| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.987937273823885, | |
| "eval_steps": 500, | |
| "global_step": 4140, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0024125452352231603, | |
| "grad_norm": 388.0, | |
| "learning_rate": 4.830917874396135e-07, | |
| "loss": 49.9057, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.012062726176115802, | |
| "grad_norm": 368.0, | |
| "learning_rate": 2.4154589371980677e-06, | |
| "loss": 50.2975, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.024125452352231604, | |
| "grad_norm": 336.0, | |
| "learning_rate": 4.830917874396135e-06, | |
| "loss": 50.3862, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03618817852834741, | |
| "grad_norm": 177.0, | |
| "learning_rate": 7.246376811594203e-06, | |
| "loss": 43.802, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04825090470446321, | |
| "grad_norm": 130.0, | |
| "learning_rate": 9.66183574879227e-06, | |
| "loss": 36.945, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06031363088057901, | |
| "grad_norm": 85.0, | |
| "learning_rate": 1.2077294685990338e-05, | |
| "loss": 30.8863, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07237635705669482, | |
| "grad_norm": 25.0, | |
| "learning_rate": 1.4492753623188407e-05, | |
| "loss": 26.5704, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08443908323281062, | |
| "grad_norm": 18.875, | |
| "learning_rate": 1.6908212560386476e-05, | |
| "loss": 25.2905, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09650180940892641, | |
| "grad_norm": 16.5, | |
| "learning_rate": 1.932367149758454e-05, | |
| "loss": 23.3032, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10856453558504221, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 2.173913043478261e-05, | |
| "loss": 21.9795, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12062726176115803, | |
| "grad_norm": 6.5, | |
| "learning_rate": 2.4154589371980676e-05, | |
| "loss": 20.9041, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13268998793727382, | |
| "grad_norm": 5.0, | |
| "learning_rate": 2.6570048309178748e-05, | |
| "loss": 20.1174, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.14475271411338964, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 2.8985507246376814e-05, | |
| "loss": 19.4134, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15681544028950542, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 3.140096618357488e-05, | |
| "loss": 19.0099, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.16887816646562123, | |
| "grad_norm": 6.5625, | |
| "learning_rate": 3.381642512077295e-05, | |
| "loss": 18.6242, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.18094089264173704, | |
| "grad_norm": 11.0, | |
| "learning_rate": 3.6231884057971014e-05, | |
| "loss": 17.8084, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19300361881785283, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 3.864734299516908e-05, | |
| "loss": 17.1619, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.20506634499396864, | |
| "grad_norm": 19.875, | |
| "learning_rate": 4.106280193236715e-05, | |
| "loss": 15.4623, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.21712907117008443, | |
| "grad_norm": 31.0, | |
| "learning_rate": 4.347826086956522e-05, | |
| "loss": 12.9656, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.22919179734620024, | |
| "grad_norm": 31.625, | |
| "learning_rate": 4.589371980676328e-05, | |
| "loss": 9.1598, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.24125452352231605, | |
| "grad_norm": 23.25, | |
| "learning_rate": 4.830917874396135e-05, | |
| "loss": 5.018, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.25331724969843183, | |
| "grad_norm": 6.09375, | |
| "learning_rate": 5.072463768115943e-05, | |
| "loss": 2.6562, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.26537997587454765, | |
| "grad_norm": 2.5, | |
| "learning_rate": 5.3140096618357496e-05, | |
| "loss": 2.0877, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.27744270205066346, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 5.555555555555556e-05, | |
| "loss": 1.8147, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.28950542822677927, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 5.797101449275363e-05, | |
| "loss": 1.646, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.30156815440289503, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 6.0386473429951696e-05, | |
| "loss": 1.5332, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.31363088057901084, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 6.280193236714976e-05, | |
| "loss": 1.4679, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.32569360675512665, | |
| "grad_norm": 0.6015625, | |
| "learning_rate": 6.521739130434783e-05, | |
| "loss": 1.4093, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.33775633293124246, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 6.76328502415459e-05, | |
| "loss": 1.3534, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3498190591073583, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 7.004830917874396e-05, | |
| "loss": 1.3102, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3618817852834741, | |
| "grad_norm": 1.0, | |
| "learning_rate": 7.246376811594203e-05, | |
| "loss": 1.2869, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.37394451145958985, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 7.48792270531401e-05, | |
| "loss": 1.2508, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.38600723763570566, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 7.729468599033817e-05, | |
| "loss": 1.2457, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.39806996381182147, | |
| "grad_norm": 0.99609375, | |
| "learning_rate": 7.971014492753623e-05, | |
| "loss": 1.2096, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4101326899879373, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 8.21256038647343e-05, | |
| "loss": 1.1968, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4221954161640531, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 8.454106280193237e-05, | |
| "loss": 1.1807, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.43425814234016885, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 8.695652173913044e-05, | |
| "loss": 1.167, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.44632086851628466, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 8.937198067632851e-05, | |
| "loss": 1.173, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4583835946924005, | |
| "grad_norm": 5.8125, | |
| "learning_rate": 9.178743961352657e-05, | |
| "loss": 1.1547, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4704463208685163, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 9.420289855072463e-05, | |
| "loss": 1.1475, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4825090470446321, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 9.66183574879227e-05, | |
| "loss": 1.1483, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4945717732207479, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 9.903381642512077e-05, | |
| "loss": 1.1399, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.5066344993968637, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 0.00010144927536231885, | |
| "loss": 1.1287, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.5186972255729795, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 0.00010386473429951691, | |
| "loss": 1.1176, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.5307599517490953, | |
| "grad_norm": 3.171875, | |
| "learning_rate": 0.00010628019323671499, | |
| "loss": 1.1057, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.5428226779252111, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 0.00010869565217391305, | |
| "loss": 1.1042, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.5548854041013269, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 1.091, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5669481302774427, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 0.00011352657004830917, | |
| "loss": 1.0847, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5790108564535585, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 0.00011594202898550725, | |
| "loss": 1.0686, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5910735826296744, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 0.00011835748792270531, | |
| "loss": 1.0752, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.6031363088057901, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 0.00012077294685990339, | |
| "loss": 1.06, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.6151990349819059, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 0.00012318840579710145, | |
| "loss": 1.0583, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.6272617611580217, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 0.00012560386473429953, | |
| "loss": 1.0598, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.6393244873341375, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 0.00012801932367149759, | |
| "loss": 1.0482, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.6513872135102533, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 0.00013043478260869567, | |
| "loss": 1.0542, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.6634499396863691, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 0.00013285024154589372, | |
| "loss": 1.0398, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6755126658624849, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 0.0001352657004830918, | |
| "loss": 1.0327, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6875753920386007, | |
| "grad_norm": 3.140625, | |
| "learning_rate": 0.00013768115942028986, | |
| "loss": 1.0287, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6996381182147166, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 0.00014009661835748792, | |
| "loss": 1.035, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.7117008443908324, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 0.000142512077294686, | |
| "loss": 1.0149, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.7237635705669482, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 0.00014492753623188405, | |
| "loss": 1.0279, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.7358262967430639, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 0.00014734299516908214, | |
| "loss": 1.0261, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.7478890229191797, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 0.0001497584541062802, | |
| "loss": 1.0113, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.7599517490952955, | |
| "grad_norm": 4.03125, | |
| "learning_rate": 0.00015217391304347827, | |
| "loss": 1.0244, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.7720144752714113, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.00015458937198067633, | |
| "loss": 1.0105, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.7840772014475271, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.0001570048309178744, | |
| "loss": 1.005, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7961399276236429, | |
| "grad_norm": 5.6875, | |
| "learning_rate": 0.00015942028985507247, | |
| "loss": 1.0101, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.8082026537997588, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 0.00016183574879227055, | |
| "loss": 1.0173, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.8202653799758746, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 0.0001642512077294686, | |
| "loss": 1.0132, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.8323281061519904, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 1.0205, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.8443908323281062, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 0.00016908212560386474, | |
| "loss": 1.0074, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.856453558504222, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 0.00017149758454106283, | |
| "loss": 0.9998, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.8685162846803377, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 0.00017391304347826088, | |
| "loss": 0.9886, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.8805790108564535, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 0.00017632850241545896, | |
| "loss": 0.9902, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.8926417370325693, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 0.00017874396135265702, | |
| "loss": 1.0085, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.9047044632086851, | |
| "grad_norm": 5.5, | |
| "learning_rate": 0.00018115942028985507, | |
| "loss": 1.0021, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.916767189384801, | |
| "grad_norm": 8.875, | |
| "learning_rate": 0.00018357487922705313, | |
| "loss": 1.0323, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.9288299155609168, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 0.0001859903381642512, | |
| "loss": 1.0051, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.9408926417370326, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 0.00018840579710144927, | |
| "loss": 1.0105, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.9529553679131484, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 0.00019082125603864735, | |
| "loss": 1.0062, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.9650180940892642, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 0.0001932367149758454, | |
| "loss": 0.9982, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.97708082026538, | |
| "grad_norm": 1.125, | |
| "learning_rate": 0.0001956521739130435, | |
| "loss": 0.9794, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.9891435464414958, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 0.00019806763285024154, | |
| "loss": 0.9973, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.9987937273823885, | |
| "eval_loss": 2.522090435028076, | |
| "eval_runtime": 0.3154, | |
| "eval_samples_per_second": 31.706, | |
| "eval_steps_per_second": 3.171, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.0012062726176116, | |
| "grad_norm": 2.546875, | |
| "learning_rate": 0.00019999996445454953, | |
| "loss": 0.9904, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.0132689987937273, | |
| "grad_norm": 4.5, | |
| "learning_rate": 0.00019999872036643513, | |
| "loss": 1.0068, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.0253317249698433, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 0.00019999569903106498, | |
| "loss": 0.9834, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.037394451145959, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 0.00019999090050213636, | |
| "loss": 0.9873, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.0494571773220749, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 0.00019998432486493205, | |
| "loss": 0.9593, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.0615199034981906, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 0.00019997597223631895, | |
| "loss": 0.9534, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.0735826296743065, | |
| "grad_norm": 10.9375, | |
| "learning_rate": 0.0001999658427647457, | |
| "loss": 0.9689, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.0856453558504222, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 0.00019995393663024054, | |
| "loss": 0.9786, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.097708082026538, | |
| "grad_norm": 6.15625, | |
| "learning_rate": 0.00019994025404440743, | |
| "loss": 0.9602, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.1097708082026538, | |
| "grad_norm": 4.0, | |
| "learning_rate": 0.00019992479525042303, | |
| "loss": 0.9507, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.1218335343787695, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 0.00019990756052303173, | |
| "loss": 0.9581, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 1.1338962605548855, | |
| "grad_norm": 2.5, | |
| "learning_rate": 0.0001998885501685412, | |
| "loss": 0.9755, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.1459589867310012, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.0001998677645248168, | |
| "loss": 0.9557, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.158021712907117, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 0.00019984520396127553, | |
| "loss": 0.975, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.1700844390832328, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 0.00019982086887887948, | |
| "loss": 0.9619, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 1.1821471652593487, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 0.0001997947597101288, | |
| "loss": 0.9619, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.1942098914354644, | |
| "grad_norm": 1.875, | |
| "learning_rate": 0.00019976687691905393, | |
| "loss": 0.9504, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 1.2062726176115803, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 0.0001997372210012073, | |
| "loss": 0.9357, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.218335343787696, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 0.00019970579248365468, | |
| "loss": 0.9435, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.2303980699638117, | |
| "grad_norm": 5.125, | |
| "learning_rate": 0.0001996725919249657, | |
| "loss": 0.9404, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.2424607961399277, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.00019963761991520387, | |
| "loss": 0.9383, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.2545235223160434, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 0.00019960087707591626, | |
| "loss": 0.9493, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.2665862484921593, | |
| "grad_norm": 3.125, | |
| "learning_rate": 0.00019956236406012232, | |
| "loss": 0.9341, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.278648974668275, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 0.00019952208155230234, | |
| "loss": 0.9472, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.290711700844391, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 0.00019948003026838525, | |
| "loss": 0.9308, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.3027744270205066, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 0.00019943621095573586, | |
| "loss": 0.9416, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.3148371531966223, | |
| "grad_norm": 2.875, | |
| "learning_rate": 0.00019939062439314174, | |
| "loss": 0.9424, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.3268998793727382, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 0.00019934327139079915, | |
| "loss": 0.9347, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.3389626055488542, | |
| "grad_norm": 14.125, | |
| "learning_rate": 0.00019929415279029873, | |
| "loss": 0.9494, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.3510253317249699, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 0.00019924326946461074, | |
| "loss": 0.9426, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.3630880579010856, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 0.00019919062231806922, | |
| "loss": 0.928, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.3751507840772015, | |
| "grad_norm": 3.703125, | |
| "learning_rate": 0.0001991362122863561, | |
| "loss": 0.942, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.3872135102533172, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 0.00019908004033648453, | |
| "loss": 0.9355, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.399276236429433, | |
| "grad_norm": 9.375, | |
| "learning_rate": 0.0001990221074667818, | |
| "loss": 0.9308, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.4113389626055488, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 0.0001989624147068713, | |
| "loss": 0.9226, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.4234016887816647, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 0.00019890096311765465, | |
| "loss": 0.9179, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.4354644149577804, | |
| "grad_norm": 2.0, | |
| "learning_rate": 0.00019883775379129248, | |
| "loss": 0.9161, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.4475271411338961, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 0.00019877278785118517, | |
| "loss": 0.9137, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.459589867310012, | |
| "grad_norm": 2.765625, | |
| "learning_rate": 0.0001987060664519529, | |
| "loss": 0.909, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.471652593486128, | |
| "grad_norm": 4.9375, | |
| "learning_rate": 0.00019863759077941504, | |
| "loss": 0.9055, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.4837153196622437, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 0.0001985673620505692, | |
| "loss": 0.9221, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.4957780458383594, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 0.00019849538151356955, | |
| "loss": 0.9249, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.5078407720144753, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.00019842165044770452, | |
| "loss": 0.9188, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.5199034981905912, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.0001983461701633742, | |
| "loss": 0.9134, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.531966224366707, | |
| "grad_norm": 2.6875, | |
| "learning_rate": 0.00019826894200206715, | |
| "loss": 0.9084, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.5440289505428226, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 0.00019818996733633618, | |
| "loss": 0.9136, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.5560916767189386, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 0.00019810924756977444, | |
| "loss": 0.9372, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.5681544028950543, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 0.00019802678413699006, | |
| "loss": 0.8959, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.58021712907117, | |
| "grad_norm": 3.578125, | |
| "learning_rate": 0.0001979425785035809, | |
| "loss": 0.8991, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.5922798552472859, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 0.00019785663216610844, | |
| "loss": 0.9074, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.6043425814234018, | |
| "grad_norm": 3.046875, | |
| "learning_rate": 0.00019776894665207113, | |
| "loss": 0.9085, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.6164053075995175, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 0.0001976795235198773, | |
| "loss": 0.9085, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.6284680337756332, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 0.00019758836435881746, | |
| "loss": 0.9026, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.6405307599517491, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 0.00019749547078903604, | |
| "loss": 0.9033, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.652593486127865, | |
| "grad_norm": 15.25, | |
| "learning_rate": 0.00019740084446150257, | |
| "loss": 0.8952, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.6646562123039808, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 0.921, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.6767189384800965, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 0.00019720640029100673, | |
| "loss": 0.9014, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.6887816646562124, | |
| "grad_norm": 0.859375, | |
| "learning_rate": 0.00019710658590384227, | |
| "loss": 0.8948, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.700844390832328, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.00019700504567046013, | |
| "loss": 0.8953, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.7129071170084438, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 0.00019690178139550443, | |
| "loss": 0.8933, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.7249698431845597, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.0001967967949142602, | |
| "loss": 0.9286, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.7370325693606756, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 0.00019669008809262062, | |
| "loss": 0.8984, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.7490952955367913, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 0.00019658166282705405, | |
| "loss": 0.9069, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.761158021712907, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 0.00019647152104457013, | |
| "loss": 0.8887, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.773220747889023, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 0.00019635966470268583, | |
| "loss": 0.8827, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.7852834740651389, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 0.00019624609578939027, | |
| "loss": 0.8757, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.7973462002412546, | |
| "grad_norm": 4.78125, | |
| "learning_rate": 0.0001961308163231097, | |
| "loss": 0.8809, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.8094089264173703, | |
| "grad_norm": 1.25, | |
| "learning_rate": 0.0001960138283526715, | |
| "loss": 0.8817, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.8214716525934862, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 0.00019589513395726777, | |
| "loss": 0.881, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.833534378769602, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 0.0001957747352464184, | |
| "loss": 0.8839, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.8455971049457176, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.0001956526343599335, | |
| "loss": 0.8854, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.8576598311218335, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.00019552883346787552, | |
| "loss": 0.8799, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.8697225572979495, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 0.0001954033347705206, | |
| "loss": 0.873, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.8817852834740652, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 0.0001952761404983194, | |
| "loss": 0.864, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.8938480096501809, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.00019514725291185762, | |
| "loss": 0.8806, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.9059107358262968, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 0.0001950166743018156, | |
| "loss": 0.8879, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.9179734620024127, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 0.0001948844069889278, | |
| "loss": 0.892, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.9300361881785284, | |
| "grad_norm": 16.875, | |
| "learning_rate": 0.00019475045332394153, | |
| "loss": 0.8922, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.942098914354644, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 0.00019461481568757506, | |
| "loss": 0.876, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.95416164053076, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 0.00019447749649047542, | |
| "loss": 0.8787, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.9662243667068757, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00019433849817317552, | |
| "loss": 0.8625, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.9782870928829914, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 0.0001941978232060507, | |
| "loss": 0.8643, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.9903498190591074, | |
| "grad_norm": 15.75, | |
| "learning_rate": 0.00019405547408927502, | |
| "loss": 0.9036, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.435774326324463, | |
| "eval_runtime": 0.2379, | |
| "eval_samples_per_second": 42.031, | |
| "eval_steps_per_second": 4.203, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 2.0024125452352233, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 0.00019391145335277655, | |
| "loss": 0.8698, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.0144752714113388, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 0.0001937657635561927, | |
| "loss": 0.8737, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 2.0265379975874547, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 0.00019361840728882447, | |
| "loss": 0.8449, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.0386007237635706, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.0001934693871695906, | |
| "loss": 0.853, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 2.0506634499396865, | |
| "grad_norm": 9.375, | |
| "learning_rate": 0.00019331870584698093, | |
| "loss": 0.8422, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.062726176115802, | |
| "grad_norm": 4.4375, | |
| "learning_rate": 0.00019316636599900946, | |
| "loss": 0.8357, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 2.074788902291918, | |
| "grad_norm": 4.25, | |
| "learning_rate": 0.00019301237033316659, | |
| "loss": 0.8618, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.086851628468034, | |
| "grad_norm": 6.125, | |
| "learning_rate": 0.00019285672158637104, | |
| "loss": 0.8538, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 2.0989143546441498, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 0.00019269942252492133, | |
| "loss": 0.8476, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.1109770808202653, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 0.0001925404759444465, | |
| "loss": 0.8442, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.123039806996381, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.0001923798846698564, | |
| "loss": 0.8363, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.135102533172497, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 0.00019221765155529158, | |
| "loss": 0.8512, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 2.147165259348613, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 0.00019205377948407258, | |
| "loss": 0.8464, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.1592279855247285, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 0.00019188827136864842, | |
| "loss": 0.8341, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 2.1712907117008444, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 0.00019172113015054532, | |
| "loss": 0.837, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.1833534378769603, | |
| "grad_norm": 3.421875, | |
| "learning_rate": 0.0001915523588003139, | |
| "loss": 0.8768, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 2.195416164053076, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 0.00019138196031747681, | |
| "loss": 0.8494, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.2074788902291917, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 0.00019120993773047513, | |
| "loss": 0.8934, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 2.2195416164053077, | |
| "grad_norm": 0.80859375, | |
| "learning_rate": 0.0001910362940966147, | |
| "loss": 0.8258, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.2316043425814236, | |
| "grad_norm": 6.5, | |
| "learning_rate": 0.00019086103250201165, | |
| "loss": 0.8256, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 2.243667068757539, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.00019068415606153787, | |
| "loss": 0.8347, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.255729794933655, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 0.00019050566791876516, | |
| "loss": 0.8264, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 2.267792521109771, | |
| "grad_norm": 0.90625, | |
| "learning_rate": 0.00019032557124590974, | |
| "loss": 0.8232, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.2798552472858864, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 0.00019014386924377582, | |
| "loss": 0.8218, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 2.2919179734620023, | |
| "grad_norm": 0.69921875, | |
| "learning_rate": 0.00018996056514169844, | |
| "loss": 0.8146, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.3039806996381182, | |
| "grad_norm": 13.4375, | |
| "learning_rate": 0.00018977566219748653, | |
| "loss": 0.817, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 2.316043425814234, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 0.0001895891636973646, | |
| "loss": 0.8263, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 2.3281061519903496, | |
| "grad_norm": 13.0, | |
| "learning_rate": 0.00018940107295591454, | |
| "loss": 0.8203, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 2.3401688781664656, | |
| "grad_norm": 3.640625, | |
| "learning_rate": 0.00018921139331601667, | |
| "loss": 0.8183, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 2.3522316043425815, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.0001890201281487903, | |
| "loss": 0.8013, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 2.3642943305186974, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00018882728085353392, | |
| "loss": 0.8068, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 2.376357056694813, | |
| "grad_norm": 6.0625, | |
| "learning_rate": 0.0001886328548576646, | |
| "loss": 0.801, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 2.388419782870929, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.00018843685361665723, | |
| "loss": 0.8269, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 2.4004825090470447, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 0.00018823928061398313, | |
| "loss": 0.8141, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 2.4125452352231607, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00018804013936104792, | |
| "loss": 0.8157, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.424607961399276, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00018783943339712938, | |
| "loss": 0.8016, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 2.436670687575392, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00018763716628931437, | |
| "loss": 0.7957, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.448733413751508, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.0001874333416324356, | |
| "loss": 0.8014, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 2.4607961399276235, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 0.0001872279630490074, | |
| "loss": 0.8054, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.4728588661037394, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 0.00018702103418916183, | |
| "loss": 0.7972, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 2.4849215922798553, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 0.00018681255873058338, | |
| "loss": 0.8115, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.4969843184559712, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 0.00018660254037844388, | |
| "loss": 0.8134, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 2.5090470446320867, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 0.00018639098286533644, | |
| "loss": 0.8286, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.5211097708082026, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 0.00018617788995120931, | |
| "loss": 0.804, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 2.5331724969843186, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 0.00018596326542329888, | |
| "loss": 0.8048, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.545235223160434, | |
| "grad_norm": 0.87109375, | |
| "learning_rate": 0.00018574711309606249, | |
| "loss": 0.7941, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 2.55729794933655, | |
| "grad_norm": 0.7578125, | |
| "learning_rate": 0.00018552943681111067, | |
| "loss": 0.795, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.569360675512666, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 0.00018531024043713868, | |
| "loss": 0.7905, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 2.581423401688782, | |
| "grad_norm": 3.765625, | |
| "learning_rate": 0.0001850895278698579, | |
| "loss": 0.793, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.5934861278648977, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 0.00018486730303192658, | |
| "loss": 0.7986, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.605548854041013, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00018464356987288013, | |
| "loss": 0.7941, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.617611580217129, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 0.0001844183323690608, | |
| "loss": 0.7904, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 2.6296743063932446, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 0.0001841915945235472, | |
| "loss": 0.7932, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.6417370325693605, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00018396336036608307, | |
| "loss": 0.7942, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 2.6537997587454765, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 0.00018373363395300554, | |
| "loss": 0.7905, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.6658624849215924, | |
| "grad_norm": 0.69140625, | |
| "learning_rate": 0.0001835024193671733, | |
| "loss": 0.7929, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 2.6779252110977083, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.0001832697207178938, | |
| "loss": 0.7877, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.689987937273824, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 0.0001830355421408503, | |
| "loss": 0.7909, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 2.7020506634499397, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 0.00018279988779802833, | |
| "loss": 0.7815, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.7141133896260556, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00018256276187764197, | |
| "loss": 0.7829, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.726176115802171, | |
| "grad_norm": 0.75, | |
| "learning_rate": 0.00018232416859405895, | |
| "loss": 0.7816, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.738238841978287, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00018208411218772615, | |
| "loss": 0.7836, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 2.750301568154403, | |
| "grad_norm": 0.98828125, | |
| "learning_rate": 0.00018184259692509406, | |
| "loss": 0.785, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.762364294330519, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 0.000181599627098541, | |
| "loss": 0.7771, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 2.7744270205066344, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00018135520702629675, | |
| "loss": 0.7808, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.7864897466827503, | |
| "grad_norm": 0.76171875, | |
| "learning_rate": 0.000181109341052366, | |
| "loss": 0.7756, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 2.798552472858866, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.00018086203354645089, | |
| "loss": 0.7838, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.8106151990349817, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00018061328890387352, | |
| "loss": 0.7856, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 2.8226779252110976, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 0.00018036311154549784, | |
| "loss": 0.7889, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.8347406513872135, | |
| "grad_norm": 0.953125, | |
| "learning_rate": 0.00018011150591765092, | |
| "loss": 0.7835, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.8468033775633295, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 0.00017985847649204417, | |
| "loss": 0.7874, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.8588661037394454, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 0.0001796040277656936, | |
| "loss": 0.7743, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 2.870928829915561, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 0.00017934816426084008, | |
| "loss": 0.7652, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.882991556091677, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.00017909089052486898, | |
| "loss": 0.7671, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 2.8950542822677923, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 0.00017883221113022916, | |
| "loss": 0.7769, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.907117008443908, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 0.00017857213067435195, | |
| "loss": 0.7669, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 2.919179734620024, | |
| "grad_norm": 0.8984375, | |
| "learning_rate": 0.0001783106537795692, | |
| "loss": 0.7656, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.93124246079614, | |
| "grad_norm": 0.85546875, | |
| "learning_rate": 0.00017804778509303138, | |
| "loss": 0.7717, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 2.943305186972256, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00017778352928662474, | |
| "loss": 0.7809, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.9553679131483714, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00017751789105688846, | |
| "loss": 0.7736, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.9674306393244874, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 0.0001772508751249311, | |
| "loss": 0.7744, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.9794933655006033, | |
| "grad_norm": 0.6875, | |
| "learning_rate": 0.0001769824862363467, | |
| "loss": 0.759, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 2.9915560916767188, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 0.00017671272916113052, | |
| "loss": 0.7651, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.9987937273823886, | |
| "eval_loss": 2.3987040519714355, | |
| "eval_runtime": 0.2575, | |
| "eval_samples_per_second": 38.831, | |
| "eval_steps_per_second": 3.883, | |
| "step": 1243 | |
| }, | |
| { | |
| "epoch": 3.0036188178528347, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 0.00017644160869359404, | |
| "loss": 0.7612, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 3.0156815440289506, | |
| "grad_norm": 11.375, | |
| "learning_rate": 0.00017616912965228001, | |
| "loss": 0.7436, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 3.0277442702050665, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 0.00017589529687987674, | |
| "loss": 0.7548, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 3.039806996381182, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 0.00017562011524313185, | |
| "loss": 0.7489, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 3.051869722557298, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00017534358963276607, | |
| "loss": 0.7469, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 3.063932448733414, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 0.0001750657249633861, | |
| "loss": 0.7501, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 3.0759951749095293, | |
| "grad_norm": 1.375, | |
| "learning_rate": 0.00017478652617339738, | |
| "loss": 0.7503, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 3.0880579010856453, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00017450599822491615, | |
| "loss": 0.7587, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 3.100120627261761, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00017422414610368157, | |
| "loss": 0.7483, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 3.112183353437877, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.00017394097481896676, | |
| "loss": 0.7554, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 3.1242460796139926, | |
| "grad_norm": 0.435546875, | |
| "learning_rate": 0.00017365648940349004, | |
| "loss": 0.7484, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 3.1363088057901085, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 0.00017337069491332537, | |
| "loss": 0.7389, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.1483715319662244, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00017308359642781242, | |
| "loss": 0.7471, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 3.1604342581423404, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00017279519904946647, | |
| "loss": 0.7469, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 3.172496984318456, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 0.00017250550790388764, | |
| "loss": 0.7525, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 3.1845597104945718, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 0.0001722145281396697, | |
| "loss": 0.755, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 3.1966224366706877, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.0001719222649283087, | |
| "loss": 0.7463, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 3.2086851628468036, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.00017162872346411102, | |
| "loss": 0.7443, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 3.220747889022919, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00017133390896410106, | |
| "loss": 0.7454, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 3.232810615199035, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00017103782666792844, | |
| "loss": 0.7459, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 3.244873341375151, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 0.00017074048183777504, | |
| "loss": 0.7454, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 3.2569360675512664, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 0.00017044187975826124, | |
| "loss": 0.7462, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 3.2689987937273823, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00017014202573635228, | |
| "loss": 0.7441, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 3.2810615199034983, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00016984092510126367, | |
| "loss": 0.7378, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 3.293124246079614, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 0.00016953858320436672, | |
| "loss": 0.7529, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 3.3051869722557297, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 0.0001692350054190932, | |
| "loss": 0.7426, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 3.3172496984318456, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 0.00016893019714084008, | |
| "loss": 0.7414, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 3.3293124246079615, | |
| "grad_norm": 0.7421875, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 0.7461, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 3.341375150784077, | |
| "grad_norm": 0.6640625, | |
| "learning_rate": 0.0001683169107962321, | |
| "loss": 0.7418, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 3.353437876960193, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 0.00016800844362963147, | |
| "loss": 0.7371, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 3.365500603136309, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.0001676987677693659, | |
| "loss": 0.736, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 3.3775633293124248, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.00016738788871921152, | |
| "loss": 0.738, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 3.3896260554885402, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00016707581200432845, | |
| "loss": 0.7458, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 3.401688781664656, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00016676254317116252, | |
| "loss": 0.7379, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 3.413751507840772, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.00016644808778734668, | |
| "loss": 0.7328, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 3.425814234016888, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.0001661324514416022, | |
| "loss": 0.7389, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 3.4378769601930035, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00016581563974363902, | |
| "loss": 0.7339, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 3.4499396863691194, | |
| "grad_norm": 0.75, | |
| "learning_rate": 0.00016549765832405653, | |
| "loss": 0.7338, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 3.4620024125452353, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00016517851283424304, | |
| "loss": 0.7367, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 3.4740651387213513, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.0001648582089462756, | |
| "loss": 0.7476, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 3.4861278648974667, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00016453675235281913, | |
| "loss": 0.7285, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 3.4981905910735827, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00016421414876702518, | |
| "loss": 0.7326, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 3.5102533172496986, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00016389040392243056, | |
| "loss": 0.7314, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 3.522316043425814, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00016356552357285522, | |
| "loss": 0.7326, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 3.53437876960193, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 0.00016323951349230017, | |
| "loss": 0.7269, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 3.546441495778046, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.0001629123794748447, | |
| "loss": 0.7265, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 3.558504221954162, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.0001625841273345436, | |
| "loss": 0.734, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 3.5705669481302773, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00016225476290532374, | |
| "loss": 0.7316, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 3.5826296743063932, | |
| "grad_norm": 2.734375, | |
| "learning_rate": 0.0001619242920408802, | |
| "loss": 0.7232, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 3.594692400482509, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 0.00016159272061457255, | |
| "loss": 0.7239, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 3.6067551266586246, | |
| "grad_norm": 0.60546875, | |
| "learning_rate": 0.0001612600545193203, | |
| "loss": 0.7207, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 3.6188178528347406, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 0.0001609262996674981, | |
| "loss": 0.7236, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.6308805790108565, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00016059146199083087, | |
| "loss": 0.736, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 3.6429433051869724, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.0001602555474402881, | |
| "loss": 0.7294, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 3.6550060313630883, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 0.0001599185619859784, | |
| "loss": 0.7288, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 3.667068757539204, | |
| "grad_norm": 4.3125, | |
| "learning_rate": 0.00015958051161704307, | |
| "loss": 0.7411, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.6791314837153197, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.0001592414023415499, | |
| "loss": 0.7431, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 3.691194209891435, | |
| "grad_norm": 3.75, | |
| "learning_rate": 0.00015890124018638638, | |
| "loss": 0.7379, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 3.703256936067551, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00015856003119715242, | |
| "loss": 0.7388, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 3.715319662243667, | |
| "grad_norm": 0.49609375, | |
| "learning_rate": 0.00015821778143805296, | |
| "loss": 0.7282, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.727382388419783, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 0.00015787449699179035, | |
| "loss": 0.73, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 3.739445114595899, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00015753018395945598, | |
| "loss": 0.7288, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 3.7515078407720144, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 0.00015718484846042214, | |
| "loss": 0.7296, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 3.7635705669481303, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00015683849663223308, | |
| "loss": 0.7361, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 3.7756332931242462, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.00015649113463049586, | |
| "loss": 0.7303, | |
| "step": 1565 | |
| }, | |
| { | |
| "epoch": 3.7876960193003617, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.00015614276862877113, | |
| "loss": 0.7263, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.7997587454764776, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.00015579340481846336, | |
| "loss": 0.7193, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 3.8118214716525936, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 0.0001554430494087107, | |
| "loss": 0.7251, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 3.8238841978287095, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.00015509170862627476, | |
| "loss": 0.7284, | |
| "step": 1585 | |
| }, | |
| { | |
| "epoch": 3.835946924004825, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.00015473938871542986, | |
| "loss": 0.722, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 3.848009650180941, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.00015438609593785202, | |
| "loss": 0.7337, | |
| "step": 1595 | |
| }, | |
| { | |
| "epoch": 3.860072376357057, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 0.00015403183657250788, | |
| "loss": 0.7216, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 3.8721351025331723, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.0001536766169155428, | |
| "loss": 0.7222, | |
| "step": 1605 | |
| }, | |
| { | |
| "epoch": 3.884197828709288, | |
| "grad_norm": 0.8515625, | |
| "learning_rate": 0.00015332044328016914, | |
| "loss": 0.7214, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 3.896260554885404, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 0.0001529633219965541, | |
| "loss": 0.7264, | |
| "step": 1615 | |
| }, | |
| { | |
| "epoch": 3.90832328106152, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.00015260525941170712, | |
| "loss": 0.7232, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.920386007237636, | |
| "grad_norm": 0.703125, | |
| "learning_rate": 0.00015224626188936705, | |
| "loss": 0.7238, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 3.9324487334137515, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 0.00015188633580988926, | |
| "loss": 0.7227, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.9445114595898674, | |
| "grad_norm": 0.427734375, | |
| "learning_rate": 0.00015152548757013182, | |
| "loss": 0.7194, | |
| "step": 1635 | |
| }, | |
| { | |
| "epoch": 3.956574185765983, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00015116372358334233, | |
| "loss": 0.7223, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.9686369119420988, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.00015080105027904362, | |
| "loss": 0.7192, | |
| "step": 1645 | |
| }, | |
| { | |
| "epoch": 3.9806996381182147, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00015043747410291945, | |
| "loss": 0.7206, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.9927623642943306, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00015007300151670015, | |
| "loss": 0.7192, | |
| "step": 1655 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.397005558013916, | |
| "eval_runtime": 0.2387, | |
| "eval_samples_per_second": 41.901, | |
| "eval_steps_per_second": 4.19, | |
| "step": 1658 | |
| }, | |
| { | |
| "epoch": 4.0048250904704465, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.00014970763899804763, | |
| "loss": 0.7131, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 4.0168878166465625, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00014934139304044033, | |
| "loss": 0.6955, | |
| "step": 1665 | |
| }, | |
| { | |
| "epoch": 4.0289505428226775, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 0.0001489742701530578, | |
| "loss": 0.6989, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 4.041013268998793, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.0001486062768606649, | |
| "loss": 0.6997, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 4.053075995174909, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 0.00014823741970349606, | |
| "loss": 0.7021, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 4.065138721351025, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 0.0001478677052371389, | |
| "loss": 0.7052, | |
| "step": 1685 | |
| }, | |
| { | |
| "epoch": 4.077201447527141, | |
| "grad_norm": 0.828125, | |
| "learning_rate": 0.0001474971400324177, | |
| "loss": 0.7003, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 4.089264173703257, | |
| "grad_norm": 0.65234375, | |
| "learning_rate": 0.00014712573067527664, | |
| "loss": 0.7057, | |
| "step": 1695 | |
| }, | |
| { | |
| "epoch": 4.101326899879373, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00014675348376666278, | |
| "loss": 0.7061, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 4.113389626055489, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00014638040592240877, | |
| "loss": 0.6999, | |
| "step": 1705 | |
| }, | |
| { | |
| "epoch": 4.125452352231604, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 0.00014600650377311522, | |
| "loss": 0.6965, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 4.13751507840772, | |
| "grad_norm": 0.8203125, | |
| "learning_rate": 0.00014563178396403284, | |
| "loss": 0.7121, | |
| "step": 1715 | |
| }, | |
| { | |
| "epoch": 4.149577804583836, | |
| "grad_norm": 0.66015625, | |
| "learning_rate": 0.00014525625315494435, | |
| "loss": 0.7075, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 4.161640530759952, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 0.00014487991802004623, | |
| "loss": 0.7059, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 4.173703256936068, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00014450278524782986, | |
| "loss": 0.7055, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 4.185765983112184, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 0.00014412486154096284, | |
| "loss": 0.7039, | |
| "step": 1735 | |
| }, | |
| { | |
| "epoch": 4.1978287092882995, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00014374615361616985, | |
| "loss": 0.6998, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 4.209891435464415, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00014336666820411327, | |
| "loss": 0.7046, | |
| "step": 1745 | |
| }, | |
| { | |
| "epoch": 4.2219541616405305, | |
| "grad_norm": 0.74609375, | |
| "learning_rate": 0.00014298641204927342, | |
| "loss": 0.7002, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 4.234016887816646, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00014260539190982886, | |
| "loss": 0.6991, | |
| "step": 1755 | |
| }, | |
| { | |
| "epoch": 4.246079613992762, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.0001422236145575362, | |
| "loss": 0.6973, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 4.258142340168878, | |
| "grad_norm": 0.58203125, | |
| "learning_rate": 0.00014184108677760984, | |
| "loss": 0.701, | |
| "step": 1765 | |
| }, | |
| { | |
| "epoch": 4.270205066344994, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00014145781536860122, | |
| "loss": 0.6959, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 4.28226779252111, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 0.00014107380714227806, | |
| "loss": 0.6982, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 4.294330518697226, | |
| "grad_norm": 0.4921875, | |
| "learning_rate": 0.00014068906892350343, | |
| "loss": 0.7017, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 4.306393244873341, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 0.00014030360755011424, | |
| "loss": 0.6932, | |
| "step": 1785 | |
| }, | |
| { | |
| "epoch": 4.318455971049457, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.0001399174298727998, | |
| "loss": 0.7008, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 4.330518697225573, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00013953054275498013, | |
| "loss": 0.6944, | |
| "step": 1795 | |
| }, | |
| { | |
| "epoch": 4.342581423401689, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00013914295307268396, | |
| "loss": 0.6977, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 4.354644149577805, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.0001387546677144263, | |
| "loss": 0.7053, | |
| "step": 1805 | |
| }, | |
| { | |
| "epoch": 4.366706875753921, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00013836569358108647, | |
| "loss": 0.7112, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 4.378769601930037, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00013797603758578496, | |
| "loss": 0.7034, | |
| "step": 1815 | |
| }, | |
| { | |
| "epoch": 4.390832328106152, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 0.00013758570665376086, | |
| "loss": 0.7081, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 4.402895054282268, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 0.0001371947077222488, | |
| "loss": 0.7029, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 4.4149577804583835, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 0.00013680304774035538, | |
| "loss": 0.699, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 4.427020506634499, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00013641073366893607, | |
| "loss": 0.7041, | |
| "step": 1835 | |
| }, | |
| { | |
| "epoch": 4.439083232810615, | |
| "grad_norm": 0.412109375, | |
| "learning_rate": 0.00013601777248047105, | |
| "loss": 0.6954, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 4.451145958986731, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00013562417115894172, | |
| "loss": 0.6941, | |
| "step": 1845 | |
| }, | |
| { | |
| "epoch": 4.463208685162847, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 0.0001352299366997062, | |
| "loss": 0.7027, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 4.475271411338962, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.0001348350761093753, | |
| "loss": 0.6979, | |
| "step": 1855 | |
| }, | |
| { | |
| "epoch": 4.487334137515078, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 0.0001344395964056878, | |
| "loss": 0.7028, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 4.499396863691194, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00013404350461738586, | |
| "loss": 0.6999, | |
| "step": 1865 | |
| }, | |
| { | |
| "epoch": 4.51145958986731, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00013364680778409, | |
| "loss": 0.6981, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 4.523522316043426, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00013324951295617398, | |
| "loss": 0.6984, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 4.535585042219542, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 0.00013285162719463961, | |
| "loss": 0.6999, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 4.547647768395658, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00013245315757099118, | |
| "loss": 0.6966, | |
| "step": 1885 | |
| }, | |
| { | |
| "epoch": 4.559710494571773, | |
| "grad_norm": 0.54296875, | |
| "learning_rate": 0.00013205411116710972, | |
| "loss": 0.6985, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 4.571773220747889, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00013165449507512725, | |
| "loss": 0.6873, | |
| "step": 1895 | |
| }, | |
| { | |
| "epoch": 4.583835946924005, | |
| "grad_norm": 0.515625, | |
| "learning_rate": 0.0001312543163973007, | |
| "loss": 0.6906, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 4.595898673100121, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 0.00013085358224588565, | |
| "loss": 0.7018, | |
| "step": 1905 | |
| }, | |
| { | |
| "epoch": 4.6079613992762365, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.00013045229974300993, | |
| "loss": 0.6973, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 4.620024125452352, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 0.0001300504760205471, | |
| "loss": 0.6942, | |
| "step": 1915 | |
| }, | |
| { | |
| "epoch": 4.632086851628468, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.0001296481182199896, | |
| "loss": 0.692, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 4.644149577804583, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00012924523349232189, | |
| "loss": 0.6993, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 4.656212303980699, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 0.00012884182899789343, | |
| "loss": 0.7002, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 4.668275030156815, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 0.0001284379119062912, | |
| "loss": 0.694, | |
| "step": 1935 | |
| }, | |
| { | |
| "epoch": 4.680337756332931, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 0.00012803348939621252, | |
| "loss": 0.6981, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 4.692400482509047, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 0.0001276285686553373, | |
| "loss": 0.6957, | |
| "step": 1945 | |
| }, | |
| { | |
| "epoch": 4.704463208685163, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 0.00012722315688020047, | |
| "loss": 0.6988, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 4.716525934861279, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00012681726127606376, | |
| "loss": 0.6962, | |
| "step": 1955 | |
| }, | |
| { | |
| "epoch": 4.728588661037395, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 0.00012641088905678802, | |
| "loss": 0.6877, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 4.74065138721351, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.0001260040474447048, | |
| "loss": 0.6978, | |
| "step": 1965 | |
| }, | |
| { | |
| "epoch": 4.752714113389626, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 0.000125596743670488, | |
| "loss": 0.6903, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 4.764776839565742, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 0.0001251889849730254, | |
| "loss": 0.6989, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 4.776839565741858, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 0.00012478077859929, | |
| "loss": 0.7012, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 4.788902291917974, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 0.00012437213180421122, | |
| "loss": 0.6962, | |
| "step": 1985 | |
| }, | |
| { | |
| "epoch": 4.8009650180940895, | |
| "grad_norm": 0.486328125, | |
| "learning_rate": 0.0001239630518505459, | |
| "loss": 0.7007, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 4.813027744270205, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.0001235535460087494, | |
| "loss": 0.6958, | |
| "step": 1995 | |
| }, | |
| { | |
| "epoch": 4.825090470446321, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00012314362155684612, | |
| "loss": 0.6963, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.837153196622436, | |
| "grad_norm": 0.3359375, | |
| "learning_rate": 0.0001227332857803004, | |
| "loss": 0.6985, | |
| "step": 2005 | |
| }, | |
| { | |
| "epoch": 4.849215922798552, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00012232254597188688, | |
| "loss": 0.6881, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 4.861278648974668, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 0.00012191140943156091, | |
| "loss": 0.695, | |
| "step": 2015 | |
| }, | |
| { | |
| "epoch": 4.873341375150784, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00012149988346632894, | |
| "loss": 0.6966, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 4.8854041013269, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.00012108797539011847, | |
| "loss": 0.6937, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 4.897466827503016, | |
| "grad_norm": 0.6796875, | |
| "learning_rate": 0.00012067569252364809, | |
| "loss": 0.6939, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 4.909529553679132, | |
| "grad_norm": 0.68359375, | |
| "learning_rate": 0.00012026304219429759, | |
| "loss": 0.7004, | |
| "step": 2035 | |
| }, | |
| { | |
| "epoch": 4.921592279855247, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 0.0001198500317359774, | |
| "loss": 0.6884, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 4.933655006031363, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00011943666848899853, | |
| "loss": 0.6988, | |
| "step": 2045 | |
| }, | |
| { | |
| "epoch": 4.945717732207479, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.00011902295979994192, | |
| "loss": 0.6925, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 4.957780458383595, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 0.00011860891302152798, | |
| "loss": 0.7009, | |
| "step": 2055 | |
| }, | |
| { | |
| "epoch": 4.969843184559711, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.00011819453551248592, | |
| "loss": 0.6949, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 4.981905910735827, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 0.00011777983463742285, | |
| "loss": 0.6937, | |
| "step": 2065 | |
| }, | |
| { | |
| "epoch": 4.9939686369119425, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.6986, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 4.998793727382388, | |
| "eval_loss": 2.416297197341919, | |
| "eval_runtime": 0.261, | |
| "eval_samples_per_second": 38.317, | |
| "eval_steps_per_second": 3.832, | |
| "step": 2072 | |
| }, | |
| { | |
| "epoch": 5.0060313630880575, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 0.0001169494922762668, | |
| "loss": 0.6738, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 5.018094089264173, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00011653386554759946, | |
| "loss": 0.6829, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 5.030156815440289, | |
| "grad_norm": 0.462890625, | |
| "learning_rate": 0.00011611794496750019, | |
| "loss": 0.6748, | |
| "step": 2085 | |
| }, | |
| { | |
| "epoch": 5.042219541616405, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00011570173792800066, | |
| "loss": 0.6752, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 5.054282267792521, | |
| "grad_norm": 0.57421875, | |
| "learning_rate": 0.00011528525182622371, | |
| "loss": 0.6672, | |
| "step": 2095 | |
| }, | |
| { | |
| "epoch": 5.066344993968637, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 0.00011486849406425188, | |
| "loss": 0.6744, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 5.078407720144753, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00011445147204899587, | |
| "loss": 0.6868, | |
| "step": 2105 | |
| }, | |
| { | |
| "epoch": 5.090470446320868, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 0.00011403419319206284, | |
| "loss": 0.681, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 5.102533172496984, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00011361666490962468, | |
| "loss": 0.6833, | |
| "step": 2115 | |
| }, | |
| { | |
| "epoch": 5.1145958986731, | |
| "grad_norm": 0.41796875, | |
| "learning_rate": 0.0001131988946222863, | |
| "loss": 0.6677, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 5.126658624849216, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00011278088975495369, | |
| "loss": 0.6852, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 5.138721351025332, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00011236265773670196, | |
| "loss": 0.6755, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 5.150784077201448, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 0.00011194420600064329, | |
| "loss": 0.6803, | |
| "step": 2135 | |
| }, | |
| { | |
| "epoch": 5.162846803377564, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00011152554198379484, | |
| "loss": 0.6776, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 5.1749095295536796, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00011110667312694653, | |
| "loss": 0.6764, | |
| "step": 2145 | |
| }, | |
| { | |
| "epoch": 5.186972255729795, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.00011068760687452895, | |
| "loss": 0.6748, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 5.1990349819059105, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00011026835067448082, | |
| "loss": 0.6837, | |
| "step": 2155 | |
| }, | |
| { | |
| "epoch": 5.211097708082026, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 0.00010984891197811687, | |
| "loss": 0.6763, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 5.223160434258142, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 0.00010942929823999517, | |
| "loss": 0.685, | |
| "step": 2165 | |
| }, | |
| { | |
| "epoch": 5.235223160434258, | |
| "grad_norm": 0.5234375, | |
| "learning_rate": 0.00010900951691778481, | |
| "loss": 0.6851, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 5.247285886610374, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00010858957547213327, | |
| "loss": 0.677, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 5.25934861278649, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 0.00010816948136653386, | |
| "loss": 0.6815, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 5.271411338962605, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 0.0001077492420671931, | |
| "loss": 0.6799, | |
| "step": 2185 | |
| }, | |
| { | |
| "epoch": 5.283474065138721, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 0.000107328865042898, | |
| "loss": 0.677, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 5.295536791314837, | |
| "grad_norm": 0.4375, | |
| "learning_rate": 0.00010690835776488328, | |
| "loss": 0.6851, | |
| "step": 2195 | |
| }, | |
| { | |
| "epoch": 5.307599517490953, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 0.00010648772770669861, | |
| "loss": 0.6816, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 5.319662243667069, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 0.00010606698234407586, | |
| "loss": 0.6715, | |
| "step": 2205 | |
| }, | |
| { | |
| "epoch": 5.331724969843185, | |
| "grad_norm": 0.53515625, | |
| "learning_rate": 0.00010564612915479612, | |
| "loss": 0.6798, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 5.343787696019301, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 0.00010522517561855683, | |
| "loss": 0.6787, | |
| "step": 2215 | |
| }, | |
| { | |
| "epoch": 5.355850422195417, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 0.00010480412921683888, | |
| "loss": 0.6855, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 5.367913148371532, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 0.00010438299743277371, | |
| "loss": 0.6823, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 5.379975874547648, | |
| "grad_norm": 0.375, | |
| "learning_rate": 0.00010396178775101014, | |
| "loss": 0.6781, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 5.3920386007237635, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 0.00010354050765758147, | |
| "loss": 0.6819, | |
| "step": 2235 | |
| }, | |
| { | |
| "epoch": 5.404101326899879, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 0.00010311916463977242, | |
| "loss": 0.6793, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 5.416164053075995, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 0.00010269776618598602, | |
| "loss": 0.6783, | |
| "step": 2245 | |
| }, | |
| { | |
| "epoch": 5.428226779252111, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 0.00010227631978561056, | |
| "loss": 0.6899, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 5.440289505428227, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 0.00010185483292888654, | |
| "loss": 0.6771, | |
| "step": 2255 | |
| }, | |
| { | |
| "epoch": 5.452352231604342, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 0.00010143331310677331, | |
| "loss": 0.6756, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 5.464414957780458, | |
| "grad_norm": 0.498046875, | |
| "learning_rate": 0.00010101176781081625, | |
| "loss": 0.6761, | |
| "step": 2265 | |
| }, | |
| { | |
| "epoch": 5.476477683956574, | |
| "grad_norm": 0.8671875, | |
| "learning_rate": 0.00010059020453301345, | |
| "loss": 0.6769, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 5.48854041013269, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00010016863076568254, | |
| "loss": 0.6803, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 5.500603136308806, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 9.974705400132764e-05, | |
| "loss": 0.6773, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 5.512665862484922, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.932548173250607e-05, | |
| "loss": 0.6759, | |
| "step": 2285 | |
| }, | |
| { | |
| "epoch": 5.524728588661038, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 9.890392145169531e-05, | |
| "loss": 0.6813, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 5.536791314837153, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 9.848238065115975e-05, | |
| "loss": 0.6801, | |
| "step": 2295 | |
| }, | |
| { | |
| "epoch": 5.548854041013269, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.806086682281758e-05, | |
| "loss": 0.6768, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 5.560916767189385, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.763938745810757e-05, | |
| "loss": 0.677, | |
| "step": 2305 | |
| }, | |
| { | |
| "epoch": 5.572979493365501, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 9.721795004785605e-05, | |
| "loss": 0.6783, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 5.5850422195416165, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 9.679656208214366e-05, | |
| "loss": 0.6885, | |
| "step": 2315 | |
| }, | |
| { | |
| "epoch": 5.597104945717732, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.637523105017229e-05, | |
| "loss": 0.6786, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 5.609167671893848, | |
| "grad_norm": 0.48046875, | |
| "learning_rate": 9.595396444013205e-05, | |
| "loss": 0.6791, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 5.621230398069963, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.553276973906786e-05, | |
| "loss": 0.6731, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 5.633293124246079, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 9.511165443274688e-05, | |
| "loss": 0.6718, | |
| "step": 2335 | |
| }, | |
| { | |
| "epoch": 5.645355850422195, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 9.469062600552509e-05, | |
| "loss": 0.6723, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 5.657418576598311, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 9.426969194021437e-05, | |
| "loss": 0.6793, | |
| "step": 2345 | |
| }, | |
| { | |
| "epoch": 5.669481302774427, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.384885971794961e-05, | |
| "loss": 0.6822, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 5.681544028950543, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 9.342813681805565e-05, | |
| "loss": 0.6811, | |
| "step": 2355 | |
| }, | |
| { | |
| "epoch": 5.693606755126659, | |
| "grad_norm": 0.421875, | |
| "learning_rate": 9.300753071791434e-05, | |
| "loss": 0.6811, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 5.705669481302774, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 9.258704889283175e-05, | |
| "loss": 0.6873, | |
| "step": 2365 | |
| }, | |
| { | |
| "epoch": 5.71773220747889, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.216669881590515e-05, | |
| "loss": 0.6764, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 5.729794933655006, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 9.174648795789039e-05, | |
| "loss": 0.6735, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 5.741857659831122, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 9.132642378706894e-05, | |
| "loss": 0.6767, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 5.753920386007238, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 9.09065137691153e-05, | |
| "loss": 0.6793, | |
| "step": 2385 | |
| }, | |
| { | |
| "epoch": 5.765983112183354, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 9.048676536696425e-05, | |
| "loss": 0.6815, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 5.7780458383594695, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 9.006718604067823e-05, | |
| "loss": 0.6889, | |
| "step": 2395 | |
| }, | |
| { | |
| "epoch": 5.790108564535585, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 8.964778324731467e-05, | |
| "loss": 0.6778, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 5.8021712907117005, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 8.922856444079362e-05, | |
| "loss": 0.6748, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 5.814234016887816, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 8.880953707176514e-05, | |
| "loss": 0.6752, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 5.826296743063932, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 8.839070858747697e-05, | |
| "loss": 0.6773, | |
| "step": 2415 | |
| }, | |
| { | |
| "epoch": 5.838359469240048, | |
| "grad_norm": 0.458984375, | |
| "learning_rate": 8.797208643164212e-05, | |
| "loss": 0.6766, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 5.850422195416164, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 8.755367804430651e-05, | |
| "loss": 0.6801, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 5.86248492159228, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 8.713549086171691e-05, | |
| "loss": 0.6732, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 5.874547647768396, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 8.671753231618866e-05, | |
| "loss": 0.6822, | |
| "step": 2435 | |
| }, | |
| { | |
| "epoch": 5.886610373944512, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 8.629980983597358e-05, | |
| "loss": 0.6734, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 5.898673100120627, | |
| "grad_norm": 0.59375, | |
| "learning_rate": 8.5882330845128e-05, | |
| "loss": 0.6751, | |
| "step": 2445 | |
| }, | |
| { | |
| "epoch": 5.910735826296743, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 8.546510276338078e-05, | |
| "loss": 0.6684, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 5.922798552472859, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 8.504813300600141e-05, | |
| "loss": 0.6755, | |
| "step": 2455 | |
| }, | |
| { | |
| "epoch": 5.934861278648975, | |
| "grad_norm": 0.5546875, | |
| "learning_rate": 8.463142898366834e-05, | |
| "loss": 0.6837, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 5.946924004825091, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 8.42149981023371e-05, | |
| "loss": 0.6831, | |
| "step": 2465 | |
| }, | |
| { | |
| "epoch": 5.958986731001207, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 8.37988477631088e-05, | |
| "loss": 0.675, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 5.9710494571773225, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 8.33829853620986e-05, | |
| "loss": 0.6663, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 5.9831121833534375, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 8.296741829030418e-05, | |
| "loss": 0.6817, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 5.9951749095295535, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 8.255215393347443e-05, | |
| "loss": 0.6737, | |
| "step": 2485 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.423604965209961, | |
| "eval_runtime": 0.2373, | |
| "eval_samples_per_second": 42.143, | |
| "eval_steps_per_second": 4.214, | |
| "step": 2487 | |
| }, | |
| { | |
| "epoch": 6.007237635705669, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 8.213719967197817e-05, | |
| "loss": 0.6732, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 6.019300361881785, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 8.172256288067298e-05, | |
| "loss": 0.6612, | |
| "step": 2495 | |
| }, | |
| { | |
| "epoch": 6.031363088057901, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 8.130825092877418e-05, | |
| "loss": 0.6584, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.043425814234017, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 8.089427117972378e-05, | |
| "loss": 0.6623, | |
| "step": 2505 | |
| }, | |
| { | |
| "epoch": 6.055488540410133, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 8.04806309910597e-05, | |
| "loss": 0.6606, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 6.067551266586248, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 8.006733771428485e-05, | |
| "loss": 0.667, | |
| "step": 2515 | |
| }, | |
| { | |
| "epoch": 6.079613992762364, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 7.965439869473664e-05, | |
| "loss": 0.6666, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 6.09167671893848, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 7.924182127145642e-05, | |
| "loss": 0.6538, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 6.103739445114596, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.882961277705895e-05, | |
| "loss": 0.6693, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 6.115802171290712, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 7.841778053760211e-05, | |
| "loss": 0.6706, | |
| "step": 2535 | |
| }, | |
| { | |
| "epoch": 6.127864897466828, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 7.800633187245673e-05, | |
| "loss": 0.6676, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 6.139927623642944, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 7.759527409417653e-05, | |
| "loss": 0.6664, | |
| "step": 2545 | |
| }, | |
| { | |
| "epoch": 6.151990349819059, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 7.718461450836804e-05, | |
| "loss": 0.6557, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 6.164053075995175, | |
| "grad_norm": 0.578125, | |
| "learning_rate": 7.67743604135609e-05, | |
| "loss": 0.6618, | |
| "step": 2555 | |
| }, | |
| { | |
| "epoch": 6.1761158021712905, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 7.636451910107806e-05, | |
| "loss": 0.6667, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 6.1881785283474064, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 7.595509785490617e-05, | |
| "loss": 0.6557, | |
| "step": 2565 | |
| }, | |
| { | |
| "epoch": 6.200241254523522, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 7.554610395156624e-05, | |
| "loss": 0.6644, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 6.212303980699638, | |
| "grad_norm": 0.40234375, | |
| "learning_rate": 7.513754465998418e-05, | |
| "loss": 0.6622, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 6.224366706875754, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 7.472942724136174e-05, | |
| "loss": 0.6657, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 6.23642943305187, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 7.432175894904733e-05, | |
| "loss": 0.669, | |
| "step": 2585 | |
| }, | |
| { | |
| "epoch": 6.248492159227985, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 7.391454702840722e-05, | |
| "loss": 0.6641, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 6.260554885404101, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 7.350779871669669e-05, | |
| "loss": 0.6673, | |
| "step": 2595 | |
| }, | |
| { | |
| "epoch": 6.272617611580217, | |
| "grad_norm": 0.439453125, | |
| "learning_rate": 7.310152124293146e-05, | |
| "loss": 0.6639, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 6.284680337756333, | |
| "grad_norm": 0.4140625, | |
| "learning_rate": 7.269572182775921e-05, | |
| "loss": 0.6649, | |
| "step": 2605 | |
| }, | |
| { | |
| "epoch": 6.296743063932449, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 7.229040768333115e-05, | |
| "loss": 0.6606, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 6.308805790108565, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 7.188558601317396e-05, | |
| "loss": 0.6597, | |
| "step": 2615 | |
| }, | |
| { | |
| "epoch": 6.320868516284681, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 7.14812640120618e-05, | |
| "loss": 0.6671, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 6.332931242460796, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 7.107744886588824e-05, | |
| "loss": 0.664, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 6.344993968636912, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 7.067414775153871e-05, | |
| "loss": 0.6565, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 6.357056694813028, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 7.027136783676295e-05, | |
| "loss": 0.6623, | |
| "step": 2635 | |
| }, | |
| { | |
| "epoch": 6.3691194209891435, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 6.986911628004753e-05, | |
| "loss": 0.6641, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 6.381182147165259, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 6.94674002304887e-05, | |
| "loss": 0.6664, | |
| "step": 2645 | |
| }, | |
| { | |
| "epoch": 6.393244873341375, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 6.906622682766526e-05, | |
| "loss": 0.6601, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 6.405307599517491, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 6.866560320151179e-05, | |
| "loss": 0.6696, | |
| "step": 2655 | |
| }, | |
| { | |
| "epoch": 6.417370325693607, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 6.826553647219175e-05, | |
| "loss": 0.6652, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 6.429433051869722, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 6.786603374997111e-05, | |
| "loss": 0.6721, | |
| "step": 2665 | |
| }, | |
| { | |
| "epoch": 6.441495778045838, | |
| "grad_norm": 0.4609375, | |
| "learning_rate": 6.74671021350919e-05, | |
| "loss": 0.6717, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 6.453558504221954, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.706874871764603e-05, | |
| "loss": 0.6646, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 6.46562123039807, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 6.667098057744927e-05, | |
| "loss": 0.6673, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 6.477683956574186, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 6.627380478391543e-05, | |
| "loss": 0.66, | |
| "step": 2685 | |
| }, | |
| { | |
| "epoch": 6.489746682750302, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 6.587722839593073e-05, | |
| "loss": 0.6636, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 6.501809408926418, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 6.548125846172836e-05, | |
| "loss": 0.6635, | |
| "step": 2695 | |
| }, | |
| { | |
| "epoch": 6.513872135102533, | |
| "grad_norm": 0.45703125, | |
| "learning_rate": 6.508590201876317e-05, | |
| "loss": 0.6708, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 6.525934861278649, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 6.469116609358654e-05, | |
| "loss": 0.669, | |
| "step": 2705 | |
| }, | |
| { | |
| "epoch": 6.537997587454765, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 6.429705770172168e-05, | |
| "loss": 0.6638, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 6.550060313630881, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 6.390358384753881e-05, | |
| "loss": 0.6574, | |
| "step": 2715 | |
| }, | |
| { | |
| "epoch": 6.5621230398069965, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 6.351075152413068e-05, | |
| "loss": 0.6666, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 6.574185765983112, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 6.311856771318832e-05, | |
| "loss": 0.6734, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 6.586248492159228, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 6.272703938487694e-05, | |
| "loss": 0.6633, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 6.598311218335343, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 6.233617349771205e-05, | |
| "loss": 0.6625, | |
| "step": 2735 | |
| }, | |
| { | |
| "epoch": 6.610373944511459, | |
| "grad_norm": 0.41015625, | |
| "learning_rate": 6.194597699843581e-05, | |
| "loss": 0.6596, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 6.622436670687575, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 6.155645682189351e-05, | |
| "loss": 0.6679, | |
| "step": 2745 | |
| }, | |
| { | |
| "epoch": 6.634499396863691, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 6.116761989091042e-05, | |
| "loss": 0.6646, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 6.646562123039807, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 6.0779473116168627e-05, | |
| "loss": 0.6662, | |
| "step": 2755 | |
| }, | |
| { | |
| "epoch": 6.658624849215923, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 0.6681, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 6.670687575392039, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 6.000527761668513e-05, | |
| "loss": 0.661, | |
| "step": 2765 | |
| }, | |
| { | |
| "epoch": 6.682750301568154, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 5.961924265148777e-05, | |
| "loss": 0.6679, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 6.69481302774427, | |
| "grad_norm": 0.5390625, | |
| "learning_rate": 5.9233925361375864e-05, | |
| "loss": 0.6653, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 6.706875753920386, | |
| "grad_norm": 0.470703125, | |
| "learning_rate": 5.884933259447798e-05, | |
| "loss": 0.6685, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 6.718938480096502, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 5.8465471186046015e-05, | |
| "loss": 0.6678, | |
| "step": 2785 | |
| }, | |
| { | |
| "epoch": 6.731001206272618, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 5.8082347958333625e-05, | |
| "loss": 0.6661, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 6.743063932448734, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.769996972047491e-05, | |
| "loss": 0.6664, | |
| "step": 2795 | |
| }, | |
| { | |
| "epoch": 6.7551266586248495, | |
| "grad_norm": 0.375, | |
| "learning_rate": 5.731834326836366e-05, | |
| "loss": 0.6703, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 6.7671893848009645, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 5.693747538453229e-05, | |
| "loss": 0.6661, | |
| "step": 2805 | |
| }, | |
| { | |
| "epoch": 6.7792521109770805, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 5.6557372838031384e-05, | |
| "loss": 0.6598, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 6.791314837153196, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 5.6178042384309546e-05, | |
| "loss": 0.6663, | |
| "step": 2815 | |
| }, | |
| { | |
| "epoch": 6.803377563329312, | |
| "grad_norm": 0.375, | |
| "learning_rate": 5.579949076509305e-05, | |
| "loss": 0.6664, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 6.815440289505428, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.542172470826632e-05, | |
| "loss": 0.6623, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 6.827503015681544, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.5044750927752106e-05, | |
| "loss": 0.6643, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 6.83956574185766, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.466857612339229e-05, | |
| "loss": 0.6641, | |
| "step": 2835 | |
| }, | |
| { | |
| "epoch": 6.851628468033776, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 5.429320698082887e-05, | |
| "loss": 0.6631, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 6.863691194209892, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 5.391865017138493e-05, | |
| "loss": 0.6648, | |
| "step": 2845 | |
| }, | |
| { | |
| "epoch": 6.875753920386007, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 5.354491235194635e-05, | |
| "loss": 0.6698, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 6.887816646562123, | |
| "grad_norm": 0.42578125, | |
| "learning_rate": 5.3172000164843195e-05, | |
| "loss": 0.6569, | |
| "step": 2855 | |
| }, | |
| { | |
| "epoch": 6.899879372738239, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 5.279992023773195e-05, | |
| "loss": 0.6653, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 6.911942098914355, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 5.2428679183477505e-05, | |
| "loss": 0.6678, | |
| "step": 2865 | |
| }, | |
| { | |
| "epoch": 6.924004825090471, | |
| "grad_norm": 0.5625, | |
| "learning_rate": 5.205828360003568e-05, | |
| "loss": 0.6667, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 6.936067551266587, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 5.168874007033615e-05, | |
| "loss": 0.6647, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 6.9481302774427025, | |
| "grad_norm": 0.419921875, | |
| "learning_rate": 5.1320055162165115e-05, | |
| "loss": 0.6611, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 6.9601930036188175, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 5.0952235428048966e-05, | |
| "loss": 0.6668, | |
| "step": 2885 | |
| }, | |
| { | |
| "epoch": 6.9722557297949335, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.0585287405137305e-05, | |
| "loss": 0.6656, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 6.984318455971049, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 5.021921761508739e-05, | |
| "loss": 0.6631, | |
| "step": 2895 | |
| }, | |
| { | |
| "epoch": 6.996381182147165, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.9854032563947714e-05, | |
| "loss": 0.6633, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 6.998793727382388, | |
| "eval_loss": 2.4494216442108154, | |
| "eval_runtime": 0.2569, | |
| "eval_samples_per_second": 38.924, | |
| "eval_steps_per_second": 3.892, | |
| "step": 2901 | |
| }, | |
| { | |
| "epoch": 7.008443908323281, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 4.9489738742042616e-05, | |
| "loss": 0.6572, | |
| "step": 2905 | |
| }, | |
| { | |
| "epoch": 7.020506634499397, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.912634262385695e-05, | |
| "loss": 0.6561, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 7.032569360675513, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 4.876385066792084e-05, | |
| "loss": 0.659, | |
| "step": 2915 | |
| }, | |
| { | |
| "epoch": 7.044632086851628, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 4.8402269316695134e-05, | |
| "loss": 0.6543, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 7.056694813027744, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 4.804160499645667e-05, | |
| "loss": 0.6483, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 7.06875753920386, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.768186411718417e-05, | |
| "loss": 0.6515, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 7.080820265379976, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 4.732305307244444e-05, | |
| "loss": 0.6613, | |
| "step": 2935 | |
| }, | |
| { | |
| "epoch": 7.092882991556092, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 4.696517823927842e-05, | |
| "loss": 0.6636, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 7.104945717732208, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 4.660824597808825e-05, | |
| "loss": 0.6584, | |
| "step": 2945 | |
| }, | |
| { | |
| "epoch": 7.117008443908324, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.625226263252386e-05, | |
| "loss": 0.6458, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 7.129071170084439, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.589723452937049e-05, | |
| "loss": 0.652, | |
| "step": 2955 | |
| }, | |
| { | |
| "epoch": 7.141133896260555, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 4.554316797843609e-05, | |
| "loss": 0.6545, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 7.1531966224366705, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.519006927243922e-05, | |
| "loss": 0.6529, | |
| "step": 2965 | |
| }, | |
| { | |
| "epoch": 7.1652593486127865, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 4.483794468689728e-05, | |
| "loss": 0.6543, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 7.177322074788902, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 4.448680048001485e-05, | |
| "loss": 0.6566, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 7.189384800965018, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 4.413664289257265e-05, | |
| "loss": 0.6579, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 7.201447527141134, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 4.3787478147816296e-05, | |
| "loss": 0.6578, | |
| "step": 2985 | |
| }, | |
| { | |
| "epoch": 7.213510253317249, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 4.343931245134616e-05, | |
| "loss": 0.6566, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 7.225572979493365, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 4.3092151991006654e-05, | |
| "loss": 0.6571, | |
| "step": 2995 | |
| }, | |
| { | |
| "epoch": 7.237635705669481, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.274600293677647e-05, | |
| "loss": 0.6571, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.249698431845597, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.240087144065895e-05, | |
| "loss": 0.6607, | |
| "step": 3005 | |
| }, | |
| { | |
| "epoch": 7.261761158021713, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.2056763636572574e-05, | |
| "loss": 0.6583, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 7.273823884197829, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 4.1713685640242165e-05, | |
| "loss": 0.6566, | |
| "step": 3015 | |
| }, | |
| { | |
| "epoch": 7.285886610373945, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 4.137164354908999e-05, | |
| "loss": 0.651, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 7.297949336550061, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.103064344212748e-05, | |
| "loss": 0.6496, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 7.310012062726176, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.069069137984731e-05, | |
| "loss": 0.6603, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 7.322074788902292, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 4.035179340411541e-05, | |
| "loss": 0.6599, | |
| "step": 3035 | |
| }, | |
| { | |
| "epoch": 7.334137515078408, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.001395553806391e-05, | |
| "loss": 0.6504, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 7.3462002412545235, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 3.967718378598376e-05, | |
| "loss": 0.66, | |
| "step": 3045 | |
| }, | |
| { | |
| "epoch": 7.3582629674306395, | |
| "grad_norm": 0.48828125, | |
| "learning_rate": 3.9341484133218366e-05, | |
| "loss": 0.6586, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 7.370325693606755, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 3.9006862546056876e-05, | |
| "loss": 0.6655, | |
| "step": 3055 | |
| }, | |
| { | |
| "epoch": 7.382388419782871, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.8673324971628357e-05, | |
| "loss": 0.657, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 7.394451145958986, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 3.834087733779611e-05, | |
| "loss": 0.6577, | |
| "step": 3065 | |
| }, | |
| { | |
| "epoch": 7.406513872135102, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.800952555305216e-05, | |
| "loss": 0.6587, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 7.418576598311218, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.767927550641237e-05, | |
| "loss": 0.6615, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 7.430639324487334, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 3.7350133067311686e-05, | |
| "loss": 0.6537, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 7.44270205066345, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 3.702210408550002e-05, | |
| "loss": 0.6545, | |
| "step": 3085 | |
| }, | |
| { | |
| "epoch": 7.454764776839566, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.669519439093801e-05, | |
| "loss": 0.6571, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 7.466827503015682, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 3.6369409793693544e-05, | |
| "loss": 0.6555, | |
| "step": 3095 | |
| }, | |
| { | |
| "epoch": 7.478890229191798, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 3.604475608383858e-05, | |
| "loss": 0.659, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 7.490952955367913, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 3.5721239031346066e-05, | |
| "loss": 0.6569, | |
| "step": 3105 | |
| }, | |
| { | |
| "epoch": 7.503015681544029, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.539886438598756e-05, | |
| "loss": 0.6622, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 7.515078407720145, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.507763787723086e-05, | |
| "loss": 0.6545, | |
| "step": 3115 | |
| }, | |
| { | |
| "epoch": 7.527141133896261, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 3.475756521413839e-05, | |
| "loss": 0.6543, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 7.5392038600723765, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 3.443865208526554e-05, | |
| "loss": 0.6597, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 7.5512665862484925, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.412090415855963e-05, | |
| "loss": 0.656, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 7.563329312424608, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.3804327081259304e-05, | |
| "loss": 0.6556, | |
| "step": 3135 | |
| }, | |
| { | |
| "epoch": 7.575392038600723, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.348892647979389e-05, | |
| "loss": 0.6591, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 7.587454764776839, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 3.317470795968376e-05, | |
| "loss": 0.6617, | |
| "step": 3145 | |
| }, | |
| { | |
| "epoch": 7.599517490952955, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.2861677105440336e-05, | |
| "loss": 0.6562, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 7.611580217129071, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 3.254983948046705e-05, | |
| "loss": 0.6485, | |
| "step": 3155 | |
| }, | |
| { | |
| "epoch": 7.623642943305187, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 3.223920062696052e-05, | |
| "loss": 0.6595, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 7.635705669481303, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 3.192976606581186e-05, | |
| "loss": 0.6538, | |
| "step": 3165 | |
| }, | |
| { | |
| "epoch": 7.647768395657419, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 3.1621541296508695e-05, | |
| "loss": 0.6535, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 7.659831121833534, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 3.131453179703734e-05, | |
| "loss": 0.6542, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 7.67189384800965, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 3.100874302378559e-05, | |
| "loss": 0.657, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 7.683956574185766, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.0704180411445524e-05, | |
| "loss": 0.6587, | |
| "step": 3185 | |
| }, | |
| { | |
| "epoch": 7.696019300361882, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.0400849372917073e-05, | |
| "loss": 0.6587, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 7.708082026537998, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 3.009875529921181e-05, | |
| "loss": 0.6564, | |
| "step": 3195 | |
| }, | |
| { | |
| "epoch": 7.720144752714114, | |
| "grad_norm": 0.375, | |
| "learning_rate": 2.979790355935703e-05, | |
| "loss": 0.659, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 7.7322074788902295, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 2.9498299500300518e-05, | |
| "loss": 0.6536, | |
| "step": 3205 | |
| }, | |
| { | |
| "epoch": 7.744270205066345, | |
| "grad_norm": 0.58984375, | |
| "learning_rate": 2.919994844681524e-05, | |
| "loss": 0.6572, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 7.7563329312424605, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 2.890285570140504e-05, | |
| "loss": 0.6549, | |
| "step": 3215 | |
| }, | |
| { | |
| "epoch": 7.768395657418576, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.8607026544210114e-05, | |
| "loss": 0.6529, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 7.780458383594692, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 2.8312466232913282e-05, | |
| "loss": 0.6607, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 7.792521109770808, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.801918000264665e-05, | |
| "loss": 0.6606, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 7.804583835946924, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.7727173065898347e-05, | |
| "loss": 0.6537, | |
| "step": 3235 | |
| }, | |
| { | |
| "epoch": 7.81664656212304, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 2.7436450612420095e-05, | |
| "loss": 0.6492, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 7.828709288299155, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 2.7147017809134822e-05, | |
| "loss": 0.6536, | |
| "step": 3245 | |
| }, | |
| { | |
| "epoch": 7.840772014475271, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.6858879800044866e-05, | |
| "loss": 0.6572, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 7.852834740651387, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 2.6572041706140683e-05, | |
| "loss": 0.6598, | |
| "step": 3255 | |
| }, | |
| { | |
| "epoch": 7.864897466827503, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 2.6286508625309624e-05, | |
| "loss": 0.6583, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 7.876960193003619, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 2.6002285632245482e-05, | |
| "loss": 0.6595, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 7.889022919179735, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.57193777783582e-05, | |
| "loss": 0.6511, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 7.901085645355851, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 2.5437790091684244e-05, | |
| "loss": 0.6534, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 7.913148371531967, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.515752757679707e-05, | |
| "loss": 0.6527, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 7.9252110977080825, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 2.4878595214718236e-05, | |
| "loss": 0.6579, | |
| "step": 3285 | |
| }, | |
| { | |
| "epoch": 7.9372738238841976, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 2.4600997962828987e-05, | |
| "loss": 0.6571, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 7.9493365500603135, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 2.432474075478194e-05, | |
| "loss": 0.6556, | |
| "step": 3295 | |
| }, | |
| { | |
| "epoch": 7.961399276236429, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.404982850041363e-05, | |
| "loss": 0.6585, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 7.973462002412545, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.3776266085657018e-05, | |
| "loss": 0.6544, | |
| "step": 3305 | |
| }, | |
| { | |
| "epoch": 7.985524728588661, | |
| "grad_norm": 0.375, | |
| "learning_rate": 2.3504058372454884e-05, | |
| "loss": 0.6539, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 7.997587454764777, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 2.3233210198673218e-05, | |
| "loss": 0.661, | |
| "step": 3315 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.462099313735962, | |
| "eval_runtime": 0.2392, | |
| "eval_samples_per_second": 41.81, | |
| "eval_steps_per_second": 4.181, | |
| "step": 3316 | |
| }, | |
| { | |
| "epoch": 8.009650180940893, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.2963726378015327e-05, | |
| "loss": 0.6547, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 8.021712907117008, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 2.269561169993637e-05, | |
| "loss": 0.6514, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 8.033775633293125, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 2.242887092955801e-05, | |
| "loss": 0.6449, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 8.04583835946924, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 2.2163508807583998e-05, | |
| "loss": 0.6495, | |
| "step": 3335 | |
| }, | |
| { | |
| "epoch": 8.057901085645355, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 2.189953005021569e-05, | |
| "loss": 0.6556, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 8.069963811821472, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.1636939349068308e-05, | |
| "loss": 0.6537, | |
| "step": 3345 | |
| }, | |
| { | |
| "epoch": 8.082026537997587, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.1375741371087677e-05, | |
| "loss": 0.6504, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 8.094089264173704, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 2.111594075846701e-05, | |
| "loss": 0.6589, | |
| "step": 3355 | |
| }, | |
| { | |
| "epoch": 8.106151990349819, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.085754212856471e-05, | |
| "loss": 0.6536, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 8.118214716525936, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.0600550073822056e-05, | |
| "loss": 0.6512, | |
| "step": 3365 | |
| }, | |
| { | |
| "epoch": 8.13027744270205, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.0344969161681792e-05, | |
| "loss": 0.6497, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 8.142340168878167, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.0090803934506764e-05, | |
| "loss": 0.6492, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 8.154402895054282, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 1.983805890949927e-05, | |
| "loss": 0.6446, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 8.166465621230397, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 1.9586738578620855e-05, | |
| "loss": 0.6518, | |
| "step": 3385 | |
| }, | |
| { | |
| "epoch": 8.178528347406514, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.9336847408512328e-05, | |
| "loss": 0.6605, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 8.19059107358263, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.908838984041452e-05, | |
| "loss": 0.6528, | |
| "step": 3395 | |
| }, | |
| { | |
| "epoch": 8.202653799758746, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 1.884137029008921e-05, | |
| "loss": 0.6508, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 8.214716525934861, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.859579314774079e-05, | |
| "loss": 0.6522, | |
| "step": 3405 | |
| }, | |
| { | |
| "epoch": 8.226779252110978, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 1.8351662777938127e-05, | |
| "loss": 0.6521, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 8.238841978287093, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 1.810898351953699e-05, | |
| "loss": 0.6537, | |
| "step": 3415 | |
| }, | |
| { | |
| "epoch": 8.250904704463208, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.7867759685603114e-05, | |
| "loss": 0.6518, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 8.262967430639325, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.762799556333524e-05, | |
| "loss": 0.6546, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 8.27503015681544, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.738969541398926e-05, | |
| "loss": 0.6564, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 8.287092882991557, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.7152863472802195e-05, | |
| "loss": 0.6515, | |
| "step": 3435 | |
| }, | |
| { | |
| "epoch": 8.299155609167672, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 1.691750394891707e-05, | |
| "loss": 0.6526, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 8.311218335343789, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.668362102530815e-05, | |
| "loss": 0.6508, | |
| "step": 3445 | |
| }, | |
| { | |
| "epoch": 8.323281061519904, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 0.6547, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 8.335343787696019, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.6220301579525798e-05, | |
| "loss": 0.652, | |
| "step": 3455 | |
| }, | |
| { | |
| "epoch": 8.347406513872135, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.59908732917899e-05, | |
| "loss": 0.6496, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 8.35946924004825, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.5762938073058853e-05, | |
| "loss": 0.654, | |
| "step": 3465 | |
| }, | |
| { | |
| "epoch": 8.371531966224367, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.5536499974356866e-05, | |
| "loss": 0.6491, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 8.383594692400482, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.5311563020100373e-05, | |
| "loss": 0.6531, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 8.395657418576599, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.5088131208026367e-05, | |
| "loss": 0.6545, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 8.407720144752714, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.4866208509121383e-05, | |
| "loss": 0.6506, | |
| "step": 3485 | |
| }, | |
| { | |
| "epoch": 8.41978287092883, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.4645798867551008e-05, | |
| "loss": 0.6495, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 8.431845597104946, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.442690620058964e-05, | |
| "loss": 0.656, | |
| "step": 3495 | |
| }, | |
| { | |
| "epoch": 8.443908323281061, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.4209534398551016e-05, | |
| "loss": 0.6574, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.455971049457178, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 1.3993687324718929e-05, | |
| "loss": 0.6553, | |
| "step": 3505 | |
| }, | |
| { | |
| "epoch": 8.468033775633293, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.3779368815278647e-05, | |
| "loss": 0.6513, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 8.48009650180941, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 1.3566582679248796e-05, | |
| "loss": 0.652, | |
| "step": 3515 | |
| }, | |
| { | |
| "epoch": 8.492159227985525, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.335533269841347e-05, | |
| "loss": 0.655, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 8.50422195416164, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.314562262725526e-05, | |
| "loss": 0.6593, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 8.516284680337757, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.2937456192888309e-05, | |
| "loss": 0.6572, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 8.528347406513872, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 1.2730837094992199e-05, | |
| "loss": 0.6437, | |
| "step": 3535 | |
| }, | |
| { | |
| "epoch": 8.540410132689988, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.252576900574618e-05, | |
| "loss": 0.655, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 8.552472858866103, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.2322255569763852e-05, | |
| "loss": 0.6562, | |
| "step": 3545 | |
| }, | |
| { | |
| "epoch": 8.56453558504222, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.2120300404028507e-05, | |
| "loss": 0.6499, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 8.576598311218335, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.1919907097828653e-05, | |
| "loss": 0.6584, | |
| "step": 3555 | |
| }, | |
| { | |
| "epoch": 8.588661037394452, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.1721079212694452e-05, | |
| "loss": 0.656, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 8.600723763570567, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.1523820282334219e-05, | |
| "loss": 0.6548, | |
| "step": 3565 | |
| }, | |
| { | |
| "epoch": 8.612786489746682, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.1328133812571784e-05, | |
| "loss": 0.6494, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 8.624849215922799, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.1134023281284023e-05, | |
| "loss": 0.6484, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 8.636911942098914, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.0941492138339183e-05, | |
| "loss": 0.6616, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 8.64897466827503, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.0750543805535518e-05, | |
| "loss": 0.6589, | |
| "step": 3585 | |
| }, | |
| { | |
| "epoch": 8.661037394451146, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 1.0561181676540444e-05, | |
| "loss": 0.6541, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 8.67310012062726, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.037340911683028e-05, | |
| "loss": 0.6455, | |
| "step": 3595 | |
| }, | |
| { | |
| "epoch": 8.685162846803378, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 1.01872294636304e-05, | |
| "loss": 0.6522, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 8.697225572979493, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 1.0002646025855888e-05, | |
| "loss": 0.6539, | |
| "step": 3605 | |
| }, | |
| { | |
| "epoch": 8.70928829915561, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.81966208405285e-06, | |
| "loss": 0.6543, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 8.721351025331725, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.638280890339945e-06, | |
| "loss": 0.6546, | |
| "step": 3615 | |
| }, | |
| { | |
| "epoch": 8.733413751507841, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 9.458505668350759e-06, | |
| "loss": 0.6543, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 8.745476477683956, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.280339613176348e-06, | |
| "loss": 0.6538, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 8.757539203860073, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.103785891308547e-06, | |
| "loss": 0.661, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 8.769601930036188, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.928847640583715e-06, | |
| "loss": 0.644, | |
| "step": 3635 | |
| }, | |
| { | |
| "epoch": 8.781664656212303, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 8.755527970126853e-06, | |
| "loss": 0.6599, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 8.79372738238842, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 8.58382996029652e-06, | |
| "loss": 0.6494, | |
| "step": 3645 | |
| }, | |
| { | |
| "epoch": 8.805790108564535, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 8.413756662629879e-06, | |
| "loss": 0.651, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 8.817852834740652, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 8.245311099788666e-06, | |
| "loss": 0.6546, | |
| "step": 3655 | |
| }, | |
| { | |
| "epoch": 8.829915560916767, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 8.07849626550531e-06, | |
| "loss": 0.656, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 8.841978287092884, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 7.91331512452983e-06, | |
| "loss": 0.6518, | |
| "step": 3665 | |
| }, | |
| { | |
| "epoch": 8.854041013268999, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 7.74977061257709e-06, | |
| "loss": 0.6529, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 8.866103739445114, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 7.587865636274594e-06, | |
| "loss": 0.6492, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 8.87816646562123, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 7.427603073110967e-06, | |
| "loss": 0.6535, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 8.890229191797346, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 7.268985771384618e-06, | |
| "loss": 0.651, | |
| "step": 3685 | |
| }, | |
| { | |
| "epoch": 8.902291917973463, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 7.1120165501533e-06, | |
| "loss": 0.6458, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 8.914354644149578, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 6.956698199183864e-06, | |
| "loss": 0.6505, | |
| "step": 3695 | |
| }, | |
| { | |
| "epoch": 8.926417370325694, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 6.803033478902765e-06, | |
| "loss": 0.6451, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 8.93848009650181, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 6.651025120346988e-06, | |
| "loss": 0.655, | |
| "step": 3705 | |
| }, | |
| { | |
| "epoch": 8.950542822677924, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.500675825115454e-06, | |
| "loss": 0.6497, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 8.962605548854041, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 6.351988265321129e-06, | |
| "loss": 0.6511, | |
| "step": 3715 | |
| }, | |
| { | |
| "epoch": 8.974668275030156, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 6.204965083543368e-06, | |
| "loss": 0.6494, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 8.986731001206273, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 6.059608892781088e-06, | |
| "loss": 0.6477, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 8.998793727382388, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 5.915922276406249e-06, | |
| "loss": 0.643, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 8.998793727382388, | |
| "eval_loss": 2.4790937900543213, | |
| "eval_runtime": 0.2579, | |
| "eval_samples_per_second": 38.775, | |
| "eval_steps_per_second": 3.877, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 9.010856453558505, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 5.77390778811796e-06, | |
| "loss": 0.65, | |
| "step": 3735 | |
| }, | |
| { | |
| "epoch": 9.02291917973462, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 5.633567951897145e-06, | |
| "loss": 0.6517, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 9.034981905910735, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 5.494905261961581e-06, | |
| "loss": 0.6468, | |
| "step": 3745 | |
| }, | |
| { | |
| "epoch": 9.047044632086852, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 5.357922182721687e-06, | |
| "loss": 0.6557, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 9.059107358262967, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 5.222621148736595e-06, | |
| "loss": 0.6555, | |
| "step": 3755 | |
| }, | |
| { | |
| "epoch": 9.071170084439084, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 5.08900456467103e-06, | |
| "loss": 0.6509, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 9.083232810615199, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 4.957074805252437e-06, | |
| "loss": 0.6506, | |
| "step": 3765 | |
| }, | |
| { | |
| "epoch": 9.095295536791316, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 4.826834215228826e-06, | |
| "loss": 0.6464, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 9.10735826296743, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.698285109327161e-06, | |
| "loss": 0.6515, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 9.119420989143546, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.5714297722121106e-06, | |
| "loss": 0.6493, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 9.131483715319662, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.446270458445545e-06, | |
| "loss": 0.6508, | |
| "step": 3785 | |
| }, | |
| { | |
| "epoch": 9.143546441495777, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 4.322809392446392e-06, | |
| "loss": 0.6547, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 9.155609167671894, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 4.20104876845111e-06, | |
| "loss": 0.6499, | |
| "step": 3795 | |
| }, | |
| { | |
| "epoch": 9.16767189384801, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 4.080990750474778e-06, | |
| "loss": 0.6468, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 9.179734620024126, | |
| "grad_norm": 0.375, | |
| "learning_rate": 3.962637472272501e-06, | |
| "loss": 0.6497, | |
| "step": 3805 | |
| }, | |
| { | |
| "epoch": 9.191797346200241, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.84599103730161e-06, | |
| "loss": 0.6578, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 9.203860072376358, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.73105351868418e-06, | |
| "loss": 0.6466, | |
| "step": 3815 | |
| }, | |
| { | |
| "epoch": 9.215922798552473, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 3.617826959170256e-06, | |
| "loss": 0.6528, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 9.227985524728588, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 3.5063133711014882e-06, | |
| "loss": 0.6565, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 9.240048250904705, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 3.3965147363754555e-06, | |
| "loss": 0.6544, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 9.25211097708082, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 3.2884330064103297e-06, | |
| "loss": 0.6532, | |
| "step": 3835 | |
| }, | |
| { | |
| "epoch": 9.264173703256937, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 3.182070102110257e-06, | |
| "loss": 0.6519, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 9.276236429433052, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 3.0774279138312657e-06, | |
| "loss": 0.6529, | |
| "step": 3845 | |
| }, | |
| { | |
| "epoch": 9.288299155609169, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 2.974508301347534e-06, | |
| "loss": 0.651, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 9.300361881785284, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.873313093818486e-06, | |
| "loss": 0.6533, | |
| "step": 3855 | |
| }, | |
| { | |
| "epoch": 9.312424607961399, | |
| "grad_norm": 0.35546875, | |
| "learning_rate": 2.7738440897561723e-06, | |
| "loss": 0.6514, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 9.324487334137515, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.676103056993362e-06, | |
| "loss": 0.6581, | |
| "step": 3865 | |
| }, | |
| { | |
| "epoch": 9.33655006031363, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 2.580091732652101e-06, | |
| "loss": 0.6529, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 9.348612786489747, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.485811823112849e-06, | |
| "loss": 0.6552, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 9.360675512665862, | |
| "grad_norm": 0.39453125, | |
| "learning_rate": 2.3932650039841687e-06, | |
| "loss": 0.6576, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 9.372738238841979, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 2.302452920072895e-06, | |
| "loss": 0.6513, | |
| "step": 3885 | |
| }, | |
| { | |
| "epoch": 9.384800965018094, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 2.213377185354959e-06, | |
| "loss": 0.648, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 9.39686369119421, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 2.126039382946676e-06, | |
| "loss": 0.6528, | |
| "step": 3895 | |
| }, | |
| { | |
| "epoch": 9.408926417370326, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.0404410650765817e-06, | |
| "loss": 0.646, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 9.420989143546441, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 1.9565837530579166e-06, | |
| "loss": 0.654, | |
| "step": 3905 | |
| }, | |
| { | |
| "epoch": 9.433051869722558, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 1.874468937261531e-06, | |
| "loss": 0.6514, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 9.445114595898673, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.7940980770894122e-06, | |
| "loss": 0.6522, | |
| "step": 3915 | |
| }, | |
| { | |
| "epoch": 9.45717732207479, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 1.71547260094872e-06, | |
| "loss": 0.6441, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 9.469240048250905, | |
| "grad_norm": 0.373046875, | |
| "learning_rate": 1.6385939062264822e-06, | |
| "loss": 0.6546, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 9.48130277442702, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.5634633592646609e-06, | |
| "loss": 0.6493, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 9.493365500603137, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 1.490082295335926e-06, | |
| "loss": 0.6557, | |
| "step": 3935 | |
| }, | |
| { | |
| "epoch": 9.505428226779252, | |
| "grad_norm": 0.404296875, | |
| "learning_rate": 1.4184520186199202e-06, | |
| "loss": 0.6489, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 9.517490952955368, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 1.348573802180053e-06, | |
| "loss": 0.6515, | |
| "step": 3945 | |
| }, | |
| { | |
| "epoch": 9.529553679131483, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.2804488879408993e-06, | |
| "loss": 0.6514, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 9.5416164053076, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 1.214078486666137e-06, | |
| "loss": 0.6548, | |
| "step": 3955 | |
| }, | |
| { | |
| "epoch": 9.553679131483715, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 1.1494637779369766e-06, | |
| "loss": 0.6505, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 9.56574185765983, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 1.0866059101312553e-06, | |
| "loss": 0.6445, | |
| "step": 3965 | |
| }, | |
| { | |
| "epoch": 9.577804583835947, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.0255060004030093e-06, | |
| "loss": 0.6528, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 9.589867310012062, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 9.661651346625889e-07, | |
| "loss": 0.6577, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 9.601930036188179, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 9.085843675574079e-07, | |
| "loss": 0.642, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 9.613992762364294, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 8.527647224531699e-07, | |
| "loss": 0.6515, | |
| "step": 3985 | |
| }, | |
| { | |
| "epoch": 9.62605548854041, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 7.987071914156596e-07, | |
| "loss": 0.6547, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 9.638118214716526, | |
| "grad_norm": 0.3515625, | |
| "learning_rate": 7.464127351931805e-07, | |
| "loss": 0.6513, | |
| "step": 3995 | |
| }, | |
| { | |
| "epoch": 9.650180940892643, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 6.958822831994005e-07, | |
| "loss": 0.6486, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.662243667068758, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 6.471167334968886e-07, | |
| "loss": 0.6516, | |
| "step": 4005 | |
| }, | |
| { | |
| "epoch": 9.674306393244873, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 6.001169527811268e-07, | |
| "loss": 0.6508, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 9.68636911942099, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 5.548837763651115e-07, | |
| "loss": 0.6512, | |
| "step": 4015 | |
| }, | |
| { | |
| "epoch": 9.698431845597105, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 5.114180081645214e-07, | |
| "loss": 0.6516, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 9.710494571773221, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 4.6972042068341714e-07, | |
| "loss": 0.6506, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 9.722557297949336, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 4.2979175500050817e-07, | |
| "loss": 0.6589, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 9.734620024125451, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 3.9163272075599664e-07, | |
| "loss": 0.6506, | |
| "step": 4035 | |
| }, | |
| { | |
| "epoch": 9.746682750301568, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 3.552439961389431e-07, | |
| "loss": 0.6492, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 9.758745476477683, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 3.206262278752314e-07, | |
| "loss": 0.6463, | |
| "step": 4045 | |
| }, | |
| { | |
| "epoch": 9.7708082026538, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 2.877800312160783e-07, | |
| "loss": 0.657, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 9.782870928829915, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 2.5670598992707516e-07, | |
| "loss": 0.6509, | |
| "step": 4055 | |
| }, | |
| { | |
| "epoch": 9.794933655006032, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.274046562778409e-07, | |
| "loss": 0.6568, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 9.806996381182147, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.9987655103217428e-07, | |
| "loss": 0.6505, | |
| "step": 4065 | |
| }, | |
| { | |
| "epoch": 9.819059107358264, | |
| "grad_norm": 0.361328125, | |
| "learning_rate": 1.7412216343885014e-07, | |
| "loss": 0.6493, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 9.831121833534379, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.5014195122287078e-07, | |
| "loss": 0.6527, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 9.843184559710494, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 1.2793634057732818e-07, | |
| "loss": 0.656, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 9.85524728588661, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 1.0750572615590982e-07, | |
| "loss": 0.6508, | |
| "step": 4085 | |
| }, | |
| { | |
| "epoch": 9.867310012062726, | |
| "grad_norm": 0.349609375, | |
| "learning_rate": 8.885047106578227e-08, | |
| "loss": 0.6539, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 9.879372738238843, | |
| "grad_norm": 0.34765625, | |
| "learning_rate": 7.197090686119623e-08, | |
| "loss": 0.6515, | |
| "step": 4095 | |
| }, | |
| { | |
| "epoch": 9.891435464414958, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 5.6867333537580226e-08, | |
| "loss": 0.6523, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 9.903498190591074, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 4.354001952621145e-08, | |
| "loss": 0.6545, | |
| "step": 4105 | |
| }, | |
| { | |
| "epoch": 9.91556091676719, | |
| "grad_norm": 0.359375, | |
| "learning_rate": 3.1989201689452967e-08, | |
| "loss": 0.6485, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 9.927623642943304, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 2.221508531652372e-08, | |
| "loss": 0.6493, | |
| "step": 4115 | |
| }, | |
| { | |
| "epoch": 9.939686369119421, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 1.4217844119857048e-08, | |
| "loss": 0.6523, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 9.951749095295536, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 7.997620232014225e-09, | |
| "loss": 0.6501, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 9.963811821471653, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 3.554524203175369e-09, | |
| "loss": 0.6438, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 9.975874547647768, | |
| "grad_norm": 0.40625, | |
| "learning_rate": 8.886349991521492e-10, | |
| "loss": 0.6541, | |
| "step": 4135 | |
| }, | |
| { | |
| "epoch": 9.987937273823885, | |
| "grad_norm": 0.365234375, | |
| "learning_rate": 0.0, | |
| "loss": 0.6511, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 9.987937273823885, | |
| "eval_loss": 2.4690446853637695, | |
| "eval_runtime": 0.2371, | |
| "eval_samples_per_second": 42.169, | |
| "eval_steps_per_second": 4.217, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 9.987937273823885, | |
| "step": 4140, | |
| "total_flos": 1.263927943748177e+19, | |
| "train_loss": 1.320238353146447, | |
| "train_runtime": 10095.6387, | |
| "train_samples_per_second": 26.261, | |
| "train_steps_per_second": 0.41 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 4140, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 1.263927943748177e+19, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |