162, |
| "grad_norm": 0.490234375, |
| "learning_rate": 0.000299997597804378, |
| "loss": 0.3958, |
| "step": 13850 |
| }, |
| { |
| "epoch": 0.001942930550466149, |
| "grad_norm": 0.89453125, |
| "learning_rate": 0.00029999759406381613, |
| "loss": 0.3903, |
| "step": 13860 |
| }, |
| { |
| "epoch": 0.0019443323762601363, |
| "grad_norm": 0.33984375, |
| "learning_rate": 0.00029999759032034424, |
| "loss": 0.3797, |
| "step": 13870 |
| }, |
| { |
| "epoch": 0.0019457342020541234, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.00029999758657396234, |
| "loss": 0.5, |
| "step": 13880 |
| }, |
| { |
| "epoch": 0.0019471360278481105, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.0002999975828246705, |
| "loss": 0.4632, |
| "step": 13890 |
| }, |
| { |
| "epoch": 0.0019485378536420975, |
| "grad_norm": 1.09375, |
| "learning_rate": 0.00029999757907246856, |
| "loss": 0.4769, |
| "step": 13900 |
| }, |
| { |
| "epoch": 0.0019499396794360848, |
| "grad_norm": 0.1591796875, |
| "learning_rate": 0.0002999975753173566, |
| "loss": 0.409, |
| "step": 13910 |
| }, |
| { |
| "epoch": 0.0019513415052300719, |
| "grad_norm": 0.625, |
| "learning_rate": 0.0002999975715593347, |
| "loss": 0.4453, |
| "step": 13920 |
| }, |
| { |
| "epoch": 0.001952743331024059, |
| "grad_norm": 0.77734375, |
| "learning_rate": 0.00029999756779840284, |
| "loss": 0.4729, |
| "step": 13930 |
| }, |
| { |
| "epoch": 0.001954145156818046, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.0002999975640345609, |
| "loss": 0.4493, |
| "step": 13940 |
| }, |
| { |
| "epoch": 0.0019555469826120333, |
| "grad_norm": 0.341796875, |
| "learning_rate": 0.00029999756026780896, |
| "loss": 0.3759, |
| "step": 13950 |
| }, |
| { |
| "epoch": 0.0019569488084060205, |
| "grad_norm": 1.171875, |
| "learning_rate": 0.0002999975564981471, |
| "loss": 0.4879, |
| "step": 13960 |
| }, |
| { |
| "epoch": 0.0019583506342000074, |
| "grad_norm": 0.9140625, |
| "learning_rate": 0.0002999975527255751, |
| "loss": 0.4767, |
| "step": 13970 |
| }, |
| { |
| "epoch": 0.0019597524599939947, |
| "grad_norm": 0.28515625, |
| "learning_rate": 0.00029999754895009315, |
| "loss": 0.4501, |
| "step": 13980 |
| }, |
| { |
| "epoch": 0.0019611542857879815, |
| "grad_norm": 0.7109375, |
| "learning_rate": 0.0002999975451717012, |
| "loss": 0.4216, |
| "step": 13990 |
| }, |
| { |
| "epoch": 0.0019625561115819688, |
| "grad_norm": 0.345703125, |
| "learning_rate": 0.0002999975413903993, |
| "loss": 0.3885, |
| "step": 14000 |
| }, |
| { |
| "epoch": 0.0019625561115819688, |
| "eval_loss": 0.17219418287277222, |
| "eval_runtime": 112.3232, |
| "eval_samples_per_second": 203.369, |
| "eval_steps_per_second": 1.594, |
| "step": 14000 |
| }, |
| { |
| "epoch": 0.001963957937375956, |
| "grad_norm": 0.56640625, |
| "learning_rate": 0.0002999975376061873, |
| "loss": 0.412, |
| "step": 14010 |
| }, |
| { |
| "epoch": 0.001965359763169943, |
| "grad_norm": 0.337890625, |
| "learning_rate": 0.0002999975338190653, |
| "loss": 0.3503, |
| "step": 14020 |
| }, |
| { |
| "epoch": 0.00196676158896393, |
| "grad_norm": 0.83984375, |
| "learning_rate": 0.00029999753002903334, |
| "loss": 0.3654, |
| "step": 14030 |
| }, |
| { |
| "epoch": 0.0019681634147579175, |
| "grad_norm": 0.6484375, |
| "learning_rate": 0.0002999975262360914, |
| "loss": 0.5271, |
| "step": 14040 |
| }, |
| { |
| "epoch": 0.0019695652405519043, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.00029999752244023943, |
| "loss": 0.3611, |
| "step": 14050 |
| }, |
| { |
| "epoch": 0.0019709670663458916, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.0002999975186414774, |
| "loss": 0.475, |
| "step": 14060 |
| }, |
| { |
| "epoch": 0.001972368892139879, |
| "grad_norm": 0.6015625, |
| "learning_rate": 0.0002999975148398054, |
| "loss": 0.4544, |
| "step": 14070 |
| }, |
| { |
| "epoch": 0.0019737707179338657, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.00029999751103522345, |
| "loss": 0.371, |
| "step": 14080 |
| }, |
| { |
| "epoch": 0.001975172543727853, |
| "grad_norm": 0.2470703125, |
| "learning_rate": 0.0002999975072277314, |
| "loss": 0.4124, |
| "step": 14090 |
| }, |
| { |
| "epoch": 0.00197657436952184, |
| "grad_norm": 0.298828125, |
| "learning_rate": 0.00029999750341732945, |
| "loss": 0.4765, |
| "step": 14100 |
| }, |
| { |
| "epoch": 0.001977976195315827, |
| "grad_norm": 0.68359375, |
| "learning_rate": 0.0002999974996040174, |
| "loss": 0.4405, |
| "step": 14110 |
| }, |
| { |
| "epoch": 0.0019793780211098144, |
| "grad_norm": 0.69140625, |
| "learning_rate": 0.00029999749578779546, |
| "loss": 0.5118, |
| "step": 14120 |
| }, |
| { |
| "epoch": 0.001980779846903801, |
| "grad_norm": 0.53125, |
| "learning_rate": 0.00029999749196866343, |
| "loss": 0.4831, |
| "step": 14130 |
| }, |
| { |
| "epoch": 0.0019821816726977885, |
| "grad_norm": 0.6328125, |
| "learning_rate": 0.00029999748814662136, |
| "loss": 0.5109, |
| "step": 14140 |
| }, |
| { |
| "epoch": 0.0019835834984917758, |
| "grad_norm": 0.58984375, |
| "learning_rate": 0.00029999748432166934, |
| "loss": 0.3219, |
| "step": 14150 |
| }, |
| { |
| "epoch": 0.0019849853242857626, |
| "grad_norm": 0.515625, |
| "learning_rate": 0.0002999974804938073, |
| "loss": 0.3513, |
| "step": 14160 |
| }, |
| { |
| "epoch": 0.00198638715007975, |
| "grad_norm": 0.90625, |
| "learning_rate": 0.0002999974766630353, |
| "loss": 0.4944, |
| "step": 14170 |
| }, |
| { |
| "epoch": 0.001987788975873737, |
| "grad_norm": 0.91015625, |
| "learning_rate": 0.0002999974728293533, |
| "loss": 0.4509, |
| "step": 14180 |
| }, |
| { |
| "epoch": 0.001989190801667724, |
| "grad_norm": 1.1171875, |
| "learning_rate": 0.0002999974689927613, |
| "loss": 0.4779, |
| "step": 14190 |
| }, |
| { |
| "epoch": 0.0019905926274617113, |
| "grad_norm": 0.578125, |
| "learning_rate": 0.0002999974651532592, |
| "loss": 0.4435, |
| "step": 14200 |
| }, |
| { |
| "epoch": 0.0019919944532556986, |
| "grad_norm": 0.62109375, |
| "learning_rate": 0.00029999746131084714, |
| "loss": 0.4092, |
| "step": 14210 |
| }, |
| { |
| "epoch": 0.0019933962790496854, |
| "grad_norm": 1.1015625, |
| "learning_rate": 0.0002999974574655251, |
| "loss": 0.4802, |
| "step": 14220 |
| }, |
| { |
| "epoch": 0.0019947981048436727, |
| "grad_norm": 0.78125, |
| "learning_rate": 0.000299997453617293, |
| "loss": 0.4033, |
| "step": 14230 |
| }, |
| { |
| "epoch": 0.0019961999306376595, |
| "grad_norm": 0.494140625, |
| "learning_rate": 0.00029999744976615095, |
| "loss": 0.3894, |
| "step": 14240 |
| }, |
| { |
| "epoch": 0.001997601756431647, |
| "grad_norm": 0.671875, |
| "learning_rate": 0.00029999744591209895, |
| "loss": 0.35, |
| "step": 14250 |
| }, |
| { |
| "epoch": 0.001999003582225634, |
| "grad_norm": 0.1357421875, |
| "learning_rate": 0.00029999744205513683, |
| "loss": 0.4277, |
| "step": 14260 |
| }, |
| { |
| "epoch": 0.002000405408019621, |
| "grad_norm": 0.193359375, |
| "learning_rate": 0.0002999974381952648, |
| "loss": 0.3722, |
| "step": 14270 |
| }, |
| { |
| "epoch": 0.002001807233813608, |
| "grad_norm": 0.83203125, |
| "learning_rate": 0.0002999974343324827, |
| "loss": 0.4317, |
| "step": 14280 |
| }, |
| { |
| "epoch": 0.0020032090596075955, |
| "grad_norm": 1.1171875, |
| "learning_rate": 0.0002999974304667906, |
| "loss": 0.4286, |
| "step": 14290 |
| }, |
| { |
| "epoch": 0.0020046108854015823, |
| "grad_norm": 0.8359375, |
| "learning_rate": 0.0002999974265981886, |
| "loss": 0.4539, |
| "step": 14300 |
| }, |
| { |
| "epoch": 0.0020060127111955696, |
| "grad_norm": 0.7265625, |
| "learning_rate": 0.0002999974227266765, |
| "loss": 0.4675, |
| "step": 14310 |
| }, |
| { |
| "epoch": 0.002007414536989557, |
| "grad_norm": 0.23046875, |
| "learning_rate": 0.00029999741885225446, |
| "loss": 0.4397, |
| "step": 14320 |
| }, |
| { |
| "epoch": 0.0020088163627835437, |
| "grad_norm": 0.890625, |
| "learning_rate": 0.0002999974149749223, |
| "loss": 0.4109, |
| "step": 14330 |
| }, |
| { |
| "epoch": 0.002010218188577531, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.00029999741109468025, |
| "loss": 0.3804, |
| "step": 14340 |
| }, |
| { |
| "epoch": 0.002011620014371518, |
| "grad_norm": 0.435546875, |
| "learning_rate": 0.00029999740721152815, |
| "loss": 0.3754, |
| "step": 14350 |
| }, |
| { |
| "epoch": 0.002013021840165505, |
| "grad_norm": 0.234375, |
| "learning_rate": 0.00029999740332546605, |
| "loss": 0.3823, |
| "step": 14360 |
| }, |
| { |
| "epoch": 0.0020144236659594924, |
| "grad_norm": 0.765625, |
| "learning_rate": 0.00029999739943649395, |
| "loss": 0.3991, |
| "step": 14370 |
| }, |
| { |
| "epoch": 0.0020158254917534793, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.0002999973955446119, |
| "loss": 0.4227, |
| "step": 14380 |
| }, |
| { |
| "epoch": 0.0020172273175474665, |
| "grad_norm": 1.1640625, |
| "learning_rate": 0.0002999973916498198, |
| "loss": 0.3951, |
| "step": 14390 |
| }, |
| { |
| "epoch": 0.002018629143341454, |
| "grad_norm": 0.55078125, |
| "learning_rate": 0.0002999973877521177, |
| "loss": 0.4999, |
| "step": 14400 |
| }, |
| { |
| "epoch": 0.0020200309691354407, |
| "grad_norm": 0.89453125, |
| "learning_rate": 0.0002999973838515056, |
| "loss": 0.4185, |
| "step": 14410 |
| }, |
| { |
| "epoch": 0.002021432794929428, |
| "grad_norm": 0.53515625, |
| "learning_rate": 0.0002999973799479835, |
| "loss": 0.4898, |
| "step": 14420 |
| }, |
| { |
| "epoch": 0.002022834620723415, |
| "grad_norm": 0.388671875, |
| "learning_rate": 0.0002999973760415514, |
| "loss": 0.4813, |
| "step": 14430 |
| }, |
| { |
| "epoch": 0.002024236446517402, |
| "grad_norm": 0.5390625, |
| "learning_rate": 0.00029999737213220925, |
| "loss": 0.4236, |
| "step": 14440 |
| }, |
| { |
| "epoch": 0.0020256382723113893, |
| "grad_norm": 0.3125, |
| "learning_rate": 0.00029999736821995716, |
| "loss": 0.4339, |
| "step": 14450 |
| }, |
| { |
| "epoch": 0.002027040098105376, |
| "grad_norm": 0.6796875, |
| "learning_rate": 0.000299997364304795, |
| "loss": 0.4094, |
| "step": 14460 |
| }, |
| { |
| "epoch": 0.0020284419238993634, |
| "grad_norm": 0.19921875, |
| "learning_rate": 0.00029999736038672294, |
| "loss": 0.4625, |
| "step": 14470 |
| }, |
| { |
| "epoch": 0.0020298437496933507, |
| "grad_norm": 0.66796875, |
| "learning_rate": 0.0002999973564657408, |
| "loss": 0.33, |
| "step": 14480 |
| }, |
| { |
| "epoch": 0.0020312455754873376, |
| "grad_norm": 0.5, |
| "learning_rate": 0.00029999735254184867, |
| "loss": 0.4033, |
| "step": 14490 |
| }, |
| { |
| "epoch": 0.002032647401281325, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.0002999973486150466, |
| "loss": 0.4106, |
| "step": 14500 |
| }, |
| { |
| "epoch": 0.002032647401281325, |
| "eval_loss": 0.1713680624961853, |
| "eval_runtime": 112.5849, |
| "eval_samples_per_second": 202.896, |
| "eval_steps_per_second": 1.59, |
| "step": 14500 |
| }, |
| { |
| "epoch": 0.002034049227075312, |
| "grad_norm": 0.84375, |
| "learning_rate": 0.00029999734468533445, |
| "loss": 0.3933, |
| "step": 14510 |
| }, |
| { |
| "epoch": 0.002035451052869299, |
| "grad_norm": 0.39453125, |
| "learning_rate": 0.0002999973407527123, |
| "loss": 0.4153, |
| "step": 14520 |
| }, |
| { |
| "epoch": 0.0020368528786632862, |
| "grad_norm": 0.234375, |
| "learning_rate": 0.0002999973368171802, |
| "loss": 0.4338, |
| "step": 14530 |
| }, |
| { |
| "epoch": 0.0020382547044572735, |
| "grad_norm": 0.470703125, |
| "learning_rate": 0.00029999733287873806, |
| "loss": 0.4352, |
| "step": 14540 |
| }, |
| { |
| "epoch": 0.0020396565302512604, |
| "grad_norm": 0.828125, |
| "learning_rate": 0.00029999732893738594, |
| "loss": 0.4688, |
| "step": 14550 |
| }, |
| { |
| "epoch": 0.0020410583560452476, |
| "grad_norm": 0.35546875, |
| "learning_rate": 0.0002999973249931238, |
| "loss": 0.436, |
| "step": 14560 |
| }, |
| { |
| "epoch": 0.0020424601818392345, |
| "grad_norm": 0.65625, |
| "learning_rate": 0.0002999973210459517, |
| "loss": 0.3661, |
| "step": 14570 |
| }, |
| { |
| "epoch": 0.0020438620076332218, |
| "grad_norm": 0.482421875, |
| "learning_rate": 0.00029999731709586956, |
| "loss": 0.3806, |
| "step": 14580 |
| }, |
| { |
| "epoch": 0.002045263833427209, |
| "grad_norm": 0.66015625, |
| "learning_rate": 0.00029999731314287744, |
| "loss": 0.4362, |
| "step": 14590 |
| }, |
| { |
| "epoch": 0.002046665659221196, |
| "grad_norm": 0.75390625, |
| "learning_rate": 0.0002999973091869753, |
| "loss": 0.3717, |
| "step": 14600 |
| }, |
| { |
| "epoch": 0.002048067485015183, |
| "grad_norm": 0.412109375, |
| "learning_rate": 0.0002999973052281632, |
| "loss": 0.4599, |
| "step": 14610 |
| }, |
| { |
| "epoch": 0.0020494693108091704, |
| "grad_norm": 0.291015625, |
| "learning_rate": 0.0002999973012664411, |
| "loss": 0.4119, |
| "step": 14620 |
| }, |
| { |
| "epoch": 0.0020508711366031573, |
| "grad_norm": 0.376953125, |
| "learning_rate": 0.0002999972973018089, |
| "loss": 0.3842, |
| "step": 14630 |
| }, |
| { |
| "epoch": 0.0020522729623971446, |
| "grad_norm": 0.53515625, |
| "learning_rate": 0.00029999729333426685, |
| "loss": 0.5399, |
| "step": 14640 |
| }, |
| { |
| "epoch": 0.002053674788191132, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.0002999972893638147, |
| "loss": 0.3464, |
| "step": 14650 |
| }, |
| { |
| "epoch": 0.0020550766139851187, |
| "grad_norm": 0.4609375, |
| "learning_rate": 0.00029999728539045257, |
| "loss": 0.466, |
| "step": 14660 |
| }, |
| { |
| "epoch": 0.002056478439779106, |
| "grad_norm": 0.455078125, |
| "learning_rate": 0.0002999972814141804, |
| "loss": 0.4817, |
| "step": 14670 |
| }, |
| { |
| "epoch": 0.002057880265573093, |
| "grad_norm": 0.32421875, |
| "learning_rate": 0.00029999727743499824, |
| "loss": 0.3673, |
| "step": 14680 |
| }, |
| { |
| "epoch": 0.00205928209136708, |
| "grad_norm": 1.0, |
| "learning_rate": 0.00029999727345290613, |
| "loss": 0.4989, |
| "step": 14690 |
| }, |
| { |
| "epoch": 0.0020606839171610674, |
| "grad_norm": 0.72265625, |
| "learning_rate": 0.000299997269467904, |
| "loss": 0.4511, |
| "step": 14700 |
| }, |
| { |
| "epoch": 0.002062085742955054, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.00029999726547999187, |
| "loss": 0.5199, |
| "step": 14710 |
| }, |
| { |
| "epoch": 0.0020634875687490415, |
| "grad_norm": 0.9140625, |
| "learning_rate": 0.00029999726148916976, |
| "loss": 0.4247, |
| "step": 14720 |
| }, |
| { |
| "epoch": 0.0020648893945430288, |
| "grad_norm": 0.2158203125, |
| "learning_rate": 0.0002999972574954376, |
| "loss": 0.3859, |
| "step": 14730 |
| }, |
| { |
| "epoch": 0.0020662912203370156, |
| "grad_norm": 0.6171875, |
| "learning_rate": 0.0002999972534987955, |
| "loss": 0.4503, |
| "step": 14740 |
| }, |
| { |
| "epoch": 0.002067693046131003, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.00029999724949924335, |
| "loss": 0.4461, |
| "step": 14750 |
| }, |
| { |
| "epoch": 0.00206909487192499, |
| "grad_norm": 0.8671875, |
| "learning_rate": 0.00029999724549678125, |
| "loss": 0.4595, |
| "step": 14760 |
| }, |
| { |
| "epoch": 0.002070496697718977, |
| "grad_norm": 0.59765625, |
| "learning_rate": 0.0002999972414914091, |
| "loss": 0.325, |
| "step": 14770 |
| }, |
| { |
| "epoch": 0.0020718985235129643, |
| "grad_norm": 0.62890625, |
| "learning_rate": 0.000299997237483127, |
| "loss": 0.384, |
| "step": 14780 |
| }, |
| { |
| "epoch": 0.0020733003493069516, |
| "grad_norm": 0.64453125, |
| "learning_rate": 0.00029999723347193485, |
| "loss": 0.3865, |
| "step": 14790 |
| }, |
| { |
| "epoch": 0.0020747021751009384, |
| "grad_norm": 0.318359375, |
| "learning_rate": 0.00029999722945783276, |
| "loss": 0.4034, |
| "step": 14800 |
| }, |
| { |
| "epoch": 0.0020761040008949257, |
| "grad_norm": 0.373046875, |
| "learning_rate": 0.0002999972254408206, |
| "loss": 0.3206, |
| "step": 14810 |
| }, |
| { |
| "epoch": 0.0020775058266889125, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.0002999972214208985, |
| "loss": 0.5055, |
| "step": 14820 |
| }, |
| { |
| "epoch": 0.0020789076524829, |
| "grad_norm": 1.09375, |
| "learning_rate": 0.0002999972173980664, |
| "loss": 0.5426, |
| "step": 14830 |
| }, |
| { |
| "epoch": 0.002080309478276887, |
| "grad_norm": 0.416015625, |
| "learning_rate": 0.00029999721337232424, |
| "loss": 0.5022, |
| "step": 14840 |
| }, |
| { |
| "epoch": 0.002081711304070874, |
| "grad_norm": 1.0390625, |
| "learning_rate": 0.0002999972093436721, |
| "loss": 0.4795, |
| "step": 14850 |
| }, |
| { |
| "epoch": 0.002083113129864861, |
| "grad_norm": 0.73828125, |
| "learning_rate": 0.00029999720531211, |
| "loss": 0.5243, |
| "step": 14860 |
| }, |
| { |
| "epoch": 0.0020845149556588485, |
| "grad_norm": 0.51953125, |
| "learning_rate": 0.00029999720127763793, |
| "loss": 0.4725, |
| "step": 14870 |
| }, |
| { |
| "epoch": 0.0020859167814528353, |
| "grad_norm": 0.73046875, |
| "learning_rate": 0.00029999719724025574, |
| "loss": 0.4518, |
| "step": 14880 |
| }, |
| { |
| "epoch": 0.0020873186072468226, |
| "grad_norm": 0.78515625, |
| "learning_rate": 0.00029999719319996366, |
| "loss": 0.3732, |
| "step": 14890 |
| }, |
| { |
| "epoch": 0.00208872043304081, |
| "grad_norm": 0.67578125, |
| "learning_rate": 0.0002999971891567615, |
| "loss": 0.5069, |
| "step": 14900 |
| }, |
| { |
| "epoch": 0.0020901222588347967, |
| "grad_norm": 0.486328125, |
| "learning_rate": 0.00029999718511064945, |
| "loss": 0.4604, |
| "step": 14910 |
| }, |
| { |
| "epoch": 0.002091524084628784, |
| "grad_norm": 0.5078125, |
| "learning_rate": 0.0002999971810616273, |
| "loss": 0.4016, |
| "step": 14920 |
| }, |
| { |
| "epoch": 0.002092925910422771, |
| "grad_norm": 0.2080078125, |
| "learning_rate": 0.00029999717700969524, |
| "loss": 0.3554, |
| "step": 14930 |
| }, |
| { |
| "epoch": 0.002094327736216758, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.0002999971729548531, |
| "loss": 0.3677, |
| "step": 14940 |
| }, |
| { |
| "epoch": 0.0020957295620107454, |
| "grad_norm": 0.703125, |
| "learning_rate": 0.000299997168897101, |
| "loss": 0.4108, |
| "step": 14950 |
| }, |
| { |
| "epoch": 0.0020971313878047322, |
| "grad_norm": 0.82421875, |
| "learning_rate": 0.0002999971648364389, |
| "loss": 0.5519, |
| "step": 14960 |
| }, |
| { |
| "epoch": 0.0020985332135987195, |
| "grad_norm": 0.296875, |
| "learning_rate": 0.0002999971607728668, |
| "loss": 0.4011, |
| "step": 14970 |
| }, |
| { |
| "epoch": 0.002099935039392707, |
| "grad_norm": 0.20703125, |
| "learning_rate": 0.0002999971567063847, |
| "loss": 0.2955, |
| "step": 14980 |
| }, |
| { |
| "epoch": 0.0021013368651866936, |
| "grad_norm": 0.859375, |
| "learning_rate": 0.0002999971526369926, |
| "loss": 0.4985, |
| "step": 14990 |
| }, |
| { |
| "epoch": 0.002102738690980681, |
| "grad_norm": 0.87890625, |
| "learning_rate": 0.00029999714856469053, |
| "loss": 0.3839, |
| "step": 15000 |
| }, |
| { |
| "epoch": 0.002102738690980681, |
| "eval_loss": 0.17203140258789062, |
| "eval_runtime": 114.2354, |
| "eval_samples_per_second": 199.964, |
| "eval_steps_per_second": 1.567, |
| "step": 15000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 7133554, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 5000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.0473171539266765e+17, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
