| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 4390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002277904328018223, | |
| "grad_norm": 386.0, | |
| "learning_rate": 4.5558086560364467e-07, | |
| "loss": 47.0249, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011389521640091117, | |
| "grad_norm": 430.0, | |
| "learning_rate": 2.2779043280182233e-06, | |
| "loss": 48.5743, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.022779043280182234, | |
| "grad_norm": 324.0, | |
| "learning_rate": 4.555808656036447e-06, | |
| "loss": 47.0262, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03416856492027335, | |
| "grad_norm": 178.0, | |
| "learning_rate": 6.83371298405467e-06, | |
| "loss": 40.673, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04555808656036447, | |
| "grad_norm": 125.5, | |
| "learning_rate": 9.111617312072893e-06, | |
| "loss": 35.8229, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05694760820045558, | |
| "grad_norm": 76.5, | |
| "learning_rate": 1.1389521640091117e-05, | |
| "loss": 29.5881, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0683371298405467, | |
| "grad_norm": 24.75, | |
| "learning_rate": 1.366742596810934e-05, | |
| "loss": 26.1961, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07972665148063782, | |
| "grad_norm": 20.25, | |
| "learning_rate": 1.5945330296127563e-05, | |
| "loss": 24.1324, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09111617312072894, | |
| "grad_norm": 14.75, | |
| "learning_rate": 1.8223234624145787e-05, | |
| "loss": 22.6848, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10250569476082004, | |
| "grad_norm": 9.375, | |
| "learning_rate": 2.050113895216401e-05, | |
| "loss": 21.496, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.11389521640091116, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 2.2779043280182233e-05, | |
| "loss": 20.4174, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1252847380410023, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 2.505694760820046e-05, | |
| "loss": 19.6892, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1366742596810934, | |
| "grad_norm": 4.25, | |
| "learning_rate": 2.733485193621868e-05, | |
| "loss": 18.9022, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1480637813211845, | |
| "grad_norm": 4.71875, | |
| "learning_rate": 2.96127562642369e-05, | |
| "loss": 18.4344, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15945330296127563, | |
| "grad_norm": 6.53125, | |
| "learning_rate": 3.189066059225513e-05, | |
| "loss": 18.1733, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17084282460136674, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 3.416856492027335e-05, | |
| "loss": 17.6679, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.18223234624145787, | |
| "grad_norm": 9.875, | |
| "learning_rate": 3.6446469248291574e-05, | |
| "loss": 16.8998, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19362186788154898, | |
| "grad_norm": 14.125, | |
| "learning_rate": 3.87243735763098e-05, | |
| "loss": 15.9599, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.20501138952164008, | |
| "grad_norm": 23.625, | |
| "learning_rate": 4.100227790432802e-05, | |
| "loss": 14.2835, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2164009111617312, | |
| "grad_norm": 29.875, | |
| "learning_rate": 4.3280182232346244e-05, | |
| "loss": 11.33, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.22779043280182232, | |
| "grad_norm": 29.0, | |
| "learning_rate": 4.555808656036447e-05, | |
| "loss": 7.3568, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23917995444191345, | |
| "grad_norm": 15.375, | |
| "learning_rate": 4.783599088838269e-05, | |
| "loss": 3.988, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2505694760820046, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 5.011389521640092e-05, | |
| "loss": 2.5546, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2619589977220957, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 5.239179954441914e-05, | |
| "loss": 2.0898, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2733485193621868, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 5.466970387243736e-05, | |
| "loss": 1.8634, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2847380410022779, | |
| "grad_norm": 1.375, | |
| "learning_rate": 5.6947608200455584e-05, | |
| "loss": 1.7283, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.296127562642369, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 5.92255125284738e-05, | |
| "loss": 1.652, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.30751708428246016, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 6.150341685649203e-05, | |
| "loss": 1.5754, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.31890660592255127, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 6.378132118451025e-05, | |
| "loss": 1.5234, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.33029612756264237, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 6.605922551252848e-05, | |
| "loss": 1.461, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3416856492027335, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 6.83371298405467e-05, | |
| "loss": 1.4367, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3530751708428246, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 7.061503416856492e-05, | |
| "loss": 1.4151, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.36446469248291574, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 7.289293849658315e-05, | |
| "loss": 1.3879, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.37585421412300685, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 7.517084282460137e-05, | |
| "loss": 1.3668, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.38724373576309795, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 7.74487471526196e-05, | |
| "loss": 1.3422, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.39863325740318906, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.972665148063782e-05, | |
| "loss": 1.3269, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.41002277904328016, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 8.200455580865604e-05, | |
| "loss": 1.3143, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4214123006833713, | |
| "grad_norm": 3.1875, | |
| "learning_rate": 8.428246013667426e-05, | |
| "loss": 1.2994, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4328018223234624, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 8.656036446469249e-05, | |
| "loss": 1.2815, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.44419134396355353, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 8.883826879271071e-05, | |
| "loss": 1.2721, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.45558086560364464, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 9.111617312072893e-05, | |
| "loss": 1.2607, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.46697038724373574, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 9.339407744874716e-05, | |
| "loss": 1.2512, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4783599088838269, | |
| "grad_norm": 3.0, | |
| "learning_rate": 9.567198177676538e-05, | |
| "loss": 1.2626, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.489749430523918, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 9.79498861047836e-05, | |
| "loss": 1.2506, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5011389521640092, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00010022779043280184, | |
| "loss": 1.2473, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5125284738041003, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 0.00010250569476082006, | |
| "loss": 1.2319, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5239179954441914, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 0.00010478359908883827, | |
| "loss": 1.2246, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5353075170842825, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 0.0001070615034168565, | |
| "loss": 1.2092, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5466970387243736, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.00010933940774487472, | |
| "loss": 1.1988, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5580865603644647, | |
| "grad_norm": 2.875, | |
| "learning_rate": 0.00011161731207289294, | |
| "loss": 1.1988, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.00011389521640091117, | |
| "loss": 1.1985, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5808656036446469, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.00011617312072892939, | |
| "loss": 1.197, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.592255125284738, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 0.0001184510250569476, | |
| "loss": 1.1889, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6036446469248291, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.00012072892938496582, | |
| "loss": 1.1738, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6150341685649203, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00012300683371298406, | |
| "loss": 1.1693, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6264236902050114, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 0.00012528473804100228, | |
| "loss": 1.1681, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6378132118451025, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 0.0001275626423690205, | |
| "loss": 1.1611, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6492027334851936, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 0.00012984054669703873, | |
| "loss": 1.1557, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6605922551252847, | |
| "grad_norm": 4.84375, | |
| "learning_rate": 0.00013211845102505695, | |
| "loss": 1.151, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6719817767653758, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 0.00013439635535307518, | |
| "loss": 1.1527, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.683371298405467, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 0.0001366742596810934, | |
| "loss": 1.143, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6947608200455581, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 0.00013895216400911162, | |
| "loss": 1.1439, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7061503416856492, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 0.00014123006833712985, | |
| "loss": 1.1345, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7175398633257403, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 0.00014350797266514807, | |
| "loss": 1.125, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7289293849658315, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 0.0001457858769931663, | |
| "loss": 1.1364, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7403189066059226, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 0.00014806378132118452, | |
| "loss": 1.1339, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7517084282460137, | |
| "grad_norm": 21.25, | |
| "learning_rate": 0.00015034168564920274, | |
| "loss": 1.1417, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7630979498861048, | |
| "grad_norm": 4.375, | |
| "learning_rate": 0.00015261958997722096, | |
| "loss": 1.1475, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7744874715261959, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 0.0001548974943052392, | |
| "loss": 1.1267, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.785876993166287, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.0001571753986332574, | |
| "loss": 1.1184, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7972665148063781, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 0.00015945330296127563, | |
| "loss": 1.123, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.8086560364464692, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.00016173120728929386, | |
| "loss": 1.1052, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8200455580865603, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 0.00016400911161731208, | |
| "loss": 1.1108, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8314350797266514, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 0.0001662870159453303, | |
| "loss": 1.1204, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8428246013667426, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 0.00016856492027334853, | |
| "loss": 1.1326, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8542141230068337, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 0.00017084282460136675, | |
| "loss": 1.1137, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8656036446469249, | |
| "grad_norm": 3.75, | |
| "learning_rate": 0.00017312072892938497, | |
| "loss": 1.1272, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.876993166287016, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 0.0001753986332574032, | |
| "loss": 1.108, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8883826879271071, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 0.00017767653758542142, | |
| "loss": 1.1221, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8997722095671982, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 0.00017995444191343964, | |
| "loss": 1.0942, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "grad_norm": 0.79296875, | |
| "learning_rate": 0.00018223234624145787, | |
| "loss": 1.0881, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.9225512528473804, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 0.0001845102505694761, | |
| "loss": 1.0857, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9339407744874715, | |
| "grad_norm": 1.875, | |
| "learning_rate": 0.00018678815489749431, | |
| "loss": 1.0843, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9453302961275627, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.00018906605922551254, | |
| "loss": 1.0782, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9567198177676538, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 0.00019134396355353076, | |
| "loss": 1.0762, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9681093394077449, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 0.00019362186788154898, | |
| "loss": 1.0681, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.979498861047836, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.0001958997722095672, | |
| "loss": 1.0742, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9908883826879271, | |
| "grad_norm": 18.625, | |
| "learning_rate": 0.00019817767653758543, | |
| "loss": 1.0805, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.521019220352173, | |
| "eval_runtime": 0.2677, | |
| "eval_samples_per_second": 37.357, | |
| "eval_steps_per_second": 3.736, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.0022779043280183, | |
| "grad_norm": 6.6875, | |
| "learning_rate": 0.0001999999683877311, | |
| "loss": 1.1284, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0136674259681093, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 0.0001999988619604182, | |
| "loss": 1.0864, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.0250569476082005, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 0.00019999617493964692, | |
| "loss": 1.0719, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.0364464692482915, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 0.00019999190736788865, | |
| "loss": 1.1239, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.0478359908883828, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 0.0001999860593125971, | |
| "loss": 1.1026, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.0592255125284737, | |
| "grad_norm": 1.75, | |
| "learning_rate": 0.00019997863086620727, | |
| "loss": 1.0776, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.070615034168565, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 0.0001999696221461341, | |
| "loss": 1.072, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.082004555808656, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 0.0001999590332947704, | |
| "loss": 1.073, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.0933940774487472, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 0.0001999468644794848, | |
| "loss": 1.0755, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1047835990888384, | |
| "grad_norm": 14.125, | |
| "learning_rate": 0.00019993311589261897, | |
| "loss": 1.0603, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.1161731207289294, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 0.00019991778775148465, | |
| "loss": 1.0802, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.1275626423690206, | |
| "grad_norm": 3.25, | |
| "learning_rate": 0.00019990088029836017, | |
| "loss": 1.0647, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 3.9375, | |
| "learning_rate": 0.00019988239380048674, | |
| "loss": 1.062, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.1503416856492028, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 0.000199862328550064, | |
| "loss": 1.0499, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.1617312072892938, | |
| "grad_norm": 1.875, | |
| "learning_rate": 0.00019984068486424557, | |
| "loss": 1.0475, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.173120728929385, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 0.0001998174630851341, | |
| "loss": 1.0381, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.184510250569476, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 0.00019979266357977564, | |
| "loss": 1.0617, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.1958997722095672, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.000199766286740154, | |
| "loss": 1.051, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.2072892938496582, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 0.0001997383329831846, | |
| "loss": 1.0657, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.2186788154897494, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 0.00019970880275070762, | |
| "loss": 1.0574, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.2300683371298406, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.00019967769650948135, | |
| "loss": 1.0469, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.2414578587699316, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 0.00019964501475117462, | |
| "loss": 1.0483, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.2528473804100229, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 0.00019961075799235903, | |
| "loss": 1.0401, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.2642369020501139, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.0001995749267745008, | |
| "loss": 1.0436, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.275626423690205, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 0.00019953752166395228, | |
| "loss": 1.0291, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.287015945330296, | |
| "grad_norm": 2.625, | |
| "learning_rate": 0.00019949854325194294, | |
| "loss": 1.0726, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.2984054669703873, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 0.00019945799215456998, | |
| "loss": 1.0269, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.3097949886104785, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 0.00019941586901278875, | |
| "loss": 1.0222, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.3211845102505695, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 0.0001993721744924024, | |
| "loss": 1.0212, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.3325740318906605, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 0.00019932690928405153, | |
| "loss": 1.0125, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.3439635535307517, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 0.00019928007410320323, | |
| "loss": 1.0043, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.355353075170843, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.0001992316696901397, | |
| "loss": 1.0229, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.366742596810934, | |
| "grad_norm": 4.625, | |
| "learning_rate": 0.00019918169680994667, | |
| "loss": 1.0179, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.3781321184510251, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 0.00019913015625250114, | |
| "loss": 1.0123, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.3895216400911161, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 0.00019907704883245916, | |
| "loss": 1.0104, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.4009111617312073, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.00019902237538924256, | |
| "loss": 1.0195, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.4123006833712983, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.00019896613678702617, | |
| "loss": 1.0101, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.4236902050113895, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 0.0001989083339147237, | |
| "loss": 1.0192, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.4350797266514808, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 0.000198848967685974, | |
| "loss": 1.0109, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.4464692482915718, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 0.0001987880390391264, | |
| "loss": 1.0048, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.4578587699316627, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.00019872554893722618, | |
| "loss": 0.9957, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.469248291571754, | |
| "grad_norm": 2.875, | |
| "learning_rate": 0.00019866149836799896, | |
| "loss": 1.0112, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.4806378132118452, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 0.0001985958883438354, | |
| "loss": 1.0237, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.4920273348519362, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 0.00019852871990177503, | |
| "loss": 1.0246, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.5034168564920274, | |
| "grad_norm": 11.625, | |
| "learning_rate": 0.00019845999410349002, | |
| "loss": 1.0251, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.5148063781321186, | |
| "grad_norm": 2.5, | |
| "learning_rate": 0.00019838971203526808, | |
| "loss": 1.0168, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.5261958997722096, | |
| "grad_norm": 13.4375, | |
| "learning_rate": 0.00019831787480799568, | |
| "loss": 1.0081, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.5375854214123006, | |
| "grad_norm": 10.25, | |
| "learning_rate": 0.0001982444835571403, | |
| "loss": 1.0271, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.5489749430523918, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 0.00019816953944273237, | |
| "loss": 1.0059, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.560364464692483, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 0.0001980930436493472, | |
| "loss": 1.0287, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.571753986332574, | |
| "grad_norm": 4.875, | |
| "learning_rate": 0.00019801499738608604, | |
| "loss": 1.0337, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.583143507972665, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 0.00019793540188655704, | |
| "loss": 1.0124, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.5945330296127562, | |
| "grad_norm": 10.375, | |
| "learning_rate": 0.0001978542584088558, | |
| "loss": 1.0009, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.6059225512528474, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 0.00019777156823554544, | |
| "loss": 0.9936, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.6173120728929384, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 0.00019768733267363624, | |
| "loss": 1.0044, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.6287015945330297, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.0001976015530545652, | |
| "loss": 0.9959, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.6400911161731209, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 0.00019751423073417475, | |
| "loss": 0.9893, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.6514806378132119, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 0.0001974253670926915, | |
| "loss": 0.9896, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.6628701594533029, | |
| "grad_norm": 2.25, | |
| "learning_rate": 0.00019733496353470433, | |
| "loss": 0.9965, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.674259681093394, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 0.00019724302148914222, | |
| "loss": 0.9817, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.6856492027334853, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 0.00019714954240925172, | |
| "loss": 0.9811, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.6970387243735763, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 0.00019705452777257377, | |
| "loss": 0.9798, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "grad_norm": 2.125, | |
| "learning_rate": 0.0001969579790809207, | |
| "loss": 0.994, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.7198177676537585, | |
| "grad_norm": 2.625, | |
| "learning_rate": 0.00019685989786035211, | |
| "loss": 0.9838, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.7312072892938497, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 0.00019676028566115102, | |
| "loss": 0.9868, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.7425968109339407, | |
| "grad_norm": 0.890625, | |
| "learning_rate": 0.00019665914405779923, | |
| "loss": 0.9933, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.753986332574032, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.00019655647464895254, | |
| "loss": 0.9949, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.7653758542141231, | |
| "grad_norm": 1.5, | |
| "learning_rate": 0.00019645227905741534, | |
| "loss": 0.9871, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.7767653758542141, | |
| "grad_norm": 3.03125, | |
| "learning_rate": 0.00019634655893011513, | |
| "loss": 0.9855, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.7881548974943051, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 0.0001962393159380763, | |
| "loss": 1.0098, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.7995444191343963, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 0.00019613055177639384, | |
| "loss": 0.9833, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.8109339407744875, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 0.0001960202681642066, | |
| "loss": 0.9712, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 0.00019590846684466992, | |
| "loss": 0.9792, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.8337129840546698, | |
| "grad_norm": 12.25, | |
| "learning_rate": 0.00019579514958492826, | |
| "loss": 0.978, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.845102505694761, | |
| "grad_norm": 18.125, | |
| "learning_rate": 0.00019568031817608725, | |
| "loss": 0.9891, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.856492027334852, | |
| "grad_norm": 5.34375, | |
| "learning_rate": 0.00019556397443318523, | |
| "loss": 0.9708, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.867881548974943, | |
| "grad_norm": 5.46875, | |
| "learning_rate": 0.00019544612019516472, | |
| "loss": 0.9601, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.8792710706150342, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 0.00019532675732484333, | |
| "loss": 1.0184, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.8906605922551254, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.00019520588770888424, | |
| "loss": 0.9742, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.9020501138952164, | |
| "grad_norm": 9.75, | |
| "learning_rate": 0.00019508351325776642, | |
| "loss": 0.9652, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.9134396355353074, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 0.00019495963590575443, | |
| "loss": 1.003, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.9248291571753986, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 0.00019483425761086793, | |
| "loss": 0.9781, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.9362186788154898, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 0.00019470738035485058, | |
| "loss": 0.9943, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.9476082004555808, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 0.0001945790061431388, | |
| "loss": 0.9864, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.958997722095672, | |
| "grad_norm": 1.9609375, | |
| "learning_rate": 0.00019444913700483008, | |
| "loss": 0.9491, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.9703872437357632, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 0.00019431777499265087, | |
| "loss": 0.9748, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.9817767653758542, | |
| "grad_norm": 8.0, | |
| "learning_rate": 0.0001941849221829242, | |
| "loss": 0.939, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.9931662870159452, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 0.00019405058067553676, | |
| "loss": 0.9397, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.4361355304718018, | |
| "eval_runtime": 0.2364, | |
| "eval_samples_per_second": 42.306, | |
| "eval_steps_per_second": 4.231, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 2.0045558086560367, | |
| "grad_norm": 12.625, | |
| "learning_rate": 0.00019391475259390584, | |
| "loss": 0.9262, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.0159453302961277, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00019377744008494555, | |
| "loss": 0.922, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.0273348519362187, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00019363864531903323, | |
| "loss": 0.9265, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.0387243735763096, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 0.00019349837048997478, | |
| "loss": 0.9572, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.050113895216401, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 0.00019335661781497024, | |
| "loss": 0.9381, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.061503416856492, | |
| "grad_norm": 3.359375, | |
| "learning_rate": 0.00019321338953457858, | |
| "loss": 0.9678, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 2.072892938496583, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 0.0001930686879126824, | |
| "loss": 0.9557, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.084282460136674, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 0.00019292251523645208, | |
| "loss": 0.9384, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 2.0956719817767655, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 0.00019277487381630975, | |
| "loss": 0.9409, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.1070615034168565, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.0001926257659858925, | |
| "loss": 0.9319, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.1184510250569475, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 0.00019247519410201585, | |
| "loss": 0.9264, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.129840546697039, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 0.00019232316054463617, | |
| "loss": 0.9069, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.14123006833713, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 0.0001921696677168133, | |
| "loss": 0.9139, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.152619589977221, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 0.00019201471804467245, | |
| "loss": 0.9152, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.164009111617312, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00019185831397736583, | |
| "loss": 0.9038, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.1753986332574033, | |
| "grad_norm": 0.796875, | |
| "learning_rate": 0.00019170045798703406, | |
| "loss": 0.9073, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.1867881548974943, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00019154115256876702, | |
| "loss": 0.9026, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.1981776765375853, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 0.00019138040024056435, | |
| "loss": 0.8967, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.2095671981776768, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00019121820354329577, | |
| "loss": 0.8996, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.2209567198177678, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00019105456504066082, | |
| "loss": 0.9049, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.2323462414578588, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.0001908894873191484, | |
| "loss": 0.913, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.2437357630979498, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00019072297298799589, | |
| "loss": 0.9099, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.255125284738041, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00019055502467914788, | |
| "loss": 0.9016, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.266514806378132, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 0.00019038564504721454, | |
| "loss": 0.89, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00019021483676942973, | |
| "loss": 0.9021, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.289293849658314, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.00019004260254560867, | |
| "loss": 0.9142, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.3006833712984056, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.00018986894509810513, | |
| "loss": 0.9014, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.3120728929384966, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 0.0001896938671717687, | |
| "loss": 0.8982, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.3234624145785876, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00018951737153390105, | |
| "loss": 0.9056, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.334851936218679, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 0.00018933946097421248, | |
| "loss": 0.8869, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.34624145785877, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00018916013830477766, | |
| "loss": 0.8821, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.357630979498861, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00018897940635999118, | |
| "loss": 0.8865, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.369020501138952, | |
| "grad_norm": 0.9765625, | |
| "learning_rate": 0.0001887972679965229, | |
| "loss": 0.8943, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.3804100227790435, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 0.00018861372609327263, | |
| "loss": 0.8912, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.3917995444191344, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.00018842878355132471, | |
| "loss": 0.8847, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.4031890660592254, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 0.0001882424432939021, | |
| "loss": 0.8833, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.4145785876993164, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.00018805470826632024, | |
| "loss": 0.8905, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.425968109339408, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00018786558143594047, | |
| "loss": 0.8828, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.437357630979499, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.00018767506579212313, | |
| "loss": 0.882, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.44874715261959, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.0001874831643461803, | |
| "loss": 0.8863, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.4601366742596813, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00018728988013132819, | |
| "loss": 0.8787, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.4715261958997723, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.0001870952162026392, | |
| "loss": 0.8755, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.4829157175398633, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.0001868991756369937, | |
| "loss": 0.8839, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.4943052391799543, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00018670176153303127, | |
| "loss": 0.8818, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.5056947608200457, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.0001865029770111019, | |
| "loss": 0.8875, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.5170842824601367, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00018630282521321645, | |
| "loss": 0.8769, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.5284738041002277, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.00018610130930299715, | |
| "loss": 0.8812, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.5398633257403187, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.00018589843246562756, | |
| "loss": 0.8837, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.55125284738041, | |
| "grad_norm": 0.80078125, | |
| "learning_rate": 0.00018569419790780218, | |
| "loss": 0.8769, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.562642369020501, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00018548860885767582, | |
| "loss": 0.8782, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.574031890660592, | |
| "grad_norm": 7.65625, | |
| "learning_rate": 0.00018528166856481254, | |
| "loss": 0.8822, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.5854214123006836, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 0.00018507338030013427, | |
| "loss": 0.8677, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.5968109339407746, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.0001848637473558692, | |
| "loss": 0.8775, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.6082004555808656, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.00018465277304549962, | |
| "loss": 0.8699, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.619589977220957, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 0.00018444046070370963, | |
| "loss": 0.8709, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.630979498861048, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00018422681368633238, | |
| "loss": 0.8821, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.642369020501139, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 0.00018401183537029714, | |
| "loss": 0.8807, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.65375854214123, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00018379552915357575, | |
| "loss": 0.8786, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.665148063781321, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00018357789845512901, | |
| "loss": 0.8744, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.6765375854214124, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 0.0001833589467148527, | |
| "loss": 0.8649, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.6879271070615034, | |
| "grad_norm": 0.76953125, | |
| "learning_rate": 0.00018313867739352304, | |
| "loss": 0.8716, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.6993166287015944, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00018291709397274218, | |
| "loss": 0.8658, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.710706150341686, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00018269419995488298, | |
| "loss": 0.874, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.722095671981777, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 0.00018246999886303383, | |
| "loss": 0.8752, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.733485193621868, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00018224449424094288, | |
| "loss": 0.8665, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.7448747152619593, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 0.00018201768965296194, | |
| "loss": 0.866, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.7562642369020502, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.00018178958868399033, | |
| "loss": 0.8602, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.7676537585421412, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00018156019493941803, | |
| "loss": 0.8618, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.7790432801822322, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00018132951204506887, | |
| "loss": 0.8658, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.7904328018223232, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00018109754364714305, | |
| "loss": 0.8646, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.8018223234624147, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.0001808642934121597, | |
| "loss": 0.8627, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.8132118451025057, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00018062976502689862, | |
| "loss": 0.8639, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.8246013667425967, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00018039396219834237, | |
| "loss": 0.8592, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.835990888382688, | |
| "grad_norm": 0.671875, | |
| "learning_rate": 0.0001801568886536174, | |
| "loss": 0.8575, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 2.847380410022779, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.0001799185481399354, | |
| "loss": 0.8581, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.85876993166287, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 0.0001796789444245337, | |
| "loss": 0.8682, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 2.8701594533029615, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.0001794380812946161, | |
| "loss": 0.8758, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.8815489749430525, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 0.00017919596255729285, | |
| "loss": 0.8691, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 2.8929384965831435, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 0.00017895259203952032, | |
| "loss": 0.8629, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.9043280182232345, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00017870797358804084, | |
| "loss": 0.8665, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.9157175398633255, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 0.00017846211106932165, | |
| "loss": 0.8631, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.927107061503417, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00017821500836949386, | |
| "loss": 0.8555, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 2.938496583143508, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.000177966669394291, | |
| "loss": 0.8555, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.949886104783599, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00017771709806898732, | |
| "loss": 0.8577, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 2.9612756264236904, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00017746629833833585, | |
| "loss": 0.8506, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.9726651480637813, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00017721427416650577, | |
| "loss": 0.8503, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 2.9840546697038723, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 0.00017696102953702, | |
| "loss": 0.8513, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.995444191343964, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00017670656845269214, | |
| "loss": 0.8628, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.405621290206909, | |
| "eval_runtime": 0.2428, | |
| "eval_samples_per_second": 41.192, | |
| "eval_steps_per_second": 4.119, | |
| "step": 1317 | |
| }, | |
| { | |
| "epoch": 3.0068337129840548, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.00017645089493556322, | |
| "loss": 0.8368, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.0182232346241458, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 0.0001761940130268381, | |
| "loss": 0.8331, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 3.0296127562642368, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00017593592678682166, | |
| "loss": 0.8446, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.041002277904328, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 0.0001756766402948545, | |
| "loss": 0.8303, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 3.052391799544419, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 0.00017541615764924868, | |
| "loss": 0.8381, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.06378132118451, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.00017515448296722262, | |
| "loss": 0.8353, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 3.075170842824601, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 0.00017489162038483637, | |
| "loss": 0.836, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.0865603644646926, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.00017462757405692597, | |
| "loss": 0.8187, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 3.0979498861047836, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00017436234815703788, | |
| "loss": 0.8301, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.1093394077448746, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.000174095946877363, | |
| "loss": 0.8356, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 3.120728929384966, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00017382837442867055, | |
| "loss": 0.824, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.132118451025057, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.00017355963504024123, | |
| "loss": 0.8378, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 3.143507972665148, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 0.00017328973295980052, | |
| "loss": 0.8334, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.154897494305239, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00017301867245345172, | |
| "loss": 0.8412, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 3.1662870159453305, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.0001727464578056081, | |
| "loss": 0.8302, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.1776765375854215, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 0.0001724730933189256, | |
| "loss": 0.8294, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 3.1890660592255125, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 0.0001721985833142346, | |
| "loss": 0.8354, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.2004555808656034, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 0.0001719229321304716, | |
| "loss": 0.8379, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 3.211845102505695, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00017164614412461084, | |
| "loss": 0.8245, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.223234624145786, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00017136822367159516, | |
| "loss": 0.8202, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 3.234624145785877, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00017108917516426704, | |
| "loss": 0.822, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.2460136674259683, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.0001708090030132992, | |
| "loss": 0.8299, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.2574031890660593, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00017052771164712465, | |
| "loss": 0.8365, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.2687927107061503, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 0.00017024530551186702, | |
| "loss": 0.8254, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 3.2801822323462413, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 0.0001699617890712699, | |
| "loss": 0.8257, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.2915717539863327, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00016967716680662667, | |
| "loss": 0.8315, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 3.3029612756264237, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.0001693914432167094, | |
| "loss": 0.8265, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.3143507972665147, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00016910462281769783, | |
| "loss": 0.8228, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 3.3257403189066057, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.0001688167101431081, | |
| "loss": 0.8262, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.337129840546697, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 0.0001685277097437208, | |
| "loss": 0.8284, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 3.348519362186788, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.00016823762618750938, | |
| "loss": 0.8327, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.359908883826879, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.00016794646405956774, | |
| "loss": 0.8211, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.3712984054669706, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.0001676542279620378, | |
| "loss": 0.8266, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.3826879271070616, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00016736092251403673, | |
| "loss": 0.8247, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 3.3940774487471526, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00016706655235158407, | |
| "loss": 0.8243, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.4054669703872436, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 0.00016677112212752824, | |
| "loss": 0.8186, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 3.416856492027335, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 0.0001664746365114732, | |
| "loss": 0.8274, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.428246013667426, | |
| "grad_norm": 0.88671875, | |
| "learning_rate": 0.00016617710018970453, | |
| "loss": 0.8175, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 3.439635535307517, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00016587851786511543, | |
| "loss": 0.8212, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.451025056947608, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00016557889425713226, | |
| "loss": 0.8185, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 3.4624145785876994, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.0001652782341016401, | |
| "loss": 0.8175, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.4738041002277904, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00016497654215090772, | |
| "loss": 0.8192, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.4851936218678814, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00016467382317351267, | |
| "loss": 0.8139, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.496583143507973, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00016437008195426578, | |
| "loss": 0.8217, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 3.507972665148064, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.00016406532329413546, | |
| "loss": 0.8182, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.519362186788155, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 0.000163759552010172, | |
| "loss": 0.8187, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 3.5307517084282463, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 0.00016345277293543136, | |
| "loss": 0.8114, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.5421412300683373, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.0001631449909188987, | |
| "loss": 0.8195, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 3.5535307517084282, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00016283621082541173, | |
| "loss": 0.8122, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.5649202733485192, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.000162526437535584, | |
| "loss": 0.8166, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 3.5763097949886102, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00016221567594572762, | |
| "loss": 0.8194, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.5876993166287017, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.0001619039309677758, | |
| "loss": 0.82, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.5990888382687927, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.0001615912075292054, | |
| "loss": 0.8184, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.6104783599088837, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.0001612775105729588, | |
| "loss": 0.8138, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 3.621867881548975, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.0001609628450573661, | |
| "loss": 0.8121, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.633257403189066, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00016064721595606635, | |
| "loss": 0.8157, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 3.644646924829157, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00016033062825792935, | |
| "loss": 0.8063, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.6560364464692485, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00016001308696697643, | |
| "loss": 0.8207, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 3.6674259681093395, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00015969459710230162, | |
| "loss": 0.8067, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.6788154897494305, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00015937516369799216, | |
| "loss": 0.8105, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 3.6902050113895215, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.00015905479180304896, | |
| "loss": 0.8212, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.7015945330296125, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 0.00015873348648130694, | |
| "loss": 0.8089, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 3.712984054669704, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00015841125281135473, | |
| "loss": 0.8173, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.724373576309795, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.00015808809588645467, | |
| "loss": 0.8156, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 3.735763097949886, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 0.00015776402081446204, | |
| "loss": 0.8083, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.7471526195899774, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00015743903271774455, | |
| "loss": 0.8165, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 3.7585421412300684, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00015711313673310125, | |
| "loss": 0.8237, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.7699316628701594, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00015678633801168137, | |
| "loss": 0.8177, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 3.781321184510251, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00015645864171890295, | |
| "loss": 0.8086, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.792710706150342, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00015613005303437104, | |
| "loss": 0.8118, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 3.8041002277904328, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00015580057715179605, | |
| "loss": 0.8208, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 3.8154897494305238, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00015547021927891144, | |
| "loss": 0.8076, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 3.8268792710706148, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.0001551389846373916, | |
| "loss": 0.8137, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.838268792710706, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 0.00015480687846276917, | |
| "loss": 0.8117, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 3.849658314350797, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.00015447390600435238, | |
| "loss": 0.8035, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.861047835990888, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 0.00015414007252514202, | |
| "loss": 0.8105, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 3.8724373576309796, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 0.00015380538330174827, | |
| "loss": 0.8043, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.8838268792710706, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.0001534698436243073, | |
| "loss": 0.8124, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 3.8952164009111616, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00015313345879639764, | |
| "loss": 0.8198, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.906605922551253, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00015279623413495642, | |
| "loss": 0.8057, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 3.917995444191344, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00015245817497019524, | |
| "loss": 0.806, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.929384965831435, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00015211928664551593, | |
| "loss": 0.8033, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 3.940774487471526, | |
| "grad_norm": 0.62890625, | |
| "learning_rate": 0.00015177957451742612, | |
| "loss": 0.8137, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.9521640091116175, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00015143904395545466, | |
| "loss": 0.8075, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 3.9635535307517085, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.0001510977003420665, | |
| "loss": 0.8065, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.9749430523917995, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00015075554907257796, | |
| "loss": 0.8129, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 3.9863325740318905, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.00015041259555507108, | |
| "loss": 0.8064, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.997722095671982, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00015006884521030848, | |
| "loss": 0.8131, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.417691946029663, | |
| "eval_runtime": 0.2354, | |
| "eval_samples_per_second": 42.48, | |
| "eval_steps_per_second": 4.248, | |
| "step": 1756 | |
| }, | |
| { | |
| "epoch": 4.009111617312073, | |
| "grad_norm": 0.7734375, | |
| "learning_rate": 0.00014972430347164742, | |
| "loss": 0.7909, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 4.020501138952164, | |
| "grad_norm": 0.7265625, | |
| "learning_rate": 0.0001493789757849541, | |
| "loss": 0.7848, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 4.031890660592255, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 0.00014903286760851737, | |
| "loss": 0.7893, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 4.043280182232346, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.0001486859844129628, | |
| "loss": 0.7902, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 4.054669703872437, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00014833833168116582, | |
| "loss": 0.7859, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 4.066059225512529, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00014798991490816532, | |
| "loss": 0.782, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 4.077448747152619, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00014764073960107666, | |
| "loss": 0.793, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 4.088838268792711, | |
| "grad_norm": 0.490234375, | |
| "learning_rate": 0.00014729081127900476, | |
| "loss": 0.783, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 4.100227790432802, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00014694013547295672, | |
| "loss": 0.7908, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.111617312072893, | |
| "grad_norm": 0.89453125, | |
| "learning_rate": 0.0001465887177257545, | |
| "loss": 0.7923, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 4.123006833712984, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.00014623656359194712, | |
| "loss": 0.7904, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 4.134396355353076, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 0.00014588367863772325, | |
| "loss": 0.7881, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 4.145785876993166, | |
| "grad_norm": 0.71875, | |
| "learning_rate": 0.00014553006844082283, | |
| "loss": 0.7831, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 4.157175398633258, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.00014517573859044907, | |
| "loss": 0.788, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 4.168564920273348, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00014482069468718022, | |
| "loss": 0.7853, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 4.17995444191344, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00014446494234288083, | |
| "loss": 0.7931, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 4.191343963553531, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 0.00014410848718061312, | |
| "loss": 0.7942, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 4.2027334851936216, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.0001437513348345482, | |
| "loss": 0.7834, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 4.214123006833713, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.00014339349094987699, | |
| "loss": 0.7797, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.225512528473804, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00014303496118272084, | |
| "loss": 0.7876, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 4.236902050113895, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00014267575120004231, | |
| "loss": 0.7943, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.248291571753986, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00014231586667955552, | |
| "loss": 0.7929, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 4.259681093394078, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00014195531330963635, | |
| "loss": 0.7842, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 4.271070615034168, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.00014159409678923265, | |
| "loss": 0.7878, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.28246013667426, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.0001412322228277741, | |
| "loss": 0.7843, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.29384965831435, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00014086969714508196, | |
| "loss": 0.7829, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 4.305239179954442, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00014050652547127864, | |
| "loss": 0.784, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.316628701594533, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00014014271354669718, | |
| "loss": 0.7815, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 4.328018223234624, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00013977826712179058, | |
| "loss": 0.7865, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.339407744874715, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 0.0001394131919570407, | |
| "loss": 0.7784, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 4.350797266514807, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00013904749382286734, | |
| "loss": 0.7846, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.362186788154897, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.0001386811784995371, | |
| "loss": 0.7905, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 4.373576309794989, | |
| "grad_norm": 0.6953125, | |
| "learning_rate": 0.00013831425177707193, | |
| "loss": 0.7936, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.38496583143508, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00013794671945515757, | |
| "loss": 0.7828, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 4.396355353075171, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 0.00013757858734305203, | |
| "loss": 0.7888, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.407744874715262, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00013720986125949353, | |
| "loss": 0.7852, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.4191343963553535, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 0.00013684054703260882, | |
| "loss": 0.7921, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.430523917995444, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00013647065049982078, | |
| "loss": 0.7872, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 4.4419134396355355, | |
| "grad_norm": 2.25, | |
| "learning_rate": 0.00013610017750775643, | |
| "loss": 0.7883, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.453302961275626, | |
| "grad_norm": 0.73828125, | |
| "learning_rate": 0.0001357291339121542, | |
| "loss": 0.7848, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 4.4646924829157175, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.0001353575255777717, | |
| "loss": 0.7894, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.476082004555809, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 0.00013498535837829276, | |
| "loss": 0.7911, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 4.4874715261958995, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 0.00013461263819623476, | |
| "loss": 0.7897, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.498861047835991, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00013423937092285555, | |
| "loss": 0.7833, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 4.510250569476082, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00013386556245806034, | |
| "loss": 0.8, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.521640091116173, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00013349121871030856, | |
| "loss": 0.7984, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 4.533029612756264, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00013311634559652036, | |
| "loss": 0.7938, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.544419134396355, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.000132740949041983, | |
| "loss": 0.7826, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 4.555808656036446, | |
| "grad_norm": 0.765625, | |
| "learning_rate": 0.00013236503498025747, | |
| "loss": 0.7922, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.567198177676538, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 0.00013198860935308444, | |
| "loss": 0.7796, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 4.578587699316628, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0001316116781102904, | |
| "loss": 0.7926, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.58997722095672, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.0001312342472096938, | |
| "loss": 0.7877, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 4.601366742596811, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 0.00013085632261701063, | |
| "loss": 0.7903, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.612756264236902, | |
| "grad_norm": 0.72265625, | |
| "learning_rate": 0.00013047791030576023, | |
| "loss": 0.7826, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 4.624145785876993, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 0.00013009901625717093, | |
| "loss": 0.7823, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.635535307517085, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00012971964646008542, | |
| "loss": 0.7884, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 4.646924829157175, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.0001293398069108662, | |
| "loss": 0.7846, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.658314350797267, | |
| "grad_norm": 0.64453125, | |
| "learning_rate": 0.00012895950361330058, | |
| "loss": 0.7822, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 4.669703872437358, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00012857874257850605, | |
| "loss": 0.7899, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.681093394077449, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00012819752982483508, | |
| "loss": 0.7914, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 4.69248291571754, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00012781587137778013, | |
| "loss": 0.7859, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.703872437357631, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00012743377326987826, | |
| "loss": 0.7849, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 4.715261958997722, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00012705124154061597, | |
| "loss": 0.7852, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 4.7266514806378135, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00012666828223633348, | |
| "loss": 0.7802, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 4.738041002277904, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00012628490141012937, | |
| "loss": 0.792, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 4.7494305239179955, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 0.00012590110512176498, | |
| "loss": 0.7915, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 4.760820045558087, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.0001255168994375683, | |
| "loss": 0.7859, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 4.7722095671981775, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.0001251322904303383, | |
| "loss": 0.7901, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 4.783599088838269, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.0001247472841792491, | |
| "loss": 0.7866, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 4.7949886104783594, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00012436188676975346, | |
| "loss": 0.7846, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 4.806378132118451, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 0.000123976104293487, | |
| "loss": 0.789, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 4.817767653758542, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00012358994284817167, | |
| "loss": 0.7765, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 4.829157175398633, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00012320340853751952, | |
| "loss": 0.7877, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 4.840546697038724, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00012281650747113612, | |
| "loss": 0.7862, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 4.851936218678816, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00012242924576442388, | |
| "loss": 0.7897, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 4.863325740318906, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00012204162953848581, | |
| "loss": 0.7782, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 4.874715261958998, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00012165366492002832, | |
| "loss": 0.7796, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 4.886104783599089, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.00012126535804126451, | |
| "loss": 0.791, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 4.89749430523918, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00012087671503981741, | |
| "loss": 0.7875, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 4.908883826879271, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00012048774205862279, | |
| "loss": 0.7783, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 4.920273348519363, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00012009844524583203, | |
| "loss": 0.7865, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 4.931662870159453, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00011970883075471522, | |
| "loss": 0.7899, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 4.943052391799545, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00011931890474356358, | |
| "loss": 0.7838, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 4.954441913439636, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00011892867337559221, | |
| "loss": 0.7859, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 4.965831435079727, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00011853814281884283, | |
| "loss": 0.7794, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 4.977220956719818, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 0.00011814731924608616, | |
| "loss": 0.7793, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 4.988610478359909, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00011775620883472424, | |
| "loss": 0.7818, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.7788, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.416613817214966, | |
| "eval_runtime": 0.2352, | |
| "eval_samples_per_second": 42.522, | |
| "eval_steps_per_second": 4.252, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 5.011389521640091, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00011697315222836458, | |
| "loss": 0.7695, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.022779043280182, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00011658121841044922, | |
| "loss": 0.7684, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 5.034168564920273, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.0001161890225078977, | |
| "loss": 0.7578, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 5.045558086560365, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 0.0001157965707198034, | |
| "loss": 0.7611, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 5.056947608200455, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.00011540386924930413, | |
| "loss": 0.7611, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 5.068337129840547, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.00011501092430348435, | |
| "loss": 0.7644, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 5.079726651480637, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 0.0001146177420932768, | |
| "loss": 0.7635, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 5.091116173120729, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00011422432883336456, | |
| "loss": 0.7639, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 5.10250569476082, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00011383069074208259, | |
| "loss": 0.77, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 5.113895216400911, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00011343683404131964, | |
| "loss": 0.7643, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 5.125284738041002, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00011304276495641981, | |
| "loss": 0.7696, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.136674259681094, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.0001126484897160842, | |
| "loss": 0.7619, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 5.148063781321184, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.0001122540145522723, | |
| "loss": 0.7612, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 5.159453302961276, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.00011185934570010374, | |
| "loss": 0.7596, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 5.170842824601367, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00011146448939775962, | |
| "loss": 0.7652, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 5.182232346241458, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00011106945188638378, | |
| "loss": 0.763, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 5.193621867881549, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00011067423940998438, | |
| "loss": 0.7633, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 5.20501138952164, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 0.00011027885821533508, | |
| "loss": 0.7679, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 5.216400911161731, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00010988331455187628, | |
| "loss": 0.7661, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 5.2277904328018225, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00010948761467161637, | |
| "loss": 0.7699, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 5.239179954441913, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00010909176482903295, | |
| "loss": 0.7734, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.2505694760820045, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 0.00010869577128097404, | |
| "loss": 0.7675, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 5.261958997722096, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00010829964028655885, | |
| "loss": 0.7645, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 5.2733485193621865, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00010790337810707931, | |
| "loss": 0.767, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 5.284738041002278, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00010750699100590076, | |
| "loss": 0.7662, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 5.296127562642369, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 0.00010711048524836311, | |
| "loss": 0.7673, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 5.30751708428246, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.0001067138671016817, | |
| "loss": 0.7666, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 5.318906605922551, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00010631714283484842, | |
| "loss": 0.7687, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 5.330296127562642, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 0.00010592031871853239, | |
| "loss": 0.771, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.341685649202733, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00010552340102498104, | |
| "loss": 0.7624, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.353075170842825, | |
| "grad_norm": 0.46875, | |
| "learning_rate": 0.00010512639602792088, | |
| "loss": 0.7654, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 5.364464692482915, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.0001047293100024583, | |
| "loss": 0.7585, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 5.375854214123007, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00010433214922498047, | |
| "loss": 0.7622, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 5.387243735763098, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.00010393491997305613, | |
| "loss": 0.7711, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 5.398633257403189, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.0001035376285253363, | |
| "loss": 0.7672, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 5.41002277904328, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00010314028116145509, | |
| "loss": 0.7748, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 5.421412300683372, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00010274288416193034, | |
| "loss": 0.7648, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 5.432801822323462, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00010234544380806461, | |
| "loss": 0.7623, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 5.444191343963554, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00010194796638184558, | |
| "loss": 0.7707, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 5.455580865603645, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00010155045816584691, | |
| "loss": 0.7629, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 5.466970387243736, | |
| "grad_norm": 0.92578125, | |
| "learning_rate": 0.00010115292544312904, | |
| "loss": 0.7728, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.478359908883827, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00010075537449713963, | |
| "loss": 0.7704, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 5.489749430523918, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00010035781161161446, | |
| "loss": 0.7731, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 5.501138952164009, | |
| "grad_norm": 0.5, | |
| "learning_rate": 9.996024307047798e-05, | |
| "loss": 0.7712, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 5.5125284738041005, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.956267515774412e-05, | |
| "loss": 0.7647, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 5.523917995444191, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 9.916511415741676e-05, | |
| "loss": 0.7712, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 5.5353075170842825, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 9.876756635339058e-05, | |
| "loss": 0.77, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 5.546697038724374, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 9.83700380293517e-05, | |
| "loss": 0.7637, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 5.5580865603644645, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 9.797253546867831e-05, | |
| "loss": 0.7676, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 5.569476082004556, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.757506495434133e-05, | |
| "loss": 0.7601, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 5.5808656036446465, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 9.71776327688053e-05, | |
| "loss": 0.7657, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 5.592255125284738, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.678024519392871e-05, | |
| "loss": 0.7737, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 5.603644646924829, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 9.638290851086518e-05, | |
| "loss": 0.7676, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 5.61503416856492, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 9.598562899996375e-05, | |
| "loss": 0.7704, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 5.626423690205011, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 9.558841294066985e-05, | |
| "loss": 0.7631, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 5.637813211845103, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.519126661142597e-05, | |
| "loss": 0.7657, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 5.649202733485193, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.479419628957246e-05, | |
| "loss": 0.7668, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 5.660592255125285, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.439720825124827e-05, | |
| "loss": 0.7604, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 5.671981776765376, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.400030877129176e-05, | |
| "loss": 0.7685, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 5.683371298405467, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.360350412314157e-05, | |
| "loss": 0.7715, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 5.694760820045558, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.320680057873735e-05, | |
| "loss": 0.7628, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.70615034168565, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.281020440842079e-05, | |
| "loss": 0.7629, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 5.71753986332574, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 9.241372188083631e-05, | |
| "loss": 0.7585, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 5.728929384965832, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.201735926283213e-05, | |
| "loss": 0.768, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 5.740318906605923, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.162112281936118e-05, | |
| "loss": 0.7658, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 5.751708428246014, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.122501881338199e-05, | |
| "loss": 0.7681, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 5.763097949886105, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.082905350575986e-05, | |
| "loss": 0.7653, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 5.774487471526196, | |
| "grad_norm": 0.375, | |
| "learning_rate": 9.043323315516775e-05, | |
| "loss": 0.7711, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 5.785876993166287, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.003756401798744e-05, | |
| "loss": 0.7596, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 5.7972665148063784, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 8.96420523482106e-05, | |
| "loss": 0.7649, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 5.808656036446469, | |
| "grad_norm": 0.375, | |
| "learning_rate": 8.924670439733997e-05, | |
| "loss": 0.7686, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 5.82004555808656, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 8.885152641429049e-05, | |
| "loss": 0.771, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 5.831435079726651, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 8.845652464529057e-05, | |
| "loss": 0.7638, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 5.842824601366742, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 8.806170533378345e-05, | |
| "loss": 0.7705, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 5.854214123006834, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 8.766707472032831e-05, | |
| "loss": 0.768, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 5.865603644646924, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 8.727263904250178e-05, | |
| "loss": 0.7626, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 5.876993166287016, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 8.687840453479938e-05, | |
| "loss": 0.7728, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 5.888382687927107, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 8.648437742853685e-05, | |
| "loss": 0.7665, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 5.899772209567198, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 8.609056395175175e-05, | |
| "loss": 0.7613, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 5.911161731207289, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 8.569697032910492e-05, | |
| "loss": 0.7712, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 5.922551252847381, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 8.530360278178227e-05, | |
| "loss": 0.7704, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 5.933940774487471, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 8.491046752739624e-05, | |
| "loss": 0.7672, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 5.945330296127563, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 8.451757077988767e-05, | |
| "loss": 0.7701, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 5.956719817767654, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 8.41249187494275e-05, | |
| "loss": 0.7666, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 5.968109339407745, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.373251764231872e-05, | |
| "loss": 0.7562, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 5.979498861047836, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 8.334037366089813e-05, | |
| "loss": 0.765, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 5.990888382687928, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 8.294849300343836e-05, | |
| "loss": 0.771, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.432857036590576, | |
| "eval_runtime": 0.2435, | |
| "eval_samples_per_second": 41.072, | |
| "eval_steps_per_second": 4.107, | |
| "step": 2634 | |
| }, | |
| { | |
| "epoch": 6.002277904328018, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 8.255688186404996e-05, | |
| "loss": 0.77, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.0136674259681095, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 8.216554643258342e-05, | |
| "loss": 0.748, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.0250569476082, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 8.177449289453134e-05, | |
| "loss": 0.7503, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 6.0364464692482915, | |
| "grad_norm": 0.5, | |
| "learning_rate": 8.138372743093076e-05, | |
| "loss": 0.7419, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 6.047835990888383, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 8.099325621826526e-05, | |
| "loss": 0.7518, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 6.0592255125284735, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 8.060308542836755e-05, | |
| "loss": 0.76, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 6.070615034168565, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 8.021322122832178e-05, | |
| "loss": 0.7567, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 6.082004555808656, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 7.982366978036618e-05, | |
| "loss": 0.7548, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 6.093394077448747, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 7.943443724179548e-05, | |
| "loss": 0.7557, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 6.104783599088838, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 7.904552976486372e-05, | |
| "loss": 0.7571, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 6.116173120728929, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 7.865695349668703e-05, | |
| "loss": 0.7572, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 6.12756264236902, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 7.826871457914639e-05, | |
| "loss": 0.7415, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 6.138952164009112, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 7.788081914879051e-05, | |
| "loss": 0.7547, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 6.150341685649202, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 7.7493273336739e-05, | |
| "loss": 0.7431, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.161731207289294, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 7.710608326858535e-05, | |
| "loss": 0.7555, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 6.173120728929385, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 7.67192550643001e-05, | |
| "loss": 0.7472, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 6.184510250569476, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 7.633279483813405e-05, | |
| "loss": 0.7569, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 6.195899772209567, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 7.594670869852185e-05, | |
| "loss": 0.7494, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 6.207289293849659, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 7.556100274798519e-05, | |
| "loss": 0.7532, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 6.218678815489749, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 7.517568308303643e-05, | |
| "loss": 0.7581, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 6.230068337129841, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 7.47907557940824e-05, | |
| "loss": 0.7569, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 6.241457858769932, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 7.440622696532775e-05, | |
| "loss": 0.7549, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 6.252847380410023, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 7.402210267467928e-05, | |
| "loss": 0.7478, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 6.264236902050114, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 7.363838899364944e-05, | |
| "loss": 0.7515, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 6.275626423690205, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 7.325509198726064e-05, | |
| "loss": 0.7471, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 6.287015945330296, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 7.287221771394917e-05, | |
| "loss": 0.7565, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 6.2984054669703875, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 7.248977222546968e-05, | |
| "loss": 0.7572, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 6.309794988610478, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 7.210776156679931e-05, | |
| "loss": 0.7442, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 6.3211845102505695, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 7.172619177604223e-05, | |
| "loss": 0.7615, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 6.332574031890661, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 7.134506888433426e-05, | |
| "loss": 0.7538, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 6.3439635535307515, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 7.096439891574745e-05, | |
| "loss": 0.7445, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 6.355353075170843, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 7.058418788719491e-05, | |
| "loss": 0.7593, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 6.366742596810934, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 7.020444180833564e-05, | |
| "loss": 0.7603, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 6.378132118451025, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 6.982516668147967e-05, | |
| "loss": 0.7544, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 6.389521640091116, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 6.944636850149306e-05, | |
| "loss": 0.7508, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 6.400911161731207, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 6.906805325570316e-05, | |
| "loss": 0.7587, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 6.412300683371298, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 6.869022692380411e-05, | |
| "loss": 0.7456, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 6.42369020501139, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 6.831289547776207e-05, | |
| "loss": 0.7541, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 6.43507972665148, | |
| "grad_norm": 0.4765625, | |
| "learning_rate": 6.793606488172118e-05, | |
| "loss": 0.7477, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 6.446469248291572, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 6.75597410919089e-05, | |
| "loss": 0.7508, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 6.457858769931663, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 6.718393005654215e-05, | |
| "loss": 0.7583, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 6.469248291571754, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 6.680863771573318e-05, | |
| "loss": 0.7533, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 6.480637813211845, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 6.643387000139565e-05, | |
| "loss": 0.7506, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 6.492027334851937, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 6.6059632837151e-05, | |
| "loss": 0.7556, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 6.503416856492027, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 6.568593213823465e-05, | |
| "loss": 0.7505, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 6.514806378132119, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 6.53127738114026e-05, | |
| "loss": 0.7499, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 6.52619589977221, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 6.494016375483811e-05, | |
| "loss": 0.7614, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 6.537585421412301, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 6.456810785805842e-05, | |
| "loss": 0.7412, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 6.548974943052392, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 6.419661200182158e-05, | |
| "loss": 0.7612, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 6.560364464692483, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 6.38256820580336e-05, | |
| "loss": 0.7529, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 6.571753986332574, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 6.345532388965565e-05, | |
| "loss": 0.7536, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 6.5831435079726655, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 6.308554335061135e-05, | |
| "loss": 0.7533, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 6.594533029612756, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 6.271634628569418e-05, | |
| "loss": 0.7502, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 6.605922551252847, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 6.234773853047526e-05, | |
| "loss": 0.7513, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 6.617312072892939, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 6.19797259112109e-05, | |
| "loss": 0.7542, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 6.628701594533029, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 6.161231424475075e-05, | |
| "loss": 0.7499, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 6.640091116173121, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 6.124550933844562e-05, | |
| "loss": 0.7522, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 6.651480637813211, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 6.087931699005588e-05, | |
| "loss": 0.7573, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 6.662870159453303, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 6.0513742987659686e-05, | |
| "loss": 0.7557, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 6.674259681093394, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 6.014879310956154e-05, | |
| "loss": 0.7554, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 6.685649202733485, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 5.978447312420103e-05, | |
| "loss": 0.7524, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 6.697038724373576, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 5.9420788790061544e-05, | |
| "loss": 0.7571, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 6.708428246013668, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 5.905774585557922e-05, | |
| "loss": 0.7512, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 6.719817767653758, | |
| "grad_norm": 0.375, | |
| "learning_rate": 5.869535005905232e-05, | |
| "loss": 0.747, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 6.73120728929385, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 5.833360712855029e-05, | |
| "loss": 0.7562, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 6.742596810933941, | |
| "grad_norm": 0.5, | |
| "learning_rate": 5.7972522781823256e-05, | |
| "loss": 0.752, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 6.753986332574032, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 5.761210272621175e-05, | |
| "loss": 0.7494, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 6.765375854214123, | |
| "grad_norm": 0.375, | |
| "learning_rate": 5.7252352658556376e-05, | |
| "loss": 0.7533, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 6.776765375854215, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 5.689327826510796e-05, | |
| "loss": 0.7486, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 6.788154897494305, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.653488522143744e-05, | |
| "loss": 0.7489, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 6.7995444191343966, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 5.617717919234624e-05, | |
| "loss": 0.7518, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 6.810933940774487, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 5.582016583177687e-05, | |
| "loss": 0.7607, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 6.8223234624145785, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 5.5463850782723346e-05, | |
| "loss": 0.7518, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 6.83371298405467, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 5.5108239677142115e-05, | |
| "loss": 0.7445, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 6.8451025056947605, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 5.475333813586297e-05, | |
| "loss": 0.7496, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 6.856492027334852, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 5.439915176850037e-05, | |
| "loss": 0.7585, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 6.867881548974943, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.404568617336456e-05, | |
| "loss": 0.7485, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 6.879271070615034, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 5.369294693737319e-05, | |
| "loss": 0.755, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 6.890660592255125, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 5.334093963596294e-05, | |
| "loss": 0.7556, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 6.902050113895216, | |
| "grad_norm": 0.52734375, | |
| "learning_rate": 5.298966983300161e-05, | |
| "loss": 0.7474, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 6.913439635535307, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 5.263914308069986e-05, | |
| "loss": 0.7531, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 6.924829157175399, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 5.228936491952363e-05, | |
| "loss": 0.7501, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 6.936218678815489, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.194034087810665e-05, | |
| "loss": 0.7469, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 6.947608200455581, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 5.159207647316282e-05, | |
| "loss": 0.7559, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 6.958997722095672, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 5.12445772093992e-05, | |
| "loss": 0.748, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 6.970387243735763, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 5.089784857942892e-05, | |
| "loss": 0.7525, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 6.981776765375854, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 5.055189606368436e-05, | |
| "loss": 0.7544, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 6.993166287015946, | |
| "grad_norm": 0.423828125, | |
| "learning_rate": 5.020672513033066e-05, | |
| "loss": 0.7459, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 2.445798397064209, | |
| "eval_runtime": 0.2348, | |
| "eval_samples_per_second": 42.591, | |
| "eval_steps_per_second": 4.259, | |
| "step": 3073 | |
| }, | |
| { | |
| "epoch": 7.004555808656036, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 4.9862341235179014e-05, | |
| "loss": 0.7506, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 7.015945330296128, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 4.951874982160079e-05, | |
| "loss": 0.7452, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 7.027334851936219, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 4.917595632044113e-05, | |
| "loss": 0.7421, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 7.03872437357631, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 4.8833966149933364e-05, | |
| "loss": 0.744, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 7.050113895216401, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 4.849278471561328e-05, | |
| "loss": 0.7418, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 7.061503416856492, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 4.815241741023367e-05, | |
| "loss": 0.7488, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.072892938496583, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 4.7812869613679103e-05, | |
| "loss": 0.7373, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 7.0842824601366745, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 4.747414669288094e-05, | |
| "loss": 0.7441, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 7.095671981776765, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 4.713625400173247e-05, | |
| "loss": 0.7439, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 7.1070615034168565, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 4.679919688100423e-05, | |
| "loss": 0.7471, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 7.118451025056948, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 4.6462980658259625e-05, | |
| "loss": 0.7476, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 7.1298405466970385, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 4.6127610647770767e-05, | |
| "loss": 0.7365, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 7.14123006833713, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 4.5793092150434405e-05, | |
| "loss": 0.7422, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 7.152619589977221, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 4.545943045368826e-05, | |
| "loss": 0.7483, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 7.164009111617312, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 4.5126630831427264e-05, | |
| "loss": 0.743, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 7.175398633257403, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 4.479469854392031e-05, | |
| "loss": 0.7414, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 7.186788154897494, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 4.4463638837727196e-05, | |
| "loss": 0.7382, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 7.198177676537585, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 4.413345694561549e-05, | |
| "loss": 0.7365, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 7.209567198177677, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 4.3804158086477986e-05, | |
| "loss": 0.7412, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 7.220956719817767, | |
| "grad_norm": 0.375, | |
| "learning_rate": 4.34757474652501e-05, | |
| "loss": 0.7491, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 7.232346241457859, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 4.3148230272827784e-05, | |
| "loss": 0.7452, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 7.24373576309795, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.282161168598523e-05, | |
| "loss": 0.7496, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 7.255125284738041, | |
| "grad_norm": 0.375, | |
| "learning_rate": 4.249589686729319e-05, | |
| "loss": 0.7409, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 7.266514806378132, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 4.217109096503736e-05, | |
| "loss": 0.742, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 7.277904328018224, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.184719911313707e-05, | |
| "loss": 0.7367, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 7.289293849658314, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.152422643106396e-05, | |
| "loss": 0.7467, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 7.300683371298406, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 4.1202178023761195e-05, | |
| "loss": 0.7416, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 7.312072892938497, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 4.088105898156282e-05, | |
| "loss": 0.7483, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 7.323462414578588, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 4.0560874380113146e-05, | |
| "loss": 0.7444, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 7.334851936218679, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 4.024162928028663e-05, | |
| "loss": 0.7417, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 7.34624145785877, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.9923328728107856e-05, | |
| "loss": 0.743, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 7.357630979498861, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.960597775467177e-05, | |
| "loss": 0.7482, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 7.3690205011389525, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.928958137606421e-05, | |
| "loss": 0.7473, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 7.380410022779043, | |
| "grad_norm": 0.375, | |
| "learning_rate": 3.8974144593282534e-05, | |
| "loss": 0.7429, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 7.3917995444191344, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 3.865967239215667e-05, | |
| "loss": 0.7481, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 7.403189066059226, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.834616974327021e-05, | |
| "loss": 0.7445, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 7.414578587699316, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 3.80336416018819e-05, | |
| "loss": 0.7431, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 7.425968109339408, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 3.7722092907847305e-05, | |
| "loss": 0.7394, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 7.437357630979498, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 3.741152858554077e-05, | |
| "loss": 0.744, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 7.44874715261959, | |
| "grad_norm": 0.375, | |
| "learning_rate": 3.710195354377747e-05, | |
| "loss": 0.7408, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 7.460136674259681, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 3.679337267573597e-05, | |
| "loss": 0.7361, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 7.471526195899772, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 3.648579085888085e-05, | |
| "loss": 0.7467, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 7.482915717539863, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.6179212954885477e-05, | |
| "loss": 0.738, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 7.494305239179955, | |
| "grad_norm": 0.375, | |
| "learning_rate": 3.587364380955529e-05, | |
| "loss": 0.7475, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 7.505694760820045, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 3.556908825275117e-05, | |
| "loss": 0.7434, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 7.517084282460137, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.526555109831311e-05, | |
| "loss": 0.7477, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 7.528473804100228, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 3.4963037143984087e-05, | |
| "loss": 0.7413, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 7.539863325740319, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 3.466155117133433e-05, | |
| "loss": 0.748, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 7.55125284738041, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 3.436109794568565e-05, | |
| "loss": 0.7444, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 7.562642369020502, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 3.406168221603611e-05, | |
| "loss": 0.7387, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 7.574031890660592, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 3.3763308714984974e-05, | |
| "loss": 0.7436, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 7.585421412300684, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 3.3465982158657984e-05, | |
| "loss": 0.7413, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 7.596810933940774, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 3.3169707246632705e-05, | |
| "loss": 0.7423, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 7.6082004555808656, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.287448866186428e-05, | |
| "loss": 0.7392, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 7.619589977220957, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 3.258033107061153e-05, | |
| "loss": 0.7461, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 7.6309794988610475, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 3.228723912236291e-05, | |
| "loss": 0.7365, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 7.642369020501139, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 3.199521744976342e-05, | |
| "loss": 0.7414, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 7.65375854214123, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 3.170427066854096e-05, | |
| "loss": 0.7548, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 7.665148063781321, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 3.141440337743369e-05, | |
| "loss": 0.739, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 7.676537585421412, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 3.1125620158117186e-05, | |
| "loss": 0.7505, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 7.687927107061503, | |
| "grad_norm": 0.375, | |
| "learning_rate": 3.0837925575132024e-05, | |
| "loss": 0.7487, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 7.699316628701594, | |
| "grad_norm": 0.447265625, | |
| "learning_rate": 3.055132417581179e-05, | |
| "loss": 0.7427, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 7.710706150341686, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 3.0265820490210973e-05, | |
| "loss": 0.7384, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 7.722095671981776, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.9981419031033498e-05, | |
| "loss": 0.7402, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 7.733485193621868, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 2.9698124293561357e-05, | |
| "loss": 0.7485, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 7.744874715261959, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 2.941594075558366e-05, | |
| "loss": 0.7505, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 7.75626423690205, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 2.913487287732565e-05, | |
| "loss": 0.7446, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 7.767653758542141, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 2.8854925101378438e-05, | |
| "loss": 0.7461, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 7.779043280182233, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 2.857610185262859e-05, | |
| "loss": 0.75, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 7.790432801822323, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 2.8298407538188288e-05, | |
| "loss": 0.7469, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 7.801822323462415, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 2.8021846547325635e-05, | |
| "loss": 0.7437, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 7.813211845102506, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 2.774642325139535e-05, | |
| "loss": 0.7408, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 7.824601366742597, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 2.7472142003769495e-05, | |
| "loss": 0.7431, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 7.835990888382688, | |
| "grad_norm": 0.375, | |
| "learning_rate": 2.7199007139768928e-05, | |
| "loss": 0.7475, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 7.8473804100227795, | |
| "grad_norm": 0.375, | |
| "learning_rate": 2.6927022976594607e-05, | |
| "loss": 0.7371, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 7.85876993166287, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 2.665619381325929e-05, | |
| "loss": 0.7477, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 7.8701594533029615, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 2.638652393051976e-05, | |
| "loss": 0.7433, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 7.881548974943052, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 2.6118017590809017e-05, | |
| "loss": 0.7401, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 7.8929384965831435, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.5850679038169045e-05, | |
| "loss": 0.7415, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 7.904328018223235, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 2.5584512498183544e-05, | |
| "loss": 0.7309, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 7.9157175398633255, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.531952217791136e-05, | |
| "loss": 0.7422, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 7.927107061503417, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 2.505571226581984e-05, | |
| "loss": 0.7434, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 7.9384965831435075, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 2.4793086931718634e-05, | |
| "loss": 0.7451, | |
| "step": 3485 | |
| }, | |
| { | |
| "epoch": 7.949886104783599, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 2.4531650326693822e-05, | |
| "loss": 0.7455, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 7.96127562642369, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 2.4271406583042335e-05, | |
| "loss": 0.7393, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 7.972665148063781, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 2.401235981420653e-05, | |
| "loss": 0.7443, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 7.984054669703872, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.3754514114709304e-05, | |
| "loss": 0.7429, | |
| "step": 3505 | |
| }, | |
| { | |
| "epoch": 7.995444191343964, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.3497873560089322e-05, | |
| "loss": 0.745, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.460949420928955, | |
| "eval_runtime": 0.2436, | |
| "eval_samples_per_second": 41.058, | |
| "eval_steps_per_second": 4.106, | |
| "step": 3512 | |
| }, | |
| { | |
| "epoch": 8.006833712984054, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 2.3242442206836523e-05, | |
| "loss": 0.749, | |
| "step": 3515 | |
| }, | |
| { | |
| "epoch": 8.018223234624147, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.298822409232817e-05, | |
| "loss": 0.7467, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 8.029612756264237, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 2.2735223234764846e-05, | |
| "loss": 0.735, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 8.041002277904328, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 2.2483443633107058e-05, | |
| "loss": 0.7392, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 8.052391799544418, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 2.2232889267012038e-05, | |
| "loss": 0.7377, | |
| "step": 3535 | |
| }, | |
| { | |
| "epoch": 8.06378132118451, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 2.1983564096770725e-05, | |
| "loss": 0.748, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 8.075170842824601, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 2.1735472063245354e-05, | |
| "loss": 0.7325, | |
| "step": 3545 | |
| }, | |
| { | |
| "epoch": 8.086560364464692, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 2.1488617087806982e-05, | |
| "loss": 0.7372, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 8.097949886104784, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 2.1243003072273582e-05, | |
| "loss": 0.7391, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 8.109339407744875, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 2.0998633898848442e-05, | |
| "loss": 0.7467, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 8.120728929384965, | |
| "grad_norm": 0.375, | |
| "learning_rate": 2.0755513430058672e-05, | |
| "loss": 0.7418, | |
| "step": 3565 | |
| }, | |
| { | |
| "epoch": 8.132118451025057, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 2.0513645508694225e-05, | |
| "loss": 0.7383, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 8.143507972665148, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 2.0273033957747134e-05, | |
| "loss": 0.7417, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 8.154897494305239, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.0033682580351144e-05, | |
| "loss": 0.7386, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 8.166287015945331, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 1.9795595159721524e-05, | |
| "loss": 0.7449, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 8.177676537585421, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.955877545909528e-05, | |
| "loss": 0.7429, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 8.189066059225512, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.932322722167168e-05, | |
| "loss": 0.7391, | |
| "step": 3595 | |
| }, | |
| { | |
| "epoch": 8.200455580865604, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.9088954170553198e-05, | |
| "loss": 0.7389, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 8.211845102505695, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.8855960008686446e-05, | |
| "loss": 0.7406, | |
| "step": 3605 | |
| }, | |
| { | |
| "epoch": 8.223234624145785, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.86242484188038e-05, | |
| "loss": 0.736, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 8.234624145785878, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.8393823063365223e-05, | |
| "loss": 0.7449, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 8.246013667425968, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.816468758450024e-05, | |
| "loss": 0.7346, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 8.257403189066059, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.7936845603950447e-05, | |
| "loss": 0.743, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 8.268792710706151, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.7710300723012262e-05, | |
| "loss": 0.7421, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 8.280182232346242, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.7485056522480004e-05, | |
| "loss": 0.7365, | |
| "step": 3635 | |
| }, | |
| { | |
| "epoch": 8.291571753986332, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 1.726111656258932e-05, | |
| "loss": 0.741, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 8.302961275626423, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.7038484382960796e-05, | |
| "loss": 0.736, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 8.314350797266515, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.6817163502544208e-05, | |
| "loss": 0.7342, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 8.325740318906606, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 1.6597157419562703e-05, | |
| "loss": 0.7331, | |
| "step": 3655 | |
| }, | |
| { | |
| "epoch": 8.337129840546696, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.6378469611457592e-05, | |
| "loss": 0.7375, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 8.348519362186789, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 1.6161103534833423e-05, | |
| "loss": 0.7431, | |
| "step": 3665 | |
| }, | |
| { | |
| "epoch": 8.35990888382688, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.594506262540324e-05, | |
| "loss": 0.7431, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 8.37129840546697, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 1.5730350297934448e-05, | |
| "loss": 0.7392, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 8.382687927107062, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.5516969946194626e-05, | |
| "loss": 0.7355, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 8.394077448747153, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.5304924942898068e-05, | |
| "loss": 0.7388, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 8.405466970387243, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.509421863965237e-05, | |
| "loss": 0.7441, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 8.416856492027335, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.4884854366905455e-05, | |
| "loss": 0.7324, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 8.428246013667426, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.4676835433892989e-05, | |
| "loss": 0.7403, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 8.439635535307517, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 1.4470165128586022e-05, | |
| "loss": 0.7422, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 8.451025056947609, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.4264846717639102e-05, | |
| "loss": 0.7403, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 8.4624145785877, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 1.4060883446338502e-05, | |
| "loss": 0.7358, | |
| "step": 3715 | |
| }, | |
| { | |
| "epoch": 8.47380410022779, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 1.3858278538551018e-05, | |
| "loss": 0.7384, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 8.485193621867882, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 1.3657035196673052e-05, | |
| "loss": 0.7365, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 8.496583143507973, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.345715660157989e-05, | |
| "loss": 0.7389, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 8.507972665148063, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.3258645912575484e-05, | |
| "loss": 0.7459, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 8.519362186788156, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.3061506267342472e-05, | |
| "loss": 0.7482, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 8.530751708428246, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 1.2865740781892699e-05, | |
| "loss": 0.7441, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 8.542141230068337, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.2671352550517823e-05, | |
| "loss": 0.7379, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 8.55353075170843, | |
| "grad_norm": 0.375, | |
| "learning_rate": 1.2478344645740469e-05, | |
| "loss": 0.7386, | |
| "step": 3755 | |
| }, | |
| { | |
| "epoch": 8.56492027334852, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.2286720118265659e-05, | |
| "loss": 0.7435, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 8.57630979498861, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.209648199693264e-05, | |
| "loss": 0.7412, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 8.5876993166287, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.190763328866693e-05, | |
| "loss": 0.7414, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 8.599088838268793, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.1720176978432795e-05, | |
| "loss": 0.7393, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 8.610478359908884, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 1.1534116029186181e-05, | |
| "loss": 0.7333, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 8.621867881548974, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.1349453381827713e-05, | |
| "loss": 0.7345, | |
| "step": 3785 | |
| }, | |
| { | |
| "epoch": 8.633257403189067, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.1166191955156346e-05, | |
| "loss": 0.7531, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 8.644646924829157, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.0984334645823158e-05, | |
| "loss": 0.7359, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 8.656036446469248, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.0803884328285586e-05, | |
| "loss": 0.7441, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 8.66742596810934, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.0624843854762034e-05, | |
| "loss": 0.7353, | |
| "step": 3805 | |
| }, | |
| { | |
| "epoch": 8.67881548974943, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.0447216055186681e-05, | |
| "loss": 0.7407, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 8.690205011389521, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 1.0271003737164909e-05, | |
| "loss": 0.7372, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 8.701594533029613, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.009620968592876e-05, | |
| "loss": 0.7445, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 8.712984054669704, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.922836664293022e-06, | |
| "loss": 0.7362, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 8.724373576309794, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.750887412611508e-06, | |
| "loss": 0.7408, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 8.735763097949887, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.580364648733775e-06, | |
| "loss": 0.7347, | |
| "step": 3835 | |
| }, | |
| { | |
| "epoch": 8.747152619589977, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.411271067962124e-06, | |
| "loss": 0.738, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 8.758542141230068, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.243609343009086e-06, | |
| "loss": 0.7391, | |
| "step": 3845 | |
| }, | |
| { | |
| "epoch": 8.76993166287016, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.07738212395508e-06, | |
| "loss": 0.7405, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 8.78132118451025, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 8.912592038206546e-06, | |
| "loss": 0.7391, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 8.792710706150341, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 8.749241690454424e-06, | |
| "loss": 0.7367, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 8.804100227790432, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 8.587333662633035e-06, | |
| "loss": 0.7411, | |
| "step": 3865 | |
| }, | |
| { | |
| "epoch": 8.815489749430524, | |
| "grad_norm": 0.375, | |
| "learning_rate": 8.426870513879182e-06, | |
| "loss": 0.7365, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 8.826879271070615, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 8.267854780491747e-06, | |
| "loss": 0.7458, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 8.838268792710707, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 8.110288975891634e-06, | |
| "loss": 0.7326, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 8.849658314350798, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 7.954175590581992e-06, | |
| "loss": 0.7374, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 8.861047835990888, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 7.799517092108855e-06, | |
| "loss": 0.7345, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 8.872437357630979, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 7.646315925022152e-06, | |
| "loss": 0.7384, | |
| "step": 3895 | |
| }, | |
| { | |
| "epoch": 8.883826879271071, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 7.49457451083706e-06, | |
| "loss": 0.7319, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 8.895216400911162, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 7.344295247995725e-06, | |
| "loss": 0.7432, | |
| "step": 3905 | |
| }, | |
| { | |
| "epoch": 8.906605922551252, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 7.195480511829411e-06, | |
| "loss": 0.7392, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 8.917995444191344, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 7.048132654520856e-06, | |
| "loss": 0.7465, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 8.929384965831435, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 6.902254005067166e-06, | |
| "loss": 0.7387, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 8.940774487471526, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 6.7578468692429345e-06, | |
| "loss": 0.74, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 8.952164009111618, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 6.614913529563927e-06, | |
| "loss": 0.7346, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 8.963553530751708, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 6.4734562452508525e-06, | |
| "loss": 0.739, | |
| "step": 3935 | |
| }, | |
| { | |
| "epoch": 8.974943052391799, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 6.333477252193731e-06, | |
| "loss": 0.7418, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 8.986332574031891, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 6.19497876291657e-06, | |
| "loss": 0.7394, | |
| "step": 3945 | |
| }, | |
| { | |
| "epoch": 8.997722095671982, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 6.057962966542319e-06, | |
| "loss": 0.7433, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 2.4639732837677, | |
| "eval_runtime": 0.2359, | |
| "eval_samples_per_second": 42.387, | |
| "eval_steps_per_second": 4.239, | |
| "step": 3951 | |
| }, | |
| { | |
| "epoch": 9.009111617312072, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 5.922432028758362e-06, | |
| "loss": 0.7355, | |
| "step": 3955 | |
| }, | |
| { | |
| "epoch": 9.020501138952165, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 5.788388091782204e-06, | |
| "loss": 0.743, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 9.031890660592255, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 5.655833274327638e-06, | |
| "loss": 0.7396, | |
| "step": 3965 | |
| }, | |
| { | |
| "epoch": 9.043280182232346, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 5.524769671571317e-06, | |
| "loss": 0.734, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 9.054669703872438, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 5.395199355119518e-06, | |
| "loss": 0.7406, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 9.066059225512529, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 5.267124372975518e-06, | |
| "loss": 0.7398, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 9.07744874715262, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 5.140546749507136e-06, | |
| "loss": 0.7425, | |
| "step": 3985 | |
| }, | |
| { | |
| "epoch": 9.08883826879271, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 5.0154684854147645e-06, | |
| "loss": 0.7335, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 9.100227790432802, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 4.891891557699779e-06, | |
| "loss": 0.7502, | |
| "step": 3995 | |
| }, | |
| { | |
| "epoch": 9.111617312072893, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 4.769817919633235e-06, | |
| "loss": 0.737, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.123006833712983, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 4.649249500725017e-06, | |
| "loss": 0.7357, | |
| "step": 4005 | |
| }, | |
| { | |
| "epoch": 9.134396355353076, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 4.530188206693375e-06, | |
| "loss": 0.7419, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 9.145785876993166, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 4.412635919434749e-06, | |
| "loss": 0.7442, | |
| "step": 4015 | |
| }, | |
| { | |
| "epoch": 9.157175398633257, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 4.296594496994055e-06, | |
| "loss": 0.7385, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 9.168564920273349, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 4.182065773535271e-06, | |
| "loss": 0.7394, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 9.17995444191344, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.069051559312531e-06, | |
| "loss": 0.7384, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 9.19134396355353, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 3.957553640641442e-06, | |
| "loss": 0.7342, | |
| "step": 4035 | |
| }, | |
| { | |
| "epoch": 9.202733485193622, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.847573779870839e-06, | |
| "loss": 0.734, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 9.214123006833713, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.7391137153550137e-06, | |
| "loss": 0.742, | |
| "step": 4045 | |
| }, | |
| { | |
| "epoch": 9.225512528473804, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 3.6321751614261767e-06, | |
| "loss": 0.7449, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 9.236902050113896, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 3.5267598083673304e-06, | |
| "loss": 0.7389, | |
| "step": 4055 | |
| }, | |
| { | |
| "epoch": 9.248291571753986, | |
| "grad_norm": 0.375, | |
| "learning_rate": 3.4228693223856136e-06, | |
| "loss": 0.7377, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 9.259681093394077, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.320505345585945e-06, | |
| "loss": 0.7356, | |
| "step": 4065 | |
| }, | |
| { | |
| "epoch": 9.27107061503417, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 3.219669495945055e-06, | |
| "loss": 0.7484, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 9.28246013667426, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.120363367285917e-06, | |
| "loss": 0.7398, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 9.29384965831435, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.022588529252579e-06, | |
| "loss": 0.736, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 9.305239179954443, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.9263465272853173e-06, | |
| "loss": 0.7336, | |
| "step": 4085 | |
| }, | |
| { | |
| "epoch": 9.316628701594533, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.8316388825962324e-06, | |
| "loss": 0.7402, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 9.328018223234624, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 2.738467092145214e-06, | |
| "loss": 0.7427, | |
| "step": 4095 | |
| }, | |
| { | |
| "epoch": 9.339407744874716, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.646832628616214e-06, | |
| "loss": 0.7395, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 9.350797266514807, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 2.5567369403940776e-06, | |
| "loss": 0.7432, | |
| "step": 4105 | |
| }, | |
| { | |
| "epoch": 9.362186788154897, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.4681814515415404e-06, | |
| "loss": 0.7375, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 9.373576309794988, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.3811675617768204e-06, | |
| "loss": 0.7441, | |
| "step": 4115 | |
| }, | |
| { | |
| "epoch": 9.38496583143508, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.2956966464514175e-06, | |
| "loss": 0.7369, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 9.39635535307517, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 2.2117700565283838e-06, | |
| "loss": 0.734, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 9.407744874715261, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 2.1293891185610204e-06, | |
| "loss": 0.7374, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 9.419134396355354, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 2.04855513467187e-06, | |
| "loss": 0.7387, | |
| "step": 4135 | |
| }, | |
| { | |
| "epoch": 9.430523917995444, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.969269382532113e-06, | |
| "loss": 0.7385, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 9.441913439635535, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.8915331153414262e-06, | |
| "loss": 0.7313, | |
| "step": 4145 | |
| }, | |
| { | |
| "epoch": 9.453302961275627, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 1.8153475618081673e-06, | |
| "loss": 0.7394, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 9.464692482915718, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.7407139261299e-06, | |
| "loss": 0.736, | |
| "step": 4155 | |
| }, | |
| { | |
| "epoch": 9.476082004555808, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.667633387974421e-06, | |
| "loss": 0.7403, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 9.4874715261959, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.5961071024610752e-06, | |
| "loss": 0.7478, | |
| "step": 4165 | |
| }, | |
| { | |
| "epoch": 9.498861047835991, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 1.5261362001425138e-06, | |
| "loss": 0.7342, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 9.510250569476081, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 1.457721786986821e-06, | |
| "loss": 0.7485, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 9.521640091116174, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.3908649443600707e-06, | |
| "loss": 0.7392, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 9.533029612756264, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.3255667290091644e-06, | |
| "loss": 0.7362, | |
| "step": 4185 | |
| }, | |
| { | |
| "epoch": 9.544419134396355, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.2618281730451432e-06, | |
| "loss": 0.7383, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 9.555808656036447, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 1.1996502839269453e-06, | |
| "loss": 0.7361, | |
| "step": 4195 | |
| }, | |
| { | |
| "epoch": 9.567198177676538, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.139034044445375e-06, | |
| "loss": 0.7354, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 9.578587699316628, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.0799804127076707e-06, | |
| "loss": 0.7354, | |
| "step": 4205 | |
| }, | |
| { | |
| "epoch": 9.589977220956719, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.0224903221222938e-06, | |
| "loss": 0.7408, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 9.601366742596811, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 9.665646813842077e-07, | |
| "loss": 0.7329, | |
| "step": 4215 | |
| }, | |
| { | |
| "epoch": 9.612756264236902, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 9.12204374460468e-07, | |
| "loss": 0.729, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 9.624145785876994, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 8.59410260576321e-07, | |
| "loss": 0.734, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 9.635535307517085, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 8.081831742015822e-07, | |
| "loss": 0.7391, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 9.646924829157175, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 7.585239250374243e-07, | |
| "loss": 0.7468, | |
| "step": 4235 | |
| }, | |
| { | |
| "epoch": 9.658314350797266, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 7.104332980036211e-07, | |
| "loss": 0.7347, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 9.669703872437358, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 6.639120532261456e-07, | |
| "loss": 0.7391, | |
| "step": 4245 | |
| }, | |
| { | |
| "epoch": 9.681093394077449, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 6.189609260251139e-07, | |
| "loss": 0.742, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 9.69248291571754, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 5.755806269031827e-07, | |
| "loss": 0.7403, | |
| "step": 4255 | |
| }, | |
| { | |
| "epoch": 9.703872437357631, | |
| "grad_norm": 0.375, | |
| "learning_rate": 5.337718415343362e-07, | |
| "loss": 0.7401, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 9.715261958997722, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.935352307530062e-07, | |
| "loss": 0.7366, | |
| "step": 4265 | |
| }, | |
| { | |
| "epoch": 9.726651480637813, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 4.548714305436685e-07, | |
| "loss": 0.7305, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 9.738041002277905, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 4.1778105203078565e-07, | |
| "loss": 0.7419, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 9.749430523917995, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 3.822646814691244e-07, | |
| "loss": 0.735, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 9.760820045558086, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 3.483228802344973e-07, | |
| "loss": 0.7427, | |
| "step": 4285 | |
| }, | |
| { | |
| "epoch": 9.772209567198178, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 3.159561848149029e-07, | |
| "loss": 0.738, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 9.783599088838269, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 2.8516510680203224e-07, | |
| "loss": 0.7443, | |
| "step": 4295 | |
| }, | |
| { | |
| "epoch": 9.79498861047836, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 2.5595013288318703e-07, | |
| "loss": 0.7361, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 9.806378132118452, | |
| "grad_norm": 0.375, | |
| "learning_rate": 2.2831172483359643e-07, | |
| "loss": 0.7446, | |
| "step": 4305 | |
| }, | |
| { | |
| "epoch": 9.817767653758542, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 2.0225031950910078e-07, | |
| "loss": 0.7427, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 9.829157175398633, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.7776632883924615e-07, | |
| "loss": 0.7394, | |
| "step": 4315 | |
| }, | |
| { | |
| "epoch": 9.840546697038725, | |
| "grad_norm": 0.375, | |
| "learning_rate": 1.548601398208116e-07, | |
| "loss": 0.7364, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 9.851936218678816, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.3353211451161417e-07, | |
| "loss": 0.741, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 9.863325740318906, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 1.1378259002488013e-07, | |
| "loss": 0.733, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 9.874715261958997, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.561187852386022e-08, | |
| "loss": 0.742, | |
| "step": 4335 | |
| }, | |
| { | |
| "epoch": 9.88610478359909, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 7.902026721687828e-08, | |
| "loss": 0.7417, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 9.89749430523918, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 6.400801835286796e-08, | |
| "loss": 0.7369, | |
| "step": 4345 | |
| }, | |
| { | |
| "epoch": 9.90888382687927, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 5.05753692171318e-08, | |
| "loss": 0.7385, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 9.920273348519363, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 3.8722532127677404e-08, | |
| "loss": 0.737, | |
| "step": 4355 | |
| }, | |
| { | |
| "epoch": 9.931662870159453, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 2.844969443178691e-08, | |
| "loss": 0.7355, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 9.943052391799544, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.9757018503119285e-08, | |
| "loss": 0.7433, | |
| "step": 4365 | |
| }, | |
| { | |
| "epoch": 9.954441913439636, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.2644641739101292e-08, | |
| "loss": 0.7328, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 9.965831435079727, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 7.112676558784781e-09, | |
| "loss": 0.743, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 9.977220956719817, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.1612104010370068e-09, | |
| "loss": 0.7305, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 9.98861047835991, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 7.903057231750666e-10, | |
| "loss": 0.7394, | |
| "step": 4385 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.0, | |
| "loss": 0.7376, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.4680683612823486, | |
| "eval_runtime": 0.2341, | |
| "eval_samples_per_second": 42.714, | |
| "eval_steps_per_second": 4.271, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 4390, | |
| "total_flos": 1.3402520949471838e+19, | |
| "train_loss": 1.3662878394941533, | |
| "train_runtime": 10620.987, | |
| "train_samples_per_second": 26.444, | |
| "train_steps_per_second": 0.413 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 4390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 1.3402520949471838e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |