| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9996613613274636, | |
| "eval_steps": 500, | |
| "global_step": 1476, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006772773450728074, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 1.3513513513513515e-07, | |
| "loss": 2.4477, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0033863867253640365, | |
| "grad_norm": 11.75, | |
| "learning_rate": 6.756756756756758e-07, | |
| "loss": 2.4743, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006772773450728073, | |
| "grad_norm": 12.5625, | |
| "learning_rate": 1.3513513513513515e-06, | |
| "loss": 2.465, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01015916017609211, | |
| "grad_norm": 11.375, | |
| "learning_rate": 2.0270270270270273e-06, | |
| "loss": 2.4495, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.013545546901456146, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 2.4107, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.016931933626820182, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 3.3783783783783788e-06, | |
| "loss": 2.4134, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02031832035218422, | |
| "grad_norm": 2.96875, | |
| "learning_rate": 4.0540540540540545e-06, | |
| "loss": 2.3642, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.023704707077548254, | |
| "grad_norm": 2.25, | |
| "learning_rate": 4.72972972972973e-06, | |
| "loss": 2.3561, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.027091093802912292, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 2.3428, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03047748052827633, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 6.081081081081082e-06, | |
| "loss": 2.2938, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.033863867253640365, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 2.2739, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0372502539790044, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 7.4324324324324324e-06, | |
| "loss": 2.2748, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04063664070436844, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 2.2662, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04402302742973248, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 8.783783783783785e-06, | |
| "loss": 2.2491, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04740941415509651, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 9.45945945945946e-06, | |
| "loss": 2.2262, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05079580088046055, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 1.0135135135135136e-05, | |
| "loss": 2.2425, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.054182187605824585, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 1.0810810810810812e-05, | |
| "loss": 2.224, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05756857433118862, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.1486486486486488e-05, | |
| "loss": 2.1884, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.06095496105655266, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.2162162162162164e-05, | |
| "loss": 2.2151, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06434134778191669, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.283783783783784e-05, | |
| "loss": 2.1939, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06772773450728073, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.3513513513513515e-05, | |
| "loss": 2.1887, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07111412123264477, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.4189189189189189e-05, | |
| "loss": 2.1825, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0745005079580088, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.4864864864864865e-05, | |
| "loss": 2.1925, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07788689468337284, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.554054054054054e-05, | |
| "loss": 2.1969, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.08127328140873688, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.6216216216216218e-05, | |
| "loss": 2.1783, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.08465966813410092, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.6891891891891896e-05, | |
| "loss": 2.1641, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08804605485946496, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.756756756756757e-05, | |
| "loss": 2.2032, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.091432441584829, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.8243243243243244e-05, | |
| "loss": 2.2021, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.09481882831019302, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.891891891891892e-05, | |
| "loss": 2.1589, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09820521503555706, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.9594594594594595e-05, | |
| "loss": 2.1598, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1015916017609211, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.9999888073536733e-05, | |
| "loss": 2.1588, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.10497798848628513, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.9998628929598956e-05, | |
| "loss": 2.1711, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.10836437521164917, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.9995970910394228e-05, | |
| "loss": 2.1901, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11175076193701321, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.9991914387797266e-05, | |
| "loss": 2.1554, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.11513714866237724, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.99864599293428e-05, | |
| "loss": 2.1558, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11852353538774128, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.997960829814616e-05, | |
| "loss": 2.1718, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.12190992211310532, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.9971360452796523e-05, | |
| "loss": 2.171, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.12529630883846934, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.9961717547222775e-05, | |
| "loss": 2.1686, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.12868269556383338, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.9950680930532107e-05, | |
| "loss": 2.1569, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.13206908228919742, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.9938252146821236e-05, | |
| "loss": 2.1832, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.13545546901456146, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.9924432934960384e-05, | |
| "loss": 2.1616, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1388418557399255, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.9909225228350007e-05, | |
| "loss": 2.1798, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.14222824246528953, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.989263115465028e-05, | |
| "loss": 2.1645, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.14561462919065357, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.987465303548345e-05, | |
| "loss": 2.1599, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.1490010159160176, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.9855293386108995e-05, | |
| "loss": 2.1525, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.15238740264138165, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.9834554915071745e-05, | |
| "loss": 2.1711, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.15577378936674569, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.981244052382293e-05, | |
| "loss": 2.1597, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.15916017609210972, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.978895330631425e-05, | |
| "loss": 2.1511, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.16254656281747376, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.976409654856501e-05, | |
| "loss": 2.1579, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.1659329495428378, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.9737873728202376e-05, | |
| "loss": 2.1586, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.16931933626820184, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.9710288513974846e-05, | |
| "loss": 2.1634, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.17270572299356587, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.9681344765238958e-05, | |
| "loss": 2.1667, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.1760921097189299, | |
| "grad_norm": 1.25, | |
| "learning_rate": 1.9651046531419335e-05, | |
| "loss": 2.1732, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.17947849644429395, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.961939805144216e-05, | |
| "loss": 2.1726, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.182864883169658, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.9586403753142104e-05, | |
| "loss": 2.1538, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.186251269895022, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.9552068252642858e-05, | |
| "loss": 2.1728, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.18963765662038604, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.9516396353711297e-05, | |
| "loss": 2.1467, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.19302404334575007, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.9479393047085392e-05, | |
| "loss": 2.1676, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.1964104300711141, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.9441063509776003e-05, | |
| "loss": 2.1476, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.19979681679647815, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.9401413104342535e-05, | |
| "loss": 2.1346, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.2031832035218422, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.936044737814273e-05, | |
| "loss": 2.1629, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.20656959024720623, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.931817206255651e-05, | |
| "loss": 2.1322, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.20995597697257026, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.9274593072184152e-05, | |
| "loss": 2.1842, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2133423636979343, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.922971650401877e-05, | |
| "loss": 2.133, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.21672875042329834, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.9183548636593322e-05, | |
| "loss": 2.1499, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.22011513714866238, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.9136095929102204e-05, | |
| "loss": 2.1444, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.22350152387402641, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.908736502049754e-05, | |
| "loss": 2.1458, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.22688791059939045, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.903736272856038e-05, | |
| "loss": 2.1738, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.2302742973247545, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.8986096048946826e-05, | |
| "loss": 2.1413, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.23366068405011853, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.893357215420929e-05, | |
| "loss": 2.1614, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.23704707077548257, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.8879798392793033e-05, | |
| "loss": 2.1582, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2404334575008466, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.8824782288008038e-05, | |
| "loss": 2.1545, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.24381984422621064, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.8768531536976452e-05, | |
| "loss": 2.1497, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.24720623095157468, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.8711054009555736e-05, | |
| "loss": 2.1677, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.2505926176769387, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.865235774723758e-05, | |
| "loss": 2.1475, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.25397900440230275, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.8592450962022872e-05, | |
| "loss": 2.1422, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.25736539112766676, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.8531342035272768e-05, | |
| "loss": 2.1679, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.26075177785303083, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.84690395165361e-05, | |
| "loss": 2.1473, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.26413816457839484, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.8405552122353212e-05, | |
| "loss": 2.1614, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2675245513037589, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.8340888735036485e-05, | |
| "loss": 2.1299, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2709109380291229, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.8275058401427622e-05, | |
| "loss": 2.1553, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.274297324754487, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.8208070331631943e-05, | |
| "loss": 2.1445, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.277683711479851, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.8139933897729833e-05, | |
| "loss": 2.1562, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.28107009820521506, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.807065863246551e-05, | |
| "loss": 2.1343, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.28445648493057907, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.8000254227913346e-05, | |
| "loss": 2.1468, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.28784287165594313, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.7928730534121872e-05, | |
| "loss": 2.1492, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.29122925838130714, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.7856097557735697e-05, | |
| "loss": 2.1326, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.2946156451066712, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.77823654605955e-05, | |
| "loss": 2.1491, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.2980020318320352, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.7707544558316332e-05, | |
| "loss": 2.1425, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.30138841855739923, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.763164531884439e-05, | |
| "loss": 2.1356, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.3047748052827633, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.7554678360992475e-05, | |
| "loss": 2.1459, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3081611920081273, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.7476654452954354e-05, | |
| "loss": 2.1232, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.31154757873349137, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.7397584510798208e-05, | |
| "loss": 2.1333, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3149339654588554, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.7317479596939426e-05, | |
| "loss": 2.145, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.31832035218421945, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.723635091859287e-05, | |
| "loss": 2.1452, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.32170673890958346, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.7154209826204916e-05, | |
| "loss": 2.155, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.3250931256349475, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.7071067811865477e-05, | |
| "loss": 2.15, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.32847951236031153, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.698693650770014e-05, | |
| "loss": 2.1436, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.3318658990856756, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.690182768424279e-05, | |
| "loss": 2.1471, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.3352522858110396, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.6815753248788814e-05, | |
| "loss": 2.1616, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.3386386725364037, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.672872524372919e-05, | |
| "loss": 2.1325, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3420250592617677, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.6640755844865693e-05, | |
| "loss": 2.1278, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.34541144598713175, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.6551857359707407e-05, | |
| "loss": 2.1371, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.34879783271249576, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.6462042225748803e-05, | |
| "loss": 2.1385, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.3521842194378598, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.637132300872969e-05, | |
| "loss": 2.1161, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.35557060616322383, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.627971240087715e-05, | |
| "loss": 2.1194, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.3589569928885879, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.6187223219129825e-05, | |
| "loss": 2.1456, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.3623433796139519, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.609386840334475e-05, | |
| "loss": 2.1475, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.365729766339316, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.5999661014486956e-05, | |
| "loss": 2.144, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.36911615306468, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.5904614232802173e-05, | |
| "loss": 2.1698, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.372502539790044, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 1.5808741355972808e-05, | |
| "loss": 2.143, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.37588892651540806, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.5712055797257542e-05, | |
| "loss": 2.1519, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.37927531324077207, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.5614571083614683e-05, | |
| "loss": 2.1537, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.38266169996613614, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.5516300853809688e-05, | |
| "loss": 2.1435, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.38604808669150015, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.5417258856506995e-05, | |
| "loss": 2.151, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.3894344734168642, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.5317458948346482e-05, | |
| "loss": 2.1489, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3928208601422282, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.5216915092004847e-05, | |
| "loss": 2.136, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.3962072468675923, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.5115641354242135e-05, | |
| "loss": 2.1468, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3995936335929563, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.5013651903933686e-05, | |
| "loss": 2.1242, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.40298002031832036, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.491096101008784e-05, | |
| "loss": 2.1387, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.4063664070436844, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.4807583039849589e-05, | |
| "loss": 2.1374, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.40975279376904844, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.4703532456490518e-05, | |
| "loss": 2.1485, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.41313918049441245, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.45988238173853e-05, | |
| "loss": 2.1299, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.4165255672197765, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.449347177197503e-05, | |
| "loss": 2.1299, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.4199119539451405, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.4387491059717653e-05, | |
| "loss": 2.1355, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.4232983406705046, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.4280896508025825e-05, | |
| "loss": 2.1481, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.4266847273958686, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.4173703030192467e-05, | |
| "loss": 2.1442, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.43007111412123267, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.4065925623304267e-05, | |
| "loss": 2.1465, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.4334575008465967, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.3957579366143521e-05, | |
| "loss": 2.1553, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.43684388757196074, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.3848679417078487e-05, | |
| "loss": 2.144, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.44023027429732475, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.373924101194264e-05, | |
| "loss": 2.1277, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.4436166610226888, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.3629279461903064e-05, | |
| "loss": 2.1534, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.44700304774805283, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.351881015131833e-05, | |
| "loss": 2.1327, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.45038943447341684, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.3407848535586122e-05, | |
| "loss": 2.1315, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.4537758211987809, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.3296410138980903e-05, | |
| "loss": 2.1231, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.4571622079241449, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.3184510552481988e-05, | |
| "loss": 2.1503, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.460548594649509, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.307216543159225e-05, | |
| "loss": 2.1274, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.463934981374873, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.2959390494147814e-05, | |
| "loss": 2.1279, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.46732136810023706, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.2846201518119019e-05, | |
| "loss": 2.1436, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.47070775482560107, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.273261433940299e-05, | |
| "loss": 2.1639, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.47409414155096513, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.2618644849608068e-05, | |
| "loss": 2.1305, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.47748052827632914, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.2504308993830488e-05, | |
| "loss": 2.1595, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.4808669150016932, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.2389622768423537e-05, | |
| "loss": 2.1304, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.4842533017270572, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.2274602218759569e-05, | |
| "loss": 2.1382, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.4876396884524213, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.2159263436985139e-05, | |
| "loss": 2.1279, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.4910260751777853, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.2043622559769614e-05, | |
| "loss": 2.1422, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.49441246190314936, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.192769576604754e-05, | |
| "loss": 2.1352, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.49779884862851337, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.1811499274755099e-05, | |
| "loss": 2.1342, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.5011852353538774, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.1695049342560969e-05, | |
| "loss": 2.149, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.5045716220792414, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.1578362261591916e-05, | |
| "loss": 2.1446, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.5079580088046055, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.1461454357153407e-05, | |
| "loss": 2.166, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5113443955299696, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.1344341985445594e-05, | |
| "loss": 2.1116, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.5147307822553335, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.1227041531274978e-05, | |
| "loss": 2.1335, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.5181171689806976, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.1109569405762051e-05, | |
| "loss": 2.1368, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.5215035557060617, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.0991942044045277e-05, | |
| "loss": 2.1468, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.5248899424314256, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.0874175902981711e-05, | |
| "loss": 2.1288, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.5282763291567897, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.075628745884457e-05, | |
| "loss": 2.1461, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5316627158821537, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.0638293205018084e-05, | |
| "loss": 2.1475, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.5350491026075178, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.0520209649689979e-05, | |
| "loss": 2.1375, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.5384354893328818, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.0402053313541854e-05, | |
| "loss": 2.1267, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.5418218760582458, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.0283840727437832e-05, | |
| "loss": 2.1394, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5452082627836099, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.0165588430111787e-05, | |
| "loss": 2.1366, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.548594649508974, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.0047312965853456e-05, | |
| "loss": 2.1366, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.5519810362343379, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 9.92903088219378e-06, | |
| "loss": 2.1493, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.555367422959702, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 9.810758727589814e-06, | |
| "loss": 2.1522, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.558753809685066, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 9.69251304910945e-06, | |
| "loss": 2.1298, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.5621401964104301, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 9.574310390116421e-06, | |
| "loss": 2.1347, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.5655265831357941, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 9.45616728795571e-06, | |
| "loss": 2.1401, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.5689129698611581, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 9.338100271639932e-06, | |
| "loss": 2.128, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.5722993565865222, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 9.220125859536766e-06, | |
| "loss": 2.1333, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.5756857433118863, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 9.102260557057936e-06, | |
| "loss": 2.1386, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.5790721300372502, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 8.984520854350014e-06, | |
| "loss": 2.1107, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.5824585167626143, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 8.866923223987303e-06, | |
| "loss": 2.1386, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.5858449034879784, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 8.749484118667242e-06, | |
| "loss": 2.1313, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.5892312902133424, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 8.632219968908556e-06, | |
| "loss": 2.1424, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.5926176769387064, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 8.515147180752504e-06, | |
| "loss": 2.1406, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.5960040636640704, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 8.398282133467579e-06, | |
| "loss": 2.1324, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5993904503894345, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 8.281641177257929e-06, | |
| "loss": 2.128, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.6027768371147985, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 8.165240630975863e-06, | |
| "loss": 2.1153, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.6061632238401625, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 8.04909677983872e-06, | |
| "loss": 2.1682, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.6095496105655266, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 7.93322587315047e-06, | |
| "loss": 2.1152, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6129359972908907, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 7.81764412202833e-06, | |
| "loss": 2.1311, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.6163223840162546, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 7.702367697134701e-06, | |
| "loss": 2.1408, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.6197087707416187, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 7.5874127264148075e-06, | |
| "loss": 2.145, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.6230951574669827, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 7.4727952928402695e-06, | |
| "loss": 2.1318, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.6264815441923468, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 7.358531432159002e-06, | |
| "loss": 2.1231, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.6298679309177108, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 7.2446371306516935e-06, | |
| "loss": 2.1285, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.6332543176430748, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 7.1311283228952345e-06, | |
| "loss": 2.1384, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.6366407043684389, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 7.018020889533348e-06, | |
| "loss": 2.1409, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.640027091093803, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 6.905330655054782e-06, | |
| "loss": 2.1417, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.6434134778191669, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 6.793073385579373e-06, | |
| "loss": 2.1375, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.646799864544531, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 6.681264786652236e-06, | |
| "loss": 2.1219, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.650186251269895, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 6.569920501046474e-06, | |
| "loss": 2.1363, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.6535726379952591, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 6.459056106574646e-06, | |
| "loss": 2.1209, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.6569590247206231, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 6.348687113909304e-06, | |
| "loss": 2.1254, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.6603454114459871, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 6.238828964412989e-06, | |
| "loss": 2.1466, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.6637317981713512, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 6.129497027977829e-06, | |
| "loss": 2.1163, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.6671181848967151, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 6.020706600875237e-06, | |
| "loss": 2.1078, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.6705045716220792, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 5.912472903615821e-06, | |
| "loss": 2.1486, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.6738909583474433, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 5.804811078819956e-06, | |
| "loss": 2.1399, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.6772773450728073, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 5.69773618909923e-06, | |
| "loss": 2.137, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6806637317981713, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 5.591263214949059e-06, | |
| "loss": 2.1506, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.6840501185235354, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 5.4854070526528445e-06, | |
| "loss": 2.142, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.6874365052488994, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 5.380182512197873e-06, | |
| "loss": 2.1231, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.6908228919742635, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 5.2756043152032934e-06, | |
| "loss": 2.1146, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.6942092786996275, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 5.171687092860464e-06, | |
| "loss": 2.1446, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.6975956654249915, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 5.068445383885961e-06, | |
| "loss": 2.1376, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.7009820521503556, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 4.965893632487511e-06, | |
| "loss": 2.139, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.7043684388757196, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 4.86404618634314e-06, | |
| "loss": 2.1404, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.7077548256010836, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 4.762917294593851e-06, | |
| "loss": 2.1275, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.7111412123264477, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 4.662521105850056e-06, | |
| "loss": 2.1326, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.7145275990518117, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 4.562871666212107e-06, | |
| "loss": 2.1461, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.7179139857771758, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 4.463982917305155e-06, | |
| "loss": 2.1244, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.7213003725025398, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 4.365868694328599e-06, | |
| "loss": 2.1384, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.7246867592279038, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 4.2685427241204755e-06, | |
| "loss": 2.1444, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.7280731459532679, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 4.1720186232369595e-06, | |
| "loss": 2.1365, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.731459532678632, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 4.076309896047337e-06, | |
| "loss": 2.1302, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.7348459194039959, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 3.981429932844641e-06, | |
| "loss": 2.1665, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.73823230612936, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 3.887392007972265e-06, | |
| "loss": 2.1583, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.741618692854724, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 3.7942092779667904e-06, | |
| "loss": 2.1263, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.745005079580088, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 3.7018947797172864e-06, | |
| "loss": 2.1303, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7483914663054521, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 3.610461428641402e-06, | |
| "loss": 2.1155, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.7517778530308161, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 3.5199220168783564e-06, | |
| "loss": 2.1145, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.7551642397561802, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 3.4302892114992814e-06, | |
| "loss": 2.1419, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.7585506264815441, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 3.341575552734978e-06, | |
| "loss": 2.1448, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.7619370132069082, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 3.253793452221485e-06, | |
| "loss": 2.1375, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.7653233999322723, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 3.1669551912635874e-06, | |
| "loss": 2.1626, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.7687097866576363, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 3.0810729191165722e-06, | |
| "loss": 2.1256, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.7720961733830003, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 2.9961586512864947e-06, | |
| "loss": 2.1337, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.7754825601083644, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 2.9122242678490975e-06, | |
| "loss": 2.1191, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.7788689468337284, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 2.829281511787739e-06, | |
| "loss": 2.1523, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.7822553335590925, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 2.747341987350439e-06, | |
| "loss": 2.1494, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.7856417202844564, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 2.666417158426393e-06, | |
| "loss": 2.1255, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.7890281070098205, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 2.586518346942082e-06, | |
| "loss": 2.136, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.7924144937351846, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 2.5076567312772635e-06, | |
| "loss": 2.1349, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.7958008804605486, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 2.4298433447010495e-06, | |
| "loss": 2.1331, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.7991872671859126, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 2.353089073828255e-06, | |
| "loss": 2.1608, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.8025736539112767, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 2.2774046570963147e-06, | |
| "loss": 2.1455, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.8059600406366407, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 2.202800683262888e-06, | |
| "loss": 2.1323, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.8093464273620048, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 2.1292875899244237e-06, | |
| "loss": 2.1235, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.8127328140873687, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 2.056875662055874e-06, | |
| "loss": 2.1479, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8161192008127328, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.985575030571746e-06, | |
| "loss": 2.1424, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.8195055875380969, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.915395670908734e-06, | |
| "loss": 2.147, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.8228919742634608, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.8463474016300831e-06, | |
| "loss": 2.1461, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.8262783609888249, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.7784398830519002e-06, | |
| "loss": 2.1333, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.829664747714189, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.7116826158916212e-06, | |
| "loss": 2.1438, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.833051134439553, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.6460849399387845e-06, | |
| "loss": 2.1339, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.836437521164917, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.5816560327483466e-06, | |
| "loss": 2.1507, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.839823907890281, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.5184049083566688e-06, | |
| "loss": 2.132, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.8432102946156451, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.4563404160204008e-06, | |
| "loss": 2.1273, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.8465966813410092, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.3954712389783998e-06, | |
| "loss": 2.1127, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.8499830680663731, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.3358058932369012e-06, | |
| "loss": 2.1482, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.8533694547917372, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.2773527263780626e-06, | |
| "loss": 2.1329, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.8567558415171013, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.2201199163920829e-06, | |
| "loss": 2.1414, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.8601422282424653, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.1641154705330503e-06, | |
| "loss": 2.141, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.8635286149678293, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.1093472241986636e-06, | |
| "loss": 2.1145, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.8669150016931934, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.0558228398340188e-06, | |
| "loss": 2.1221, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.8703013884185574, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.0035498058595838e-06, | |
| "loss": 2.1412, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.8736877751439215, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 9.525354356235006e-07, | |
| "loss": 2.1334, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.8770741618692854, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 9.027868663784189e-07, | |
| "loss": 2.1355, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.8804605485946495, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 8.543110582829272e-07, | |
| "loss": 2.1282, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8838469353200136, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 8.071147934278045e-07, | |
| "loss": 2.138, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.8872333220453776, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 7.612046748871327e-07, | |
| "loss": 2.1461, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.8906197087707416, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 7.165871257945046e-07, | |
| "loss": 2.1359, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.8940060954961057, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 6.732683884443736e-07, | |
| "loss": 2.1379, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.8973924822214697, | |
| "grad_norm": 1.125, | |
| "learning_rate": 6.312545234187106e-07, | |
| "loss": 2.1351, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.9007788689468337, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 5.905514087391051e-07, | |
| "loss": 2.15, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.9041652556721977, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 5.511647390443697e-07, | |
| "loss": 2.1279, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.9075516423975618, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 5.131000247938367e-07, | |
| "loss": 2.1455, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.9109380291229259, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 4.7636259149639983e-07, | |
| "loss": 2.1343, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.9143244158482898, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 4.4095757896544747e-07, | |
| "loss": 2.1377, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.9177108025736539, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 4.0688994059976175e-07, | |
| "loss": 2.1354, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.921097189299018, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 3.7416444269050335e-07, | |
| "loss": 2.1452, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.924483576024382, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 3.427856637543825e-07, | |
| "loss": 2.1494, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.927869962749746, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 3.127579938930891e-07, | |
| "loss": 2.1434, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.93125634947511, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 2.8408563417908875e-07, | |
| "loss": 2.1375, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.9346427362004741, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 2.5677259606786686e-07, | |
| "loss": 2.1249, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.9380291229258382, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 2.3082270083669255e-07, | |
| "loss": 2.1593, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.9414155096512021, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 2.0623957905000603e-07, | |
| "loss": 2.1368, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.9448018963765662, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.8302667005146936e-07, | |
| "loss": 2.1378, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.9481882831019303, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.6118722148278586e-07, | |
| "loss": 2.1432, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9515746698272943, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.40724288829327e-07, | |
| "loss": 2.1343, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.9549610565526583, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.2164073499265404e-07, | |
| "loss": 2.1412, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.9583474432780223, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.0393922988997995e-07, | |
| "loss": 2.1336, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.9617338300033864, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 8.762225008062675e-08, | |
| "loss": 2.1514, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.9651202167287504, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 7.269207841953618e-08, | |
| "loss": 2.121, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.9685066034541144, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 5.915080373788962e-08, | |
| "loss": 2.1249, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.9718929901794785, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 4.700032055086179e-08, | |
| "loss": 2.1442, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.9752793769048426, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 3.6242328792567286e-08, | |
| "loss": 2.1205, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.9786657636302065, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 2.6878333578227446e-08, | |
| "loss": 2.1349, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.9820521503555706, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 1.890964499359327e-08, | |
| "loss": 2.1394, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.9854385370809347, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.2337377911660897e-08, | |
| "loss": 2.1281, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.9888249238062987, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 7.162451836685291e-09, | |
| "loss": 2.1459, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.9922113105316627, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 3.385590775544234e-09, | |
| "loss": 2.1328, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.9955976972570267, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 1.007323136438254e-09, | |
| "loss": 2.1426, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.9989840839823908, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 2.7981654966424686e-11, | |
| "loss": 2.1307, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.9996613613274636, | |
| "eval_loss": 2.1349875926971436, | |
| "eval_runtime": 82.2944, | |
| "eval_samples_per_second": 15.858, | |
| "eval_steps_per_second": 1.993, | |
| "step": 1476 | |
| }, | |
| { | |
| "epoch": 0.9996613613274636, | |
| "step": 1476, | |
| "total_flos": 3.002086743311647e+17, | |
| "train_loss": 2.154247896135015, | |
| "train_runtime": 6599.6501, | |
| "train_samples_per_second": 3.579, | |
| "train_steps_per_second": 0.224 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1476, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.002086743311647e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |