| { | |
| "best_metric": 25.2438, | |
| "best_model_checkpoint": "/exports/eddie/scratch/s1970716/models/longt5_xl_sfd_bp_15/checkpoint-201", | |
| "epoch": 14.608695652173914, | |
| "eval_steps": 500, | |
| "global_step": 210, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 4.218067169189453, | |
| "learning_rate": 0.001, | |
| "loss": 2.9668, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 200.38365173339844, | |
| "learning_rate": 0.001, | |
| "loss": 3.3198, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 74.55081176757812, | |
| "learning_rate": 0.001, | |
| "loss": 2.6874, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 19.85554313659668, | |
| "learning_rate": 0.001, | |
| "loss": 2.3138, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 41.041751861572266, | |
| "learning_rate": 0.001, | |
| "loss": 2.0222, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 3.789278507232666, | |
| "learning_rate": 0.001, | |
| "loss": 1.8363, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 5.874878883361816, | |
| "learning_rate": 0.001, | |
| "loss": 2.5763, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_gen_len": 509.64792899408286, | |
| "eval_loss": 2.541541814804077, | |
| "eval_rouge1": 10.6052, | |
| "eval_rouge2": 1.4494, | |
| "eval_rougeL": 10.4593, | |
| "eval_rougeLsum": 10.4801, | |
| "eval_runtime": 1798.047, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 1.338165521621704, | |
| "learning_rate": 0.001, | |
| "loss": 2.4441, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 3.755629777908325, | |
| "learning_rate": 0.001, | |
| "loss": 2.258, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 6.490938663482666, | |
| "learning_rate": 0.001, | |
| "loss": 3.0147, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 5.593593597412109, | |
| "learning_rate": 0.001, | |
| "loss": 2.4724, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.0521235466003418, | |
| "learning_rate": 0.001, | |
| "loss": 2.023, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 12.585270881652832, | |
| "learning_rate": 0.001, | |
| "loss": 2.223, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 78.1630630493164, | |
| "learning_rate": 0.001, | |
| "loss": 1.8998, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "eval_gen_len": 511.0, | |
| "eval_loss": 1.739753246307373, | |
| "eval_rouge1": 16.7989, | |
| "eval_rouge2": 4.1457, | |
| "eval_rougeL": 16.4049, | |
| "eval_rougeLsum": 15.1803, | |
| "eval_runtime": 1798.9905, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "grad_norm": 0.7915446758270264, | |
| "learning_rate": 0.001, | |
| "loss": 1.8375, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 1.5996413230895996, | |
| "learning_rate": 0.001, | |
| "loss": 2.0326, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 1.0431970357894897, | |
| "learning_rate": 0.001, | |
| "loss": 2.1242, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.5979584455490112, | |
| "learning_rate": 0.001, | |
| "loss": 2.0047, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.28407618403434753, | |
| "learning_rate": 0.001, | |
| "loss": 1.7317, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 0.3217169940471649, | |
| "learning_rate": 0.001, | |
| "loss": 1.694, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.33735284209251404, | |
| "learning_rate": 0.001, | |
| "loss": 1.6403, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_gen_len": 511.0, | |
| "eval_loss": 1.5456656217575073, | |
| "eval_rouge1": 18.4716, | |
| "eval_rouge2": 5.4633, | |
| "eval_rougeL": 17.1393, | |
| "eval_rougeLsum": 16.9242, | |
| "eval_runtime": 1798.0277, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 0.20745837688446045, | |
| "learning_rate": 0.001, | |
| "loss": 1.5256, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.24278272688388824, | |
| "learning_rate": 0.001, | |
| "loss": 1.4077, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 3.5210845470428467, | |
| "learning_rate": 0.001, | |
| "loss": 1.4244, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 0.31759026646614075, | |
| "learning_rate": 0.001, | |
| "loss": 1.3542, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 0.2855791449546814, | |
| "learning_rate": 0.001, | |
| "loss": 1.3873, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 0.30171895027160645, | |
| "learning_rate": 0.001, | |
| "loss": 1.4693, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.28778406977653503, | |
| "learning_rate": 0.001, | |
| "loss": 1.5012, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "eval_gen_len": 511.0, | |
| "eval_loss": 1.5736442804336548, | |
| "eval_rouge1": 18.2259, | |
| "eval_rouge2": 5.3524, | |
| "eval_rougeL": 17.0162, | |
| "eval_rougeLsum": 16.7948, | |
| "eval_runtime": 1799.9735, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 4.03, | |
| "grad_norm": 0.27410924434661865, | |
| "learning_rate": 0.001, | |
| "loss": 1.3865, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 0.2398337870836258, | |
| "learning_rate": 0.001, | |
| "loss": 1.198, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 0.24380528926849365, | |
| "learning_rate": 0.001, | |
| "loss": 1.1965, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 0.28130125999450684, | |
| "learning_rate": 0.001, | |
| "loss": 1.2576, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 4.59, | |
| "grad_norm": 0.22549273073673248, | |
| "learning_rate": 0.001, | |
| "loss": 1.2108, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 0.3336837589740753, | |
| "learning_rate": 0.001, | |
| "loss": 1.23, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "grad_norm": 0.39294493198394775, | |
| "learning_rate": 0.001, | |
| "loss": 1.248, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "eval_gen_len": 511.0, | |
| "eval_loss": 1.5482468605041504, | |
| "eval_rouge1": 20.8275, | |
| "eval_rouge2": 6.7412, | |
| "eval_rougeL": 18.0859, | |
| "eval_rougeLsum": 19.3113, | |
| "eval_runtime": 1798.5715, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 5.01, | |
| "grad_norm": 0.3731052577495575, | |
| "learning_rate": 0.001, | |
| "loss": 1.2523, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 5.15, | |
| "grad_norm": 0.33552882075309753, | |
| "learning_rate": 0.001, | |
| "loss": 1.0577, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 0.3163793087005615, | |
| "learning_rate": 0.001, | |
| "loss": 1.0478, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 5.43, | |
| "grad_norm": 0.21926109492778778, | |
| "learning_rate": 0.001, | |
| "loss": 1.0127, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "grad_norm": 0.24710944294929504, | |
| "learning_rate": 0.001, | |
| "loss": 1.0042, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.2397957742214203, | |
| "learning_rate": 0.001, | |
| "loss": 1.0332, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "grad_norm": 0.21428123116493225, | |
| "learning_rate": 0.001, | |
| "loss": 1.022, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "grad_norm": 0.2227003127336502, | |
| "learning_rate": 0.001, | |
| "loss": 1.0176, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 5.98, | |
| "eval_gen_len": 510.6775147928994, | |
| "eval_loss": 1.625435709953308, | |
| "eval_rouge1": 21.1937, | |
| "eval_rouge2": 6.8813, | |
| "eval_rougeL": 18.411, | |
| "eval_rougeLsum": 19.8577, | |
| "eval_runtime": 1798.5872, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 6.12, | |
| "grad_norm": 0.2959192991256714, | |
| "learning_rate": 0.001, | |
| "loss": 0.879, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "grad_norm": 0.33006206154823303, | |
| "learning_rate": 0.001, | |
| "loss": 0.8812, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.34284549951553345, | |
| "learning_rate": 0.001, | |
| "loss": 0.8742, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "grad_norm": 0.4311819076538086, | |
| "learning_rate": 0.001, | |
| "loss": 0.8357, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 6.68, | |
| "grad_norm": 0.5699031352996826, | |
| "learning_rate": 0.001, | |
| "loss": 0.8721, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 6.82, | |
| "grad_norm": 0.39324450492858887, | |
| "learning_rate": 0.001, | |
| "loss": 0.8739, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "grad_norm": 0.3442493677139282, | |
| "learning_rate": 0.001, | |
| "loss": 0.8472, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "eval_gen_len": 479.9704142011834, | |
| "eval_loss": 1.6212307214736938, | |
| "eval_rouge1": 26.1873, | |
| "eval_rouge2": 9.1581, | |
| "eval_rougeL": 20.393, | |
| "eval_rougeLsum": 24.1393, | |
| "eval_runtime": 1802.4729, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.2600483298301697, | |
| "learning_rate": 0.001, | |
| "loss": 0.7568, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 7.23, | |
| "grad_norm": 0.28727108240127563, | |
| "learning_rate": 0.001, | |
| "loss": 0.6971, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 7.37, | |
| "grad_norm": 0.3065392076969147, | |
| "learning_rate": 0.001, | |
| "loss": 0.6918, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 7.51, | |
| "grad_norm": 0.427791029214859, | |
| "learning_rate": 0.001, | |
| "loss": 0.6902, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 7.65, | |
| "grad_norm": 0.48664093017578125, | |
| "learning_rate": 0.001, | |
| "loss": 0.7415, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 7.79, | |
| "grad_norm": 0.2857199013233185, | |
| "learning_rate": 0.001, | |
| "loss": 0.7442, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "grad_norm": 0.24586661159992218, | |
| "learning_rate": 0.001, | |
| "loss": 0.7242, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_gen_len": 506.9112426035503, | |
| "eval_loss": 1.723126769065857, | |
| "eval_rouge1": 23.5881, | |
| "eval_rouge2": 7.8961, | |
| "eval_rougeL": 18.7014, | |
| "eval_rougeLsum": 22.2999, | |
| "eval_runtime": 1807.8192, | |
| "eval_samples_per_second": 0.187, | |
| "eval_steps_per_second": 0.024, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 8.07, | |
| "grad_norm": 0.21033655107021332, | |
| "learning_rate": 0.001, | |
| "loss": 0.6797, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 8.21, | |
| "grad_norm": 0.22591687738895416, | |
| "learning_rate": 0.001, | |
| "loss": 0.5446, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 8.35, | |
| "grad_norm": 0.20658165216445923, | |
| "learning_rate": 0.001, | |
| "loss": 0.5545, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "grad_norm": 0.29855239391326904, | |
| "learning_rate": 0.001, | |
| "loss": 0.6124, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 8.63, | |
| "grad_norm": 0.3976292312145233, | |
| "learning_rate": 0.001, | |
| "loss": 0.6052, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "grad_norm": 0.27770739793777466, | |
| "learning_rate": 0.001, | |
| "loss": 0.5755, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.2741471529006958, | |
| "learning_rate": 0.001, | |
| "loss": 0.5876, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "eval_gen_len": 451.698224852071, | |
| "eval_loss": 1.9400925636291504, | |
| "eval_rouge1": 32.1851, | |
| "eval_rouge2": 12.6426, | |
| "eval_rougeL": 22.8358, | |
| "eval_rougeLsum": 30.6718, | |
| "eval_runtime": 1805.3092, | |
| "eval_samples_per_second": 0.187, | |
| "eval_steps_per_second": 0.024, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "grad_norm": 0.25768765807151794, | |
| "learning_rate": 0.001, | |
| "loss": 0.5517, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 9.18, | |
| "grad_norm": 0.203142449259758, | |
| "learning_rate": 0.001, | |
| "loss": 0.4295, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 9.32, | |
| "grad_norm": 0.29351434111595154, | |
| "learning_rate": 0.001, | |
| "loss": 0.493, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 9.46, | |
| "grad_norm": 0.23967808485031128, | |
| "learning_rate": 0.001, | |
| "loss": 0.4877, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.21488718688488007, | |
| "learning_rate": 0.001, | |
| "loss": 0.4943, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 9.74, | |
| "grad_norm": 0.20587602257728577, | |
| "learning_rate": 0.001, | |
| "loss": 0.4729, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 9.88, | |
| "grad_norm": 0.2094978392124176, | |
| "learning_rate": 0.001, | |
| "loss": 0.4756, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "eval_gen_len": 455.594674556213, | |
| "eval_loss": 1.9001177549362183, | |
| "eval_rouge1": 31.353, | |
| "eval_rouge2": 12.994, | |
| "eval_rougeL": 23.1542, | |
| "eval_rougeLsum": 29.8375, | |
| "eval_runtime": 1806.0454, | |
| "eval_samples_per_second": 0.187, | |
| "eval_steps_per_second": 0.024, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 10.02, | |
| "grad_norm": 0.2443789541721344, | |
| "learning_rate": 0.001, | |
| "loss": 0.4707, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 10.16, | |
| "grad_norm": 0.21666786074638367, | |
| "learning_rate": 0.001, | |
| "loss": 0.3612, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.20268017053604126, | |
| "learning_rate": 0.001, | |
| "loss": 0.3739, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 10.43, | |
| "grad_norm": 0.22428925335407257, | |
| "learning_rate": 0.001, | |
| "loss": 0.382, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 10.57, | |
| "grad_norm": 0.21844923496246338, | |
| "learning_rate": 0.001, | |
| "loss": 0.3623, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 10.71, | |
| "grad_norm": 0.2675388753414154, | |
| "learning_rate": 0.001, | |
| "loss": 0.3674, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 10.85, | |
| "grad_norm": 0.2905120849609375, | |
| "learning_rate": 0.001, | |
| "loss": 0.39, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 10.99, | |
| "grad_norm": 0.27420204877853394, | |
| "learning_rate": 0.001, | |
| "loss": 0.4042, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 10.99, | |
| "eval_gen_len": 497.53550295857985, | |
| "eval_loss": 2.1294684410095215, | |
| "eval_rouge1": 28.6425, | |
| "eval_rouge2": 11.8399, | |
| "eval_rougeL": 21.3847, | |
| "eval_rougeLsum": 27.0508, | |
| "eval_runtime": 1807.4153, | |
| "eval_samples_per_second": 0.187, | |
| "eval_steps_per_second": 0.024, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 11.13, | |
| "grad_norm": 0.26691916584968567, | |
| "learning_rate": 0.001, | |
| "loss": 0.3127, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 11.27, | |
| "grad_norm": 0.3042663335800171, | |
| "learning_rate": 0.001, | |
| "loss": 0.305, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 11.41, | |
| "grad_norm": 0.26255106925964355, | |
| "learning_rate": 0.001, | |
| "loss": 0.3133, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 11.55, | |
| "grad_norm": 0.23816817998886108, | |
| "learning_rate": 0.001, | |
| "loss": 0.3118, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 11.69, | |
| "grad_norm": 0.22553777694702148, | |
| "learning_rate": 0.001, | |
| "loss": 0.3073, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 11.83, | |
| "grad_norm": 0.2234884351491928, | |
| "learning_rate": 0.001, | |
| "loss": 0.3346, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 11.97, | |
| "grad_norm": 0.18143154680728912, | |
| "learning_rate": 0.001, | |
| "loss": 0.3292, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 11.97, | |
| "eval_gen_len": 478.81065088757396, | |
| "eval_loss": 2.244086503982544, | |
| "eval_rouge1": 31.8393, | |
| "eval_rouge2": 13.1308, | |
| "eval_rougeL": 22.135, | |
| "eval_rougeLsum": 30.5866, | |
| "eval_runtime": 1798.3958, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.24745677411556244, | |
| "learning_rate": 0.001, | |
| "loss": 0.2539, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 12.24, | |
| "grad_norm": 0.26513755321502686, | |
| "learning_rate": 0.001, | |
| "loss": 0.2588, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 12.38, | |
| "grad_norm": 0.20156317949295044, | |
| "learning_rate": 0.001, | |
| "loss": 0.2537, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 12.52, | |
| "grad_norm": 0.21362556517124176, | |
| "learning_rate": 0.001, | |
| "loss": 0.2812, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "grad_norm": 0.5383086800575256, | |
| "learning_rate": 0.001, | |
| "loss": 0.2594, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.2891131639480591, | |
| "learning_rate": 0.001, | |
| "loss": 0.2629, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 12.94, | |
| "grad_norm": 0.265836238861084, | |
| "learning_rate": 0.001, | |
| "loss": 0.2812, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 12.94, | |
| "eval_gen_len": 429.99112426035504, | |
| "eval_loss": 2.3464245796203613, | |
| "eval_rouge1": 34.4102, | |
| "eval_rouge2": 14.3607, | |
| "eval_rougeL": 23.8634, | |
| "eval_rougeLsum": 32.9732, | |
| "eval_runtime": 1798.2194, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 13.08, | |
| "grad_norm": 0.2541401982307434, | |
| "learning_rate": 0.001, | |
| "loss": 0.2283, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "grad_norm": 9.848714828491211, | |
| "learning_rate": 0.001, | |
| "loss": 0.206, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 13.36, | |
| "grad_norm": 0.4088878333568573, | |
| "learning_rate": 0.001, | |
| "loss": 0.2014, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 13.5, | |
| "grad_norm": 0.4533099830150604, | |
| "learning_rate": 0.001, | |
| "loss": 0.2292, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 13.63, | |
| "grad_norm": 0.28066885471343994, | |
| "learning_rate": 0.001, | |
| "loss": 0.2202, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 13.77, | |
| "grad_norm": 0.38810494542121887, | |
| "learning_rate": 0.001, | |
| "loss": 0.2278, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 13.91, | |
| "grad_norm": 0.2568497657775879, | |
| "learning_rate": 0.001, | |
| "loss": 0.2443, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 13.98, | |
| "eval_gen_len": 392.53846153846155, | |
| "eval_loss": 2.2002713680267334, | |
| "eval_rouge1": 34.8239, | |
| "eval_rouge2": 14.8042, | |
| "eval_rougeL": 25.2438, | |
| "eval_rougeLsum": 33.0469, | |
| "eval_runtime": 1797.5392, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 14.05, | |
| "grad_norm": 0.279291570186615, | |
| "learning_rate": 0.001, | |
| "loss": 0.2362, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 14.19, | |
| "grad_norm": 0.18151430785655975, | |
| "learning_rate": 0.001, | |
| "loss": 0.1807, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 14.33, | |
| "grad_norm": 0.2227843850851059, | |
| "learning_rate": 0.001, | |
| "loss": 0.1708, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 14.47, | |
| "grad_norm": 0.2937067151069641, | |
| "learning_rate": 0.001, | |
| "loss": 0.1818, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 14.61, | |
| "grad_norm": 0.3238927125930786, | |
| "learning_rate": 0.001, | |
| "loss": 0.1958, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 14.61, | |
| "eval_gen_len": 503.5769230769231, | |
| "eval_loss": 2.5840089321136475, | |
| "eval_rouge1": 29.7482, | |
| "eval_rouge2": 12.0072, | |
| "eval_rougeL": 21.348, | |
| "eval_rougeLsum": 28.5849, | |
| "eval_runtime": 1799.2535, | |
| "eval_samples_per_second": 0.188, | |
| "eval_steps_per_second": 0.024, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 14.61, | |
| "step": 210, | |
| "total_flos": 3.6715210940733604e+18, | |
| "train_loss": 0.9784720075981957, | |
| "train_runtime": 78497.3694, | |
| "train_samples_per_second": 0.702, | |
| "train_steps_per_second": 0.003 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 210, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "total_flos": 3.6715210940733604e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |