{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.998984083982391, "eval_steps": 500, "global_step": 4428, "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": 4.514672686230248e-08, "loss": 2.4477, "step": 1 }, { "epoch": 0.0033863867253640365, "grad_norm": 11.75, "learning_rate": 2.2573363431151243e-07, "loss": 2.4743, "step": 5 }, { "epoch": 0.006772773450728073, "grad_norm": 12.625, "learning_rate": 4.5146726862302486e-07, "loss": 2.4675, "step": 10 }, { "epoch": 0.01015916017609211, "grad_norm": 11.8125, "learning_rate": 6.772009029345373e-07, "loss": 2.4607, "step": 15 }, { "epoch": 0.013545546901456146, "grad_norm": 11.9375, "learning_rate": 9.029345372460497e-07, "loss": 2.4413, "step": 20 }, { "epoch": 0.016931933626820182, "grad_norm": 11.4375, "learning_rate": 1.1286681715575621e-06, "loss": 2.4661, "step": 25 }, { "epoch": 0.02031832035218422, "grad_norm": 10.6875, "learning_rate": 1.3544018058690746e-06, "loss": 2.4284, "step": 30 }, { "epoch": 0.023704707077548254, "grad_norm": 9.0625, "learning_rate": 1.5801354401805871e-06, "loss": 2.4384, "step": 35 }, { "epoch": 0.027091093802912292, "grad_norm": 7.53125, "learning_rate": 1.8058690744920994e-06, "loss": 2.4398, "step": 40 }, { "epoch": 0.03047748052827633, "grad_norm": 5.40625, "learning_rate": 2.0316027088036117e-06, "loss": 2.4001, "step": 45 }, { "epoch": 0.033863867253640365, "grad_norm": 3.75, "learning_rate": 2.2573363431151243e-06, "loss": 2.3842, "step": 50 }, { "epoch": 0.0372502539790044, "grad_norm": 3.03125, "learning_rate": 2.4830699774266368e-06, "loss": 2.3908, "step": 55 }, { "epoch": 0.04063664070436844, "grad_norm": 2.609375, "learning_rate": 2.7088036117381493e-06, "loss": 2.3855, "step": 60 }, { "epoch": 0.04402302742973248, "grad_norm": 2.484375, "learning_rate": 2.9345372460496618e-06, "loss": 2.3649, "step": 65 }, { "epoch": 0.04740941415509651, "grad_norm": 2.28125, "learning_rate": 3.1602708803611743e-06, "loss": 2.3336, "step": 70 }, { "epoch": 0.05079580088046055, "grad_norm": 1.9140625, "learning_rate": 3.386004514672687e-06, "loss": 2.3418, "step": 75 }, { "epoch": 0.054182187605824585, "grad_norm": 1.765625, "learning_rate": 3.611738148984199e-06, "loss": 2.3162, "step": 80 }, { "epoch": 0.05756857433118862, "grad_norm": 1.796875, "learning_rate": 3.837471783295712e-06, "loss": 2.274, "step": 85 }, { "epoch": 0.06095496105655266, "grad_norm": 1.6875, "learning_rate": 4.0632054176072235e-06, "loss": 2.2926, "step": 90 }, { "epoch": 0.06434134778191669, "grad_norm": 1.859375, "learning_rate": 4.288939051918736e-06, "loss": 2.2671, "step": 95 }, { "epoch": 0.06772773450728073, "grad_norm": 1.578125, "learning_rate": 4.5146726862302485e-06, "loss": 2.255, "step": 100 }, { "epoch": 0.07111412123264477, "grad_norm": 1.546875, "learning_rate": 4.740406320541761e-06, "loss": 2.2431, "step": 105 }, { "epoch": 0.0745005079580088, "grad_norm": 1.4765625, "learning_rate": 4.9661399548532735e-06, "loss": 2.2486, "step": 110 }, { "epoch": 0.07788689468337284, "grad_norm": 1.5, "learning_rate": 5.191873589164786e-06, "loss": 2.249, "step": 115 }, { "epoch": 0.08127328140873688, "grad_norm": 1.3984375, "learning_rate": 5.4176072234762986e-06, "loss": 2.2278, "step": 120 }, { "epoch": 0.08465966813410092, "grad_norm": 1.375, "learning_rate": 5.643340857787811e-06, "loss": 2.2095, "step": 125 }, { "epoch": 0.08804605485946496, "grad_norm": 1.375, "learning_rate": 5.8690744920993236e-06, "loss": 2.2445, "step": 130 }, { "epoch": 0.091432441584829, "grad_norm": 1.390625, "learning_rate": 6.094808126410836e-06, "loss": 2.2414, "step": 135 }, { "epoch": 0.09481882831019302, "grad_norm": 1.3671875, "learning_rate": 6.320541760722349e-06, "loss": 2.1974, "step": 140 }, { "epoch": 0.09820521503555706, "grad_norm": 1.3359375, "learning_rate": 6.546275395033861e-06, "loss": 2.1979, "step": 145 }, { "epoch": 0.1015916017609211, "grad_norm": 1.3203125, "learning_rate": 6.772009029345374e-06, "loss": 2.1945, "step": 150 }, { "epoch": 0.10497798848628513, "grad_norm": 1.3828125, "learning_rate": 6.997742663656886e-06, "loss": 2.2061, "step": 155 }, { "epoch": 0.10836437521164917, "grad_norm": 1.328125, "learning_rate": 7.223476297968398e-06, "loss": 2.2234, "step": 160 }, { "epoch": 0.11175076193701321, "grad_norm": 1.34375, "learning_rate": 7.44920993227991e-06, "loss": 2.1885, "step": 165 }, { "epoch": 0.11513714866237724, "grad_norm": 1.3046875, "learning_rate": 7.674943566591424e-06, "loss": 2.187, "step": 170 }, { "epoch": 0.11852353538774128, "grad_norm": 1.40625, "learning_rate": 7.900677200902936e-06, "loss": 2.2022, "step": 175 }, { "epoch": 0.12190992211310532, "grad_norm": 1.359375, "learning_rate": 8.126410835214447e-06, "loss": 2.2013, "step": 180 }, { "epoch": 0.12529630883846934, "grad_norm": 1.3515625, "learning_rate": 8.35214446952596e-06, "loss": 2.1984, "step": 185 }, { "epoch": 0.12868269556383338, "grad_norm": 1.3125, "learning_rate": 8.577878103837472e-06, "loss": 2.1861, "step": 190 }, { "epoch": 0.13206908228919742, "grad_norm": 1.3359375, "learning_rate": 8.803611738148985e-06, "loss": 2.2115, "step": 195 }, { "epoch": 0.13545546901456146, "grad_norm": 1.3359375, "learning_rate": 9.029345372460497e-06, "loss": 2.1902, "step": 200 }, { "epoch": 0.1388418557399255, "grad_norm": 1.2890625, "learning_rate": 9.25507900677201e-06, "loss": 2.2061, "step": 205 }, { "epoch": 0.14222824246528953, "grad_norm": 1.3046875, "learning_rate": 9.480812641083522e-06, "loss": 2.1917, "step": 210 }, { "epoch": 0.14561462919065357, "grad_norm": 1.3046875, "learning_rate": 9.706546275395035e-06, "loss": 2.1859, "step": 215 }, { "epoch": 0.1490010159160176, "grad_norm": 1.265625, "learning_rate": 9.932279909706547e-06, "loss": 2.1786, "step": 220 }, { "epoch": 0.15238740264138165, "grad_norm": 1.3515625, "learning_rate": 1.015801354401806e-05, "loss": 2.1978, "step": 225 }, { "epoch": 0.15577378936674569, "grad_norm": 1.2890625, "learning_rate": 1.0383747178329572e-05, "loss": 2.1843, "step": 230 }, { "epoch": 0.15916017609210972, "grad_norm": 1.296875, "learning_rate": 1.0609480812641085e-05, "loss": 2.1768, "step": 235 }, { "epoch": 0.16254656281747376, "grad_norm": 1.3515625, "learning_rate": 1.0835214446952597e-05, "loss": 2.1826, "step": 240 }, { "epoch": 0.1659329495428378, "grad_norm": 1.2734375, "learning_rate": 1.106094808126411e-05, "loss": 2.1822, "step": 245 }, { "epoch": 0.16931933626820184, "grad_norm": 1.265625, "learning_rate": 1.1286681715575622e-05, "loss": 2.1865, "step": 250 }, { "epoch": 0.17270572299356587, "grad_norm": 1.2421875, "learning_rate": 1.1512415349887135e-05, "loss": 2.1902, "step": 255 }, { "epoch": 0.1760921097189299, "grad_norm": 1.3125, "learning_rate": 1.1738148984198647e-05, "loss": 2.1975, "step": 260 }, { "epoch": 0.17947849644429395, "grad_norm": 1.2890625, "learning_rate": 1.1963882618510158e-05, "loss": 2.1951, "step": 265 }, { "epoch": 0.182864883169658, "grad_norm": 1.3203125, "learning_rate": 1.2189616252821672e-05, "loss": 2.1777, "step": 270 }, { "epoch": 0.186251269895022, "grad_norm": 1.296875, "learning_rate": 1.2415349887133183e-05, "loss": 2.1964, "step": 275 }, { "epoch": 0.18963765662038604, "grad_norm": 1.265625, "learning_rate": 1.2641083521444697e-05, "loss": 2.1692, "step": 280 }, { "epoch": 0.19302404334575007, "grad_norm": 1.296875, "learning_rate": 1.2866817155756208e-05, "loss": 2.1907, "step": 285 }, { "epoch": 0.1964104300711141, "grad_norm": 1.328125, "learning_rate": 1.3092550790067722e-05, "loss": 2.17, "step": 290 }, { "epoch": 0.19979681679647815, "grad_norm": 1.2578125, "learning_rate": 1.3318284424379233e-05, "loss": 2.1558, "step": 295 }, { "epoch": 0.2031832035218422, "grad_norm": 1.3125, "learning_rate": 1.3544018058690747e-05, "loss": 2.184, "step": 300 }, { "epoch": 0.20656959024720623, "grad_norm": 1.2734375, "learning_rate": 1.3769751693002258e-05, "loss": 2.1542, "step": 305 }, { "epoch": 0.20995597697257026, "grad_norm": 1.265625, "learning_rate": 1.3995485327313772e-05, "loss": 2.2053, "step": 310 }, { "epoch": 0.2133423636979343, "grad_norm": 1.2734375, "learning_rate": 1.4221218961625283e-05, "loss": 2.1536, "step": 315 }, { "epoch": 0.21672875042329834, "grad_norm": 1.25, "learning_rate": 1.4446952595936796e-05, "loss": 2.1703, "step": 320 }, { "epoch": 0.22011513714866238, "grad_norm": 1.3046875, "learning_rate": 1.4672686230248308e-05, "loss": 2.1635, "step": 325 }, { "epoch": 0.22350152387402641, "grad_norm": 1.2734375, "learning_rate": 1.489841986455982e-05, "loss": 2.167, "step": 330 }, { "epoch": 0.22688791059939045, "grad_norm": 1.2421875, "learning_rate": 1.5124153498871333e-05, "loss": 2.193, "step": 335 }, { "epoch": 0.2302742973247545, "grad_norm": 1.328125, "learning_rate": 1.5349887133182847e-05, "loss": 2.1613, "step": 340 }, { "epoch": 0.23366068405011853, "grad_norm": 1.2109375, "learning_rate": 1.5575620767494356e-05, "loss": 2.1805, "step": 345 }, { "epoch": 0.23704707077548257, "grad_norm": 1.265625, "learning_rate": 1.5801354401805872e-05, "loss": 2.1765, "step": 350 }, { "epoch": 0.2404334575008466, "grad_norm": 1.21875, "learning_rate": 1.602708803611738e-05, "loss": 2.1734, "step": 355 }, { "epoch": 0.24381984422621064, "grad_norm": 1.265625, "learning_rate": 1.6252821670428894e-05, "loss": 2.1681, "step": 360 }, { "epoch": 0.24720623095157468, "grad_norm": 1.2421875, "learning_rate": 1.6478555304740406e-05, "loss": 2.1865, "step": 365 }, { "epoch": 0.2505926176769387, "grad_norm": 1.25, "learning_rate": 1.670428893905192e-05, "loss": 2.1665, "step": 370 }, { "epoch": 0.25397900440230275, "grad_norm": 1.1875, "learning_rate": 1.693002257336343e-05, "loss": 2.1621, "step": 375 }, { "epoch": 0.25736539112766676, "grad_norm": 1.2421875, "learning_rate": 1.7155756207674944e-05, "loss": 2.186, "step": 380 }, { "epoch": 0.26075177785303083, "grad_norm": 1.2890625, "learning_rate": 1.7381489841986457e-05, "loss": 2.1651, "step": 385 }, { "epoch": 0.26413816457839484, "grad_norm": 1.2421875, "learning_rate": 1.760722347629797e-05, "loss": 2.1786, "step": 390 }, { "epoch": 0.2675245513037589, "grad_norm": 1.25, "learning_rate": 1.783295711060948e-05, "loss": 2.1472, "step": 395 }, { "epoch": 0.2709109380291229, "grad_norm": 1.265625, "learning_rate": 1.8058690744920994e-05, "loss": 2.1712, "step": 400 }, { "epoch": 0.274297324754487, "grad_norm": 1.2109375, "learning_rate": 1.8284424379232507e-05, "loss": 2.1611, "step": 405 }, { "epoch": 0.277683711479851, "grad_norm": 1.265625, "learning_rate": 1.851015801354402e-05, "loss": 2.1727, "step": 410 }, { "epoch": 0.28107009820521506, "grad_norm": 1.21875, "learning_rate": 1.873589164785553e-05, "loss": 2.1499, "step": 415 }, { "epoch": 0.28445648493057907, "grad_norm": 1.2265625, "learning_rate": 1.8961625282167044e-05, "loss": 2.1616, "step": 420 }, { "epoch": 0.28784287165594313, "grad_norm": 1.203125, "learning_rate": 1.9187358916478557e-05, "loss": 2.1646, "step": 425 }, { "epoch": 0.29122925838130714, "grad_norm": 1.234375, "learning_rate": 1.941309255079007e-05, "loss": 2.1476, "step": 430 }, { "epoch": 0.2946156451066712, "grad_norm": 1.28125, "learning_rate": 1.963882618510158e-05, "loss": 2.1641, "step": 435 }, { "epoch": 0.2980020318320352, "grad_norm": 1.234375, "learning_rate": 1.9864559819413094e-05, "loss": 2.1577, "step": 440 }, { "epoch": 0.30138841855739923, "grad_norm": 1.2265625, "learning_rate": 1.999998756994645e-05, "loss": 2.1505, "step": 445 }, { "epoch": 0.3047748052827633, "grad_norm": 1.2109375, "learning_rate": 1.99998477321989e-05, "loss": 2.1605, "step": 450 }, { "epoch": 0.3081611920081273, "grad_norm": 1.265625, "learning_rate": 1.9999552521316842e-05, "loss": 2.1373, "step": 455 }, { "epoch": 0.31154757873349137, "grad_norm": 1.2265625, "learning_rate": 1.9999101941887125e-05, "loss": 2.1473, "step": 460 }, { "epoch": 0.3149339654588554, "grad_norm": 1.234375, "learning_rate": 1.9998496000910653e-05, "loss": 2.1594, "step": 465 }, { "epoch": 0.31832035218421945, "grad_norm": 1.2890625, "learning_rate": 1.999773470780227e-05, "loss": 2.1588, "step": 470 }, { "epoch": 0.32170673890958346, "grad_norm": 1.2734375, "learning_rate": 1.999681807439059e-05, "loss": 2.1672, "step": 475 }, { "epoch": 0.3250931256349475, "grad_norm": 1.203125, "learning_rate": 1.9995746114917866e-05, "loss": 2.1631, "step": 480 }, { "epoch": 0.32847951236031153, "grad_norm": 1.234375, "learning_rate": 1.9994518846039715e-05, "loss": 2.156, "step": 485 }, { "epoch": 0.3318658990856756, "grad_norm": 1.2265625, "learning_rate": 1.9993136286824894e-05, "loss": 2.1593, "step": 490 }, { "epoch": 0.3352522858110396, "grad_norm": 1.1640625, "learning_rate": 1.999159845875498e-05, "loss": 2.174, "step": 495 }, { "epoch": 0.3386386725364037, "grad_norm": 1.21875, "learning_rate": 1.9989905385724052e-05, "loss": 2.1444, "step": 500 }, { "epoch": 0.3420250592617677, "grad_norm": 1.2421875, "learning_rate": 1.9988057094038325e-05, "loss": 2.1395, "step": 505 }, { "epoch": 0.34541144598713175, "grad_norm": 1.2109375, "learning_rate": 1.9986053612415717e-05, "loss": 2.1486, "step": 510 }, { "epoch": 0.34879783271249576, "grad_norm": 1.2109375, "learning_rate": 1.9983894971985425e-05, "loss": 2.1498, "step": 515 }, { "epoch": 0.3521842194378598, "grad_norm": 1.234375, "learning_rate": 1.9981581206287437e-05, "loss": 2.1265, "step": 520 }, { "epoch": 0.35557060616322383, "grad_norm": 1.234375, "learning_rate": 1.9979112351271996e-05, "loss": 2.1298, "step": 525 }, { "epoch": 0.3589569928885879, "grad_norm": 1.2578125, "learning_rate": 1.9976488445299066e-05, "loss": 2.1563, "step": 530 }, { "epoch": 0.3623433796139519, "grad_norm": 1.1875, "learning_rate": 1.9973709529137717e-05, "loss": 2.1582, "step": 535 }, { "epoch": 0.365729766339316, "grad_norm": 1.1953125, "learning_rate": 1.9970775645965493e-05, "loss": 2.1549, "step": 540 }, { "epoch": 0.36911615306468, "grad_norm": 1.2265625, "learning_rate": 1.9967686841367757e-05, "loss": 2.1804, "step": 545 }, { "epoch": 0.372502539790044, "grad_norm": 1.171875, "learning_rate": 1.996444316333696e-05, "loss": 2.153, "step": 550 }, { "epoch": 0.37588892651540806, "grad_norm": 1.1640625, "learning_rate": 1.9961044662271922e-05, "loss": 2.162, "step": 555 }, { "epoch": 0.37927531324077207, "grad_norm": 1.21875, "learning_rate": 1.9957491390977007e-05, "loss": 2.164, "step": 560 }, { "epoch": 0.38266169996613614, "grad_norm": 1.2265625, "learning_rate": 1.9953783404661356e-05, "loss": 2.1532, "step": 565 }, { "epoch": 0.38604808669150015, "grad_norm": 1.2109375, "learning_rate": 1.994992076093799e-05, "loss": 2.1603, "step": 570 }, { "epoch": 0.3894344734168642, "grad_norm": 1.21875, "learning_rate": 1.9945903519822927e-05, "loss": 2.158, "step": 575 }, { "epoch": 0.3928208601422282, "grad_norm": 1.2421875, "learning_rate": 1.9941731743734258e-05, "loss": 2.1451, "step": 580 }, { "epoch": 0.3962072468675923, "grad_norm": 1.2109375, "learning_rate": 1.993740549749115e-05, "loss": 2.1554, "step": 585 }, { "epoch": 0.3995936335929563, "grad_norm": 1.1875, "learning_rate": 1.9932924848312888e-05, "loss": 2.1331, "step": 590 }, { "epoch": 0.40298002031832036, "grad_norm": 1.1953125, "learning_rate": 1.992828986581777e-05, "loss": 2.1476, "step": 595 }, { "epoch": 0.4063664070436844, "grad_norm": 1.21875, "learning_rate": 1.992350062202207e-05, "loss": 2.1458, "step": 600 }, { "epoch": 0.40975279376904844, "grad_norm": 1.203125, "learning_rate": 1.991855719133891e-05, "loss": 2.1569, "step": 605 }, { "epoch": 0.41313918049441245, "grad_norm": 1.203125, "learning_rate": 1.9913459650577082e-05, "loss": 2.1378, "step": 610 }, { "epoch": 0.4165255672197765, "grad_norm": 1.1875, "learning_rate": 1.990820807893989e-05, "loss": 2.1378, "step": 615 }, { "epoch": 0.4199119539451405, "grad_norm": 1.2109375, "learning_rate": 1.9902802558023883e-05, "loss": 2.1436, "step": 620 }, { "epoch": 0.4232983406705046, "grad_norm": 1.21875, "learning_rate": 1.989724317181762e-05, "loss": 2.1556, "step": 625 }, { "epoch": 0.4266847273958686, "grad_norm": 1.1796875, "learning_rate": 1.9891530006700335e-05, "loss": 2.152, "step": 630 }, { "epoch": 0.43007111412123267, "grad_norm": 1.2421875, "learning_rate": 1.988566315144062e-05, "loss": 2.1539, "step": 635 }, { "epoch": 0.4334575008465967, "grad_norm": 1.234375, "learning_rate": 1.9879642697195035e-05, "loss": 2.1624, "step": 640 }, { "epoch": 0.43684388757196074, "grad_norm": 1.1796875, "learning_rate": 1.987346873750669e-05, "loss": 2.1513, "step": 645 }, { "epoch": 0.44023027429732475, "grad_norm": 1.234375, "learning_rate": 1.9867141368303783e-05, "loss": 2.1344, "step": 650 }, { "epoch": 0.4436166610226888, "grad_norm": 1.2109375, "learning_rate": 1.9860660687898142e-05, "loss": 2.16, "step": 655 }, { "epoch": 0.44700304774805283, "grad_norm": 1.203125, "learning_rate": 1.9854026796983652e-05, "loss": 2.1398, "step": 660 }, { "epoch": 0.45038943447341684, "grad_norm": 1.1953125, "learning_rate": 1.9847239798634734e-05, "loss": 2.1383, "step": 665 }, { "epoch": 0.4537758211987809, "grad_norm": 1.234375, "learning_rate": 1.9840299798304708e-05, "loss": 2.1298, "step": 670 }, { "epoch": 0.4571622079241449, "grad_norm": 1.2265625, "learning_rate": 1.9833206903824177e-05, "loss": 2.1569, "step": 675 }, { "epoch": 0.460548594649509, "grad_norm": 1.2109375, "learning_rate": 1.982596122539935e-05, "loss": 2.134, "step": 680 }, { "epoch": 0.463934981374873, "grad_norm": 1.1796875, "learning_rate": 1.981856287561031e-05, "loss": 2.1341, "step": 685 }, { "epoch": 0.46732136810023706, "grad_norm": 1.21875, "learning_rate": 1.9811011969409297e-05, "loss": 2.1496, "step": 690 }, { "epoch": 0.47070775482560107, "grad_norm": 1.2109375, "learning_rate": 1.9803308624118882e-05, "loss": 2.1695, "step": 695 }, { "epoch": 0.47409414155096513, "grad_norm": 1.2109375, "learning_rate": 1.9795452959430187e-05, "loss": 2.1361, "step": 700 }, { "epoch": 0.47748052827632914, "grad_norm": 1.2109375, "learning_rate": 1.978744509740099e-05, "loss": 2.165, "step": 705 }, { "epoch": 0.4808669150016932, "grad_norm": 1.2109375, "learning_rate": 1.9779285162453853e-05, "loss": 2.1356, "step": 710 }, { "epoch": 0.4842533017270572, "grad_norm": 1.2109375, "learning_rate": 1.9770973281374168e-05, "loss": 2.1442, "step": 715 }, { "epoch": 0.4876396884524213, "grad_norm": 1.1875, "learning_rate": 1.9762509583308206e-05, "loss": 2.1332, "step": 720 }, { "epoch": 0.4910260751777853, "grad_norm": 1.1953125, "learning_rate": 1.97538941997611e-05, "loss": 2.1472, "step": 725 }, { "epoch": 0.49441246190314936, "grad_norm": 1.1953125, "learning_rate": 1.9745127264594802e-05, "loss": 2.1405, "step": 730 }, { "epoch": 0.49779884862851337, "grad_norm": 1.25, "learning_rate": 1.9736208914026004e-05, "loss": 2.1393, "step": 735 }, { "epoch": 0.5011852353538774, "grad_norm": 1.203125, "learning_rate": 1.9727139286624026e-05, "loss": 2.1532, "step": 740 }, { "epoch": 0.5045716220792414, "grad_norm": 1.2109375, "learning_rate": 1.9717918523308656e-05, "loss": 2.1491, "step": 745 }, { "epoch": 0.5079580088046055, "grad_norm": 1.1875, "learning_rate": 1.9708546767347972e-05, "loss": 2.1705, "step": 750 }, { "epoch": 0.5113443955299696, "grad_norm": 1.15625, "learning_rate": 1.9699024164356092e-05, "loss": 2.1159, "step": 755 }, { "epoch": 0.5147307822553335, "grad_norm": 1.1796875, "learning_rate": 1.9689350862290943e-05, "loss": 2.1381, "step": 760 }, { "epoch": 0.5181171689806976, "grad_norm": 1.1796875, "learning_rate": 1.9679527011451937e-05, "loss": 2.1408, "step": 765 }, { "epoch": 0.5215035557060617, "grad_norm": 1.203125, "learning_rate": 1.9669552764477644e-05, "loss": 2.1505, "step": 770 }, { "epoch": 0.5248899424314256, "grad_norm": 1.203125, "learning_rate": 1.9659428276343435e-05, "loss": 2.1325, "step": 775 }, { "epoch": 0.5282763291567897, "grad_norm": 1.203125, "learning_rate": 1.9649153704359042e-05, "loss": 2.1494, "step": 780 }, { "epoch": 0.5316627158821537, "grad_norm": 1.171875, "learning_rate": 1.963872920816615e-05, "loss": 2.1513, "step": 785 }, { "epoch": 0.5350491026075178, "grad_norm": 1.1640625, "learning_rate": 1.9628154949735882e-05, "loss": 2.1416, "step": 790 }, { "epoch": 0.5384354893328818, "grad_norm": 1.1953125, "learning_rate": 1.9617431093366312e-05, "loss": 2.1302, "step": 795 }, { "epoch": 0.5418218760582458, "grad_norm": 1.21875, "learning_rate": 1.9606557805679893e-05, "loss": 2.1431, "step": 800 }, { "epoch": 0.5452082627836099, "grad_norm": 1.2265625, "learning_rate": 1.9595535255620877e-05, "loss": 2.1397, "step": 805 }, { "epoch": 0.548594649508974, "grad_norm": 1.171875, "learning_rate": 1.958436361445269e-05, "loss": 2.1399, "step": 810 }, { "epoch": 0.5519810362343379, "grad_norm": 1.21875, "learning_rate": 1.957304305575526e-05, "loss": 2.1521, "step": 815 }, { "epoch": 0.555367422959702, "grad_norm": 1.2109375, "learning_rate": 1.9561573755422338e-05, "loss": 2.1551, "step": 820 }, { "epoch": 0.558753809685066, "grad_norm": 1.21875, "learning_rate": 1.954995589165875e-05, "loss": 2.1325, "step": 825 }, { "epoch": 0.5621401964104301, "grad_norm": 1.2578125, "learning_rate": 1.953818964497764e-05, "loss": 2.1375, "step": 830 }, { "epoch": 0.5655265831357941, "grad_norm": 1.1875, "learning_rate": 1.9526275198197647e-05, "loss": 2.1424, "step": 835 }, { "epoch": 0.5689129698611581, "grad_norm": 1.1796875, "learning_rate": 1.9514212736440094e-05, "loss": 2.1309, "step": 840 }, { "epoch": 0.5722993565865222, "grad_norm": 1.1953125, "learning_rate": 1.9502002447126073e-05, "loss": 2.1355, "step": 845 }, { "epoch": 0.5756857433118863, "grad_norm": 1.1796875, "learning_rate": 1.948964451997357e-05, "loss": 2.1404, "step": 850 }, { "epoch": 0.5790721300372502, "grad_norm": 1.1875, "learning_rate": 1.9477139146994505e-05, "loss": 2.1129, "step": 855 }, { "epoch": 0.5824585167626143, "grad_norm": 1.1796875, "learning_rate": 1.946448652249172e-05, "loss": 2.1403, "step": 860 }, { "epoch": 0.5858449034879784, "grad_norm": 1.2109375, "learning_rate": 1.9451686843056013e-05, "loss": 2.1333, "step": 865 }, { "epoch": 0.5892312902133424, "grad_norm": 1.25, "learning_rate": 1.943874030756304e-05, "loss": 2.144, "step": 870 }, { "epoch": 0.5926176769387064, "grad_norm": 1.1953125, "learning_rate": 1.9425647117170245e-05, "loss": 2.1423, "step": 875 }, { "epoch": 0.5960040636640704, "grad_norm": 1.2109375, "learning_rate": 1.9412407475313727e-05, "loss": 2.1338, "step": 880 }, { "epoch": 0.5993904503894345, "grad_norm": 1.1484375, "learning_rate": 1.939902158770509e-05, "loss": 2.1294, "step": 885 }, { "epoch": 0.6027768371147985, "grad_norm": 1.1875, "learning_rate": 1.938548966232822e-05, "loss": 2.1167, "step": 890 }, { "epoch": 0.6061632238401625, "grad_norm": 1.203125, "learning_rate": 1.9371811909436097e-05, "loss": 2.1691, "step": 895 }, { "epoch": 0.6095496105655266, "grad_norm": 1.1875, "learning_rate": 1.9357988541547486e-05, "loss": 2.1158, "step": 900 }, { "epoch": 0.6129359972908907, "grad_norm": 1.1640625, "learning_rate": 1.9344019773443655e-05, "loss": 2.1318, "step": 905 }, { "epoch": 0.6163223840162546, "grad_norm": 1.15625, "learning_rate": 1.9329905822165042e-05, "loss": 2.1414, "step": 910 }, { "epoch": 0.6197087707416187, "grad_norm": 1.1484375, "learning_rate": 1.931564690700787e-05, "loss": 2.1453, "step": 915 }, { "epoch": 0.6230951574669827, "grad_norm": 1.1875, "learning_rate": 1.9301243249520746e-05, "loss": 2.133, "step": 920 }, { "epoch": 0.6264815441923468, "grad_norm": 1.1640625, "learning_rate": 1.928669507350122e-05, "loss": 2.123, "step": 925 }, { "epoch": 0.6298679309177108, "grad_norm": 1.1796875, "learning_rate": 1.92720026049923e-05, "loss": 2.1289, "step": 930 }, { "epoch": 0.6332543176430748, "grad_norm": 1.171875, "learning_rate": 1.9257166072278957e-05, "loss": 2.1385, "step": 935 }, { "epoch": 0.6366407043684389, "grad_norm": 1.21875, "learning_rate": 1.924218570588456e-05, "loss": 2.1414, "step": 940 }, { "epoch": 0.640027091093803, "grad_norm": 1.1484375, "learning_rate": 1.9227061738567293e-05, "loss": 2.1418, "step": 945 }, { "epoch": 0.6434134778191669, "grad_norm": 1.1640625, "learning_rate": 1.9211794405316566e-05, "loss": 2.1372, "step": 950 }, { "epoch": 0.646799864544531, "grad_norm": 1.21875, "learning_rate": 1.919638394334933e-05, "loss": 2.122, "step": 955 }, { "epoch": 0.650186251269895, "grad_norm": 1.2265625, "learning_rate": 1.91808305921064e-05, "loss": 2.1357, "step": 960 }, { "epoch": 0.6535726379952591, "grad_norm": 1.203125, "learning_rate": 1.916513459324876e-05, "loss": 2.1205, "step": 965 }, { "epoch": 0.6569590247206231, "grad_norm": 1.1953125, "learning_rate": 1.9149296190653776e-05, "loss": 2.1248, "step": 970 }, { "epoch": 0.6603454114459871, "grad_norm": 1.203125, "learning_rate": 1.9133315630411423e-05, "loss": 2.1463, "step": 975 }, { "epoch": 0.6637317981713512, "grad_norm": 1.203125, "learning_rate": 1.9117193160820445e-05, "loss": 2.1155, "step": 980 }, { "epoch": 0.6671181848967151, "grad_norm": 1.1796875, "learning_rate": 1.9100929032384527e-05, "loss": 2.107, "step": 985 }, { "epoch": 0.6705045716220792, "grad_norm": 1.171875, "learning_rate": 1.908452349780838e-05, "loss": 2.1478, "step": 990 }, { "epoch": 0.6738909583474433, "grad_norm": 1.1796875, "learning_rate": 1.906797681199382e-05, "loss": 2.1388, "step": 995 }, { "epoch": 0.6772773450728073, "grad_norm": 1.1875, "learning_rate": 1.9051289232035794e-05, "loss": 2.1359, "step": 1000 }, { "epoch": 0.6806637317981713, "grad_norm": 1.203125, "learning_rate": 1.9034461017218424e-05, "loss": 2.149, "step": 1005 }, { "epoch": 0.6840501185235354, "grad_norm": 1.1875, "learning_rate": 1.9017492429010933e-05, "loss": 2.1402, "step": 1010 }, { "epoch": 0.6874365052488994, "grad_norm": 1.1875, "learning_rate": 1.900038373106361e-05, "loss": 2.1211, "step": 1015 }, { "epoch": 0.6908228919742635, "grad_norm": 1.1875, "learning_rate": 1.8983135189203708e-05, "loss": 2.1127, "step": 1020 }, { "epoch": 0.6942092786996275, "grad_norm": 1.203125, "learning_rate": 1.8965747071431307e-05, "loss": 2.1432, "step": 1025 }, { "epoch": 0.6975956654249915, "grad_norm": 1.171875, "learning_rate": 1.8948219647915157e-05, "loss": 2.1353, "step": 1030 }, { "epoch": 0.7009820521503556, "grad_norm": 1.203125, "learning_rate": 1.893055319098848e-05, "loss": 2.1372, "step": 1035 }, { "epoch": 0.7043684388757196, "grad_norm": 1.1796875, "learning_rate": 1.891274797514473e-05, "loss": 2.1385, "step": 1040 }, { "epoch": 0.7077548256010836, "grad_norm": 1.15625, "learning_rate": 1.8894804277033345e-05, "loss": 2.1252, "step": 1045 }, { "epoch": 0.7111412123264477, "grad_norm": 1.1953125, "learning_rate": 1.8876722375455426e-05, "loss": 2.1302, "step": 1050 }, { "epoch": 0.7145275990518117, "grad_norm": 1.1875, "learning_rate": 1.885850255135943e-05, "loss": 2.1437, "step": 1055 }, { "epoch": 0.7179139857771758, "grad_norm": 1.15625, "learning_rate": 1.8840145087836777e-05, "loss": 2.1213, "step": 1060 }, { "epoch": 0.7213003725025398, "grad_norm": 1.1640625, "learning_rate": 1.8821650270117485e-05, "loss": 2.1354, "step": 1065 }, { "epoch": 0.7246867592279038, "grad_norm": 1.203125, "learning_rate": 1.8803018385565707e-05, "loss": 2.1425, "step": 1070 }, { "epoch": 0.7280731459532679, "grad_norm": 1.1875, "learning_rate": 1.8784249723675273e-05, "loss": 2.1342, "step": 1075 }, { "epoch": 0.731459532678632, "grad_norm": 1.2265625, "learning_rate": 1.8765344576065222e-05, "loss": 2.1274, "step": 1080 }, { "epoch": 0.7348459194039959, "grad_norm": 1.1875, "learning_rate": 1.8746303236475212e-05, "loss": 2.1635, "step": 1085 }, { "epoch": 0.73823230612936, "grad_norm": 1.2421875, "learning_rate": 1.8727126000761025e-05, "loss": 2.1549, "step": 1090 }, { "epoch": 0.741618692854724, "grad_norm": 1.1953125, "learning_rate": 1.8707813166889915e-05, "loss": 2.1232, "step": 1095 }, { "epoch": 0.745005079580088, "grad_norm": 1.2109375, "learning_rate": 1.8688365034936002e-05, "loss": 2.1266, "step": 1100 }, { "epoch": 0.7483914663054521, "grad_norm": 1.171875, "learning_rate": 1.8668781907075618e-05, "loss": 2.1122, "step": 1105 }, { "epoch": 0.7517778530308161, "grad_norm": 1.1484375, "learning_rate": 1.864906408758258e-05, "loss": 2.1109, "step": 1110 }, { "epoch": 0.7551642397561802, "grad_norm": 1.2265625, "learning_rate": 1.8629211882823502e-05, "loss": 2.1381, "step": 1115 }, { "epoch": 0.7585506264815441, "grad_norm": 1.2109375, "learning_rate": 1.860922560125301e-05, "loss": 2.1412, "step": 1120 }, { "epoch": 0.7619370132069082, "grad_norm": 1.1953125, "learning_rate": 1.858910555340895e-05, "loss": 2.1333, "step": 1125 }, { "epoch": 0.7653233999322723, "grad_norm": 1.1953125, "learning_rate": 1.8568852051907575e-05, "loss": 2.1583, "step": 1130 }, { "epoch": 0.7687097866576363, "grad_norm": 1.1484375, "learning_rate": 1.854846541143868e-05, "loss": 2.122, "step": 1135 }, { "epoch": 0.7720961733830003, "grad_norm": 1.171875, "learning_rate": 1.8527945948760702e-05, "loss": 2.1293, "step": 1140 }, { "epoch": 0.7754825601083644, "grad_norm": 1.5546875, "learning_rate": 1.850729398269583e-05, "loss": 2.1142, "step": 1145 }, { "epoch": 0.7788689468337284, "grad_norm": 1.21875, "learning_rate": 1.8486509834125013e-05, "loss": 2.148, "step": 1150 }, { "epoch": 0.7822553335590925, "grad_norm": 1.21875, "learning_rate": 1.8465593825983005e-05, "loss": 2.145, "step": 1155 }, { "epoch": 0.7856417202844564, "grad_norm": 1.1796875, "learning_rate": 1.8444546283253325e-05, "loss": 2.1206, "step": 1160 }, { "epoch": 0.7890281070098205, "grad_norm": 1.203125, "learning_rate": 1.8423367532963224e-05, "loss": 2.1323, "step": 1165 }, { "epoch": 0.7924144937351846, "grad_norm": 1.2421875, "learning_rate": 1.840205790417859e-05, "loss": 2.1299, "step": 1170 }, { "epoch": 0.7958008804605486, "grad_norm": 1.2421875, "learning_rate": 1.838061772799885e-05, "loss": 2.1289, "step": 1175 }, { "epoch": 0.7991872671859126, "grad_norm": 1.203125, "learning_rate": 1.8359047337551815e-05, "loss": 2.1564, "step": 1180 }, { "epoch": 0.8025736539112767, "grad_norm": 1.1796875, "learning_rate": 1.8337347067988506e-05, "loss": 2.1398, "step": 1185 }, { "epoch": 0.8059600406366407, "grad_norm": 1.1875, "learning_rate": 1.8315517256477942e-05, "loss": 2.1273, "step": 1190 }, { "epoch": 0.8093464273620048, "grad_norm": 1.2265625, "learning_rate": 1.8293558242201914e-05, "loss": 2.1186, "step": 1195 }, { "epoch": 0.8127328140873687, "grad_norm": 1.1640625, "learning_rate": 1.827147036634971e-05, "loss": 2.1423, "step": 1200 }, { "epoch": 0.8161192008127328, "grad_norm": 1.15625, "learning_rate": 1.8249253972112805e-05, "loss": 2.1372, "step": 1205 }, { "epoch": 0.8195055875380969, "grad_norm": 1.1796875, "learning_rate": 1.8226909404679526e-05, "loss": 2.1415, "step": 1210 }, { "epoch": 0.8228919742634608, "grad_norm": 1.203125, "learning_rate": 1.820443701122972e-05, "loss": 2.1409, "step": 1215 }, { "epoch": 0.8262783609888249, "grad_norm": 1.1484375, "learning_rate": 1.8181837140929322e-05, "loss": 2.1275, "step": 1220 }, { "epoch": 0.829664747714189, "grad_norm": 1.1953125, "learning_rate": 1.8159110144924943e-05, "loss": 2.1377, "step": 1225 }, { "epoch": 0.833051134439553, "grad_norm": 1.1640625, "learning_rate": 1.8136256376338427e-05, "loss": 2.1274, "step": 1230 }, { "epoch": 0.836437521164917, "grad_norm": 1.15625, "learning_rate": 1.811327619026134e-05, "loss": 2.1449, "step": 1235 }, { "epoch": 0.839823907890281, "grad_norm": 1.1875, "learning_rate": 1.8090169943749477e-05, "loss": 2.1263, "step": 1240 }, { "epoch": 0.8432102946156451, "grad_norm": 1.1640625, "learning_rate": 1.8066937995817296e-05, "loss": 2.1207, "step": 1245 }, { "epoch": 0.8465966813410092, "grad_norm": 1.1796875, "learning_rate": 1.8043580707432356e-05, "loss": 2.1066, "step": 1250 }, { "epoch": 0.8499830680663731, "grad_norm": 1.1953125, "learning_rate": 1.802009844150969e-05, "loss": 2.1421, "step": 1255 }, { "epoch": 0.8533694547917372, "grad_norm": 1.2109375, "learning_rate": 1.7996491562906186e-05, "loss": 2.1265, "step": 1260 }, { "epoch": 0.8567558415171013, "grad_norm": 1.140625, "learning_rate": 1.797276043841489e-05, "loss": 2.1343, "step": 1265 }, { "epoch": 0.8601422282424653, "grad_norm": 1.2109375, "learning_rate": 1.7948905436759346e-05, "loss": 2.134, "step": 1270 }, { "epoch": 0.8635286149678293, "grad_norm": 1.1640625, "learning_rate": 1.7924926928587834e-05, "loss": 2.1073, "step": 1275 }, { "epoch": 0.8669150016931934, "grad_norm": 1.1875, "learning_rate": 1.7900825286467622e-05, "loss": 2.1149, "step": 1280 }, { "epoch": 0.8703013884185574, "grad_norm": 1.1796875, "learning_rate": 1.787660088487918e-05, "loss": 2.1335, "step": 1285 }, { "epoch": 0.8736877751439215, "grad_norm": 1.203125, "learning_rate": 1.7852254100210363e-05, "loss": 2.1261, "step": 1290 }, { "epoch": 0.8770741618692854, "grad_norm": 1.203125, "learning_rate": 1.7827785310750552e-05, "loss": 2.1274, "step": 1295 }, { "epoch": 0.8804605485946495, "grad_norm": 1.15625, "learning_rate": 1.780319489668479e-05, "loss": 2.1213, "step": 1300 }, { "epoch": 0.8838469353200136, "grad_norm": 1.1875, "learning_rate": 1.7778483240087873e-05, "loss": 2.1304, "step": 1305 }, { "epoch": 0.8872333220453776, "grad_norm": 1.171875, "learning_rate": 1.7753650724918388e-05, "loss": 2.1386, "step": 1310 }, { "epoch": 0.8906197087707416, "grad_norm": 1.1875, "learning_rate": 1.7728697737012794e-05, "loss": 2.128, "step": 1315 }, { "epoch": 0.8940060954961057, "grad_norm": 1.1484375, "learning_rate": 1.770362466407938e-05, "loss": 2.1301, "step": 1320 }, { "epoch": 0.8973924822214697, "grad_norm": 1.140625, "learning_rate": 1.7678431895692278e-05, "loss": 2.1275, "step": 1325 }, { "epoch": 0.9007788689468337, "grad_norm": 1.203125, "learning_rate": 1.7653119823285383e-05, "loss": 2.1423, "step": 1330 }, { "epoch": 0.9041652556721977, "grad_norm": 1.15625, "learning_rate": 1.7627688840146286e-05, "loss": 2.1195, "step": 1335 }, { "epoch": 0.9075516423975618, "grad_norm": 1.2109375, "learning_rate": 1.760213934141017e-05, "loss": 2.1369, "step": 1340 }, { "epoch": 0.9109380291229259, "grad_norm": 1.1875, "learning_rate": 1.7576471724053634e-05, "loss": 2.1259, "step": 1345 }, { "epoch": 0.9143244158482898, "grad_norm": 1.2109375, "learning_rate": 1.7550686386888583e-05, "loss": 2.1294, "step": 1350 }, { "epoch": 0.9177108025736539, "grad_norm": 1.1640625, "learning_rate": 1.752478373055597e-05, "loss": 2.1275, "step": 1355 }, { "epoch": 0.921097189299018, "grad_norm": 1.1953125, "learning_rate": 1.7498764157519625e-05, "loss": 2.1376, "step": 1360 }, { "epoch": 0.924483576024382, "grad_norm": 1.1953125, "learning_rate": 1.7472628072059955e-05, "loss": 2.1415, "step": 1365 }, { "epoch": 0.927869962749746, "grad_norm": 1.1875, "learning_rate": 1.744637588026771e-05, "loss": 2.1356, "step": 1370 }, { "epoch": 0.93125634947511, "grad_norm": 1.1875, "learning_rate": 1.742000799003762e-05, "loss": 2.1287, "step": 1375 }, { "epoch": 0.9346427362004741, "grad_norm": 1.1796875, "learning_rate": 1.7393524811062105e-05, "loss": 2.1163, "step": 1380 }, { "epoch": 0.9380291229258382, "grad_norm": 1.21875, "learning_rate": 1.736692675482488e-05, "loss": 2.1505, "step": 1385 }, { "epoch": 0.9414155096512021, "grad_norm": 1.1953125, "learning_rate": 1.734021423459458e-05, "loss": 2.1286, "step": 1390 }, { "epoch": 0.9448018963765662, "grad_norm": 1.1484375, "learning_rate": 1.731338766541832e-05, "loss": 2.1295, "step": 1395 }, { "epoch": 0.9481882831019303, "grad_norm": 1.2109375, "learning_rate": 1.7286447464115267e-05, "loss": 2.1351, "step": 1400 }, { "epoch": 0.9515746698272943, "grad_norm": 1.1953125, "learning_rate": 1.7259394049270138e-05, "loss": 2.1251, "step": 1405 }, { "epoch": 0.9549610565526583, "grad_norm": 1.1640625, "learning_rate": 1.7232227841226724e-05, "loss": 2.1313, "step": 1410 }, { "epoch": 0.9583474432780223, "grad_norm": 1.171875, "learning_rate": 1.7204949262081344e-05, "loss": 2.1242, "step": 1415 }, { "epoch": 0.9617338300033864, "grad_norm": 1.1796875, "learning_rate": 1.7177558735676285e-05, "loss": 2.1422, "step": 1420 }, { "epoch": 0.9651202167287504, "grad_norm": 1.1796875, "learning_rate": 1.7150056687593218e-05, "loss": 2.1116, "step": 1425 }, { "epoch": 0.9685066034541144, "grad_norm": 1.1953125, "learning_rate": 1.7122443545146598e-05, "loss": 2.1154, "step": 1430 }, { "epoch": 0.9718929901794785, "grad_norm": 1.234375, "learning_rate": 1.7094719737377007e-05, "loss": 2.1348, "step": 1435 }, { "epoch": 0.9752793769048426, "grad_norm": 1.1953125, "learning_rate": 1.7066885695044482e-05, "loss": 2.1122, "step": 1440 }, { "epoch": 0.9786657636302065, "grad_norm": 1.1640625, "learning_rate": 1.703894185062187e-05, "loss": 2.1258, "step": 1445 }, { "epoch": 0.9820521503555706, "grad_norm": 1.1953125, "learning_rate": 1.7010888638288037e-05, "loss": 2.1304, "step": 1450 }, { "epoch": 0.9854385370809347, "grad_norm": 1.1640625, "learning_rate": 1.6982726493921175e-05, "loss": 2.118, "step": 1455 }, { "epoch": 0.9888249238062987, "grad_norm": 1.1953125, "learning_rate": 1.6954455855092014e-05, "loss": 2.1361, "step": 1460 }, { "epoch": 0.9922113105316627, "grad_norm": 1.171875, "learning_rate": 1.692607716105702e-05, "loss": 2.1223, "step": 1465 }, { "epoch": 0.9955976972570267, "grad_norm": 1.203125, "learning_rate": 1.689759085275157e-05, "loss": 2.1329, "step": 1470 }, { "epoch": 0.9989840839823908, "grad_norm": 1.15625, "learning_rate": 1.6868997372783106e-05, "loss": 2.1208, "step": 1475 }, { "epoch": 0.9996613613274636, "eval_loss": 2.124814987182617, "eval_runtime": 85.1985, "eval_samples_per_second": 15.317, "eval_steps_per_second": 1.925, "step": 1476 }, { "epoch": 1.0023704707077548, "grad_norm": 1.15625, "learning_rate": 1.6840297165424254e-05, "loss": 2.0853, "step": 1480 }, { "epoch": 1.0057568574331188, "grad_norm": 1.1640625, "learning_rate": 1.6811490676605916e-05, "loss": 2.0976, "step": 1485 }, { "epoch": 1.0091432441584829, "grad_norm": 1.203125, "learning_rate": 1.6782578353910363e-05, "loss": 2.1058, "step": 1490 }, { "epoch": 1.012529630883847, "grad_norm": 1.21875, "learning_rate": 1.675356064656425e-05, "loss": 2.1074, "step": 1495 }, { "epoch": 1.015916017609211, "grad_norm": 1.1953125, "learning_rate": 1.6724438005431656e-05, "loss": 2.0834, "step": 1500 }, { "epoch": 1.019302404334575, "grad_norm": 1.1875, "learning_rate": 1.6695210883007077e-05, "loss": 2.0947, "step": 1505 }, { "epoch": 1.0226887910599392, "grad_norm": 1.171875, "learning_rate": 1.666587973340839e-05, "loss": 2.0953, "step": 1510 }, { "epoch": 1.026075177785303, "grad_norm": 1.15625, "learning_rate": 1.66364450123698e-05, "loss": 2.0828, "step": 1515 }, { "epoch": 1.029461564510667, "grad_norm": 1.1953125, "learning_rate": 1.6606907177234757e-05, "loss": 2.0834, "step": 1520 }, { "epoch": 1.0328479512360311, "grad_norm": 1.1640625, "learning_rate": 1.657726668694885e-05, "loss": 2.1086, "step": 1525 }, { "epoch": 1.0362343379613952, "grad_norm": 1.1640625, "learning_rate": 1.654752400205269e-05, "loss": 2.1209, "step": 1530 }, { "epoch": 1.0396207246867593, "grad_norm": 1.21875, "learning_rate": 1.6517679584674716e-05, "loss": 2.106, "step": 1535 }, { "epoch": 1.0430071114121233, "grad_norm": 1.2109375, "learning_rate": 1.6487733898524068e-05, "loss": 2.0924, "step": 1540 }, { "epoch": 1.0463934981374874, "grad_norm": 1.2109375, "learning_rate": 1.645768740888334e-05, "loss": 2.0973, "step": 1545 }, { "epoch": 1.0497798848628512, "grad_norm": 1.1953125, "learning_rate": 1.6427540582601368e-05, "loss": 2.0975, "step": 1550 }, { "epoch": 1.0531662715882153, "grad_norm": 1.1796875, "learning_rate": 1.6397293888085975e-05, "loss": 2.0917, "step": 1555 }, { "epoch": 1.0565526583135794, "grad_norm": 1.21875, "learning_rate": 1.636694779529669e-05, "loss": 2.1072, "step": 1560 }, { "epoch": 1.0599390450389434, "grad_norm": 1.203125, "learning_rate": 1.6336502775737443e-05, "loss": 2.0714, "step": 1565 }, { "epoch": 1.0633254317643075, "grad_norm": 1.2109375, "learning_rate": 1.6305959302449255e-05, "loss": 2.0936, "step": 1570 }, { "epoch": 1.0667118184896716, "grad_norm": 1.203125, "learning_rate": 1.6275317850002875e-05, "loss": 2.0924, "step": 1575 }, { "epoch": 1.0700982052150356, "grad_norm": 1.1875, "learning_rate": 1.624457889449139e-05, "loss": 2.1, "step": 1580 }, { "epoch": 1.0734845919403997, "grad_norm": 1.21875, "learning_rate": 1.621374291352287e-05, "loss": 2.1331, "step": 1585 }, { "epoch": 1.0768709786657635, "grad_norm": 1.1875, "learning_rate": 1.618281038621291e-05, "loss": 2.0948, "step": 1590 }, { "epoch": 1.0802573653911276, "grad_norm": 1.21875, "learning_rate": 1.6151781793177198e-05, "loss": 2.1174, "step": 1595 }, { "epoch": 1.0836437521164917, "grad_norm": 1.203125, "learning_rate": 1.6120657616524053e-05, "loss": 2.0937, "step": 1600 }, { "epoch": 1.0870301388418557, "grad_norm": 1.1953125, "learning_rate": 1.608943833984693e-05, "loss": 2.0901, "step": 1605 }, { "epoch": 1.0904165255672198, "grad_norm": 1.2265625, "learning_rate": 1.6058124448216897e-05, "loss": 2.0892, "step": 1610 }, { "epoch": 1.0938029122925839, "grad_norm": 1.1953125, "learning_rate": 1.602671642817511e-05, "loss": 2.0988, "step": 1615 }, { "epoch": 1.097189299017948, "grad_norm": 1.2109375, "learning_rate": 1.5995214767725267e-05, "loss": 2.09, "step": 1620 }, { "epoch": 1.100575685743312, "grad_norm": 1.1875, "learning_rate": 1.5963619956325977e-05, "loss": 2.1143, "step": 1625 }, { "epoch": 1.1039620724686758, "grad_norm": 1.2265625, "learning_rate": 1.593193248488321e-05, "loss": 2.0987, "step": 1630 }, { "epoch": 1.10734845919404, "grad_norm": 1.1796875, "learning_rate": 1.5900152845742638e-05, "loss": 2.0955, "step": 1635 }, { "epoch": 1.110734845919404, "grad_norm": 1.2109375, "learning_rate": 1.5868281532681995e-05, "loss": 2.1032, "step": 1640 }, { "epoch": 1.114121232644768, "grad_norm": 1.203125, "learning_rate": 1.5836319040903402e-05, "loss": 2.0949, "step": 1645 }, { "epoch": 1.117507619370132, "grad_norm": 1.1875, "learning_rate": 1.5804265867025674e-05, "loss": 2.0914, "step": 1650 }, { "epoch": 1.1208940060954962, "grad_norm": 1.1953125, "learning_rate": 1.5772122509076604e-05, "loss": 2.1035, "step": 1655 }, { "epoch": 1.1242803928208602, "grad_norm": 1.1953125, "learning_rate": 1.5739889466485227e-05, "loss": 2.1001, "step": 1660 }, { "epoch": 1.127666779546224, "grad_norm": 1.1953125, "learning_rate": 1.5707567240074052e-05, "loss": 2.1013, "step": 1665 }, { "epoch": 1.1310531662715881, "grad_norm": 1.2421875, "learning_rate": 1.5675156332051293e-05, "loss": 2.0902, "step": 1670 }, { "epoch": 1.1344395529969522, "grad_norm": 1.1953125, "learning_rate": 1.5642657246003048e-05, "loss": 2.1021, "step": 1675 }, { "epoch": 1.1378259397223163, "grad_norm": 1.2265625, "learning_rate": 1.5610070486885498e-05, "loss": 2.1041, "step": 1680 }, { "epoch": 1.1412123264476803, "grad_norm": 1.171875, "learning_rate": 1.5577396561017043e-05, "loss": 2.1088, "step": 1685 }, { "epoch": 1.1445987131730444, "grad_norm": 1.1875, "learning_rate": 1.554463597607044e-05, "loss": 2.1084, "step": 1690 }, { "epoch": 1.1479850998984085, "grad_norm": 1.1953125, "learning_rate": 1.5511789241064922e-05, "loss": 2.1016, "step": 1695 }, { "epoch": 1.1513714866237725, "grad_norm": 1.1796875, "learning_rate": 1.5478856866358266e-05, "loss": 2.0917, "step": 1700 }, { "epoch": 1.1547578733491364, "grad_norm": 1.25, "learning_rate": 1.5445839363638905e-05, "loss": 2.1034, "step": 1705 }, { "epoch": 1.1581442600745004, "grad_norm": 1.234375, "learning_rate": 1.5412737245917927e-05, "loss": 2.0877, "step": 1710 }, { "epoch": 1.1615306467998645, "grad_norm": 1.2109375, "learning_rate": 1.5379551027521154e-05, "loss": 2.0906, "step": 1715 }, { "epoch": 1.1649170335252286, "grad_norm": 1.171875, "learning_rate": 1.5346281224081113e-05, "loss": 2.0846, "step": 1720 }, { "epoch": 1.1683034202505926, "grad_norm": 1.2578125, "learning_rate": 1.531292835252903e-05, "loss": 2.102, "step": 1725 }, { "epoch": 1.1716898069759567, "grad_norm": 1.2265625, "learning_rate": 1.527949293108683e-05, "loss": 2.0851, "step": 1730 }, { "epoch": 1.1750761937013208, "grad_norm": 1.234375, "learning_rate": 1.524597547925903e-05, "loss": 2.1029, "step": 1735 }, { "epoch": 1.1784625804266846, "grad_norm": 1.2109375, "learning_rate": 1.5212376517824727e-05, "loss": 2.0947, "step": 1740 }, { "epoch": 1.1818489671520487, "grad_norm": 1.2265625, "learning_rate": 1.5178696568829465e-05, "loss": 2.1251, "step": 1745 }, { "epoch": 1.1852353538774127, "grad_norm": 1.2109375, "learning_rate": 1.5144936155577136e-05, "loss": 2.0804, "step": 1750 }, { "epoch": 1.1886217406027768, "grad_norm": 1.203125, "learning_rate": 1.5111095802621844e-05, "loss": 2.0978, "step": 1755 }, { "epoch": 1.1920081273281409, "grad_norm": 1.234375, "learning_rate": 1.5077176035759772e-05, "loss": 2.0717, "step": 1760 }, { "epoch": 1.195394514053505, "grad_norm": 1.2109375, "learning_rate": 1.5043177382020994e-05, "loss": 2.0926, "step": 1765 }, { "epoch": 1.198780900778869, "grad_norm": 1.21875, "learning_rate": 1.5009100369661294e-05, "loss": 2.1096, "step": 1770 }, { "epoch": 1.202167287504233, "grad_norm": 1.21875, "learning_rate": 1.4974945528153956e-05, "loss": 2.0994, "step": 1775 }, { "epoch": 1.2055536742295971, "grad_norm": 1.21875, "learning_rate": 1.494071338818154e-05, "loss": 2.0904, "step": 1780 }, { "epoch": 1.208940060954961, "grad_norm": 1.21875, "learning_rate": 1.4906404481627633e-05, "loss": 2.093, "step": 1785 }, { "epoch": 1.212326447680325, "grad_norm": 1.203125, "learning_rate": 1.4872019341568585e-05, "loss": 2.0988, "step": 1790 }, { "epoch": 1.2157128344056891, "grad_norm": 1.1875, "learning_rate": 1.483755850226523e-05, "loss": 2.1117, "step": 1795 }, { "epoch": 1.2190992211310532, "grad_norm": 1.234375, "learning_rate": 1.4803022499154589e-05, "loss": 2.0897, "step": 1800 }, { "epoch": 1.2224856078564172, "grad_norm": 1.1953125, "learning_rate": 1.4768411868841534e-05, "loss": 2.0905, "step": 1805 }, { "epoch": 1.2258719945817813, "grad_norm": 1.2578125, "learning_rate": 1.4733727149090467e-05, "loss": 2.1207, "step": 1810 }, { "epoch": 1.2292583813071452, "grad_norm": 1.21875, "learning_rate": 1.4698968878816952e-05, "loss": 2.0997, "step": 1815 }, { "epoch": 1.2326447680325092, "grad_norm": 1.234375, "learning_rate": 1.466413759807936e-05, "loss": 2.1071, "step": 1820 }, { "epoch": 1.2360311547578733, "grad_norm": 1.25, "learning_rate": 1.4629233848070452e-05, "loss": 2.1019, "step": 1825 }, { "epoch": 1.2394175414832374, "grad_norm": 1.2421875, "learning_rate": 1.4594258171108997e-05, "loss": 2.0857, "step": 1830 }, { "epoch": 1.2428039282086014, "grad_norm": 1.1875, "learning_rate": 1.4559211110631317e-05, "loss": 2.1097, "step": 1835 }, { "epoch": 1.2461903149339655, "grad_norm": 1.203125, "learning_rate": 1.4524093211182874e-05, "loss": 2.1162, "step": 1840 }, { "epoch": 1.2495767016593295, "grad_norm": 1.1953125, "learning_rate": 1.4488905018409792e-05, "loss": 2.0866, "step": 1845 }, { "epoch": 1.2529630883846936, "grad_norm": 1.1953125, "learning_rate": 1.4453647079050366e-05, "loss": 2.1043, "step": 1850 }, { "epoch": 1.2563494751100577, "grad_norm": 1.1875, "learning_rate": 1.4418319940926601e-05, "loss": 2.0843, "step": 1855 }, { "epoch": 1.2597358618354215, "grad_norm": 1.21875, "learning_rate": 1.438292415293567e-05, "loss": 2.1041, "step": 1860 }, { "epoch": 1.2631222485607856, "grad_norm": 1.171875, "learning_rate": 1.43474602650414e-05, "loss": 2.0916, "step": 1865 }, { "epoch": 1.2665086352861497, "grad_norm": 1.234375, "learning_rate": 1.4311928828265724e-05, "loss": 2.0774, "step": 1870 }, { "epoch": 1.2698950220115137, "grad_norm": 1.2109375, "learning_rate": 1.4276330394680122e-05, "loss": 2.1185, "step": 1875 }, { "epoch": 1.2732814087368778, "grad_norm": 1.2265625, "learning_rate": 1.4240665517397031e-05, "loss": 2.1146, "step": 1880 }, { "epoch": 1.2766677954622419, "grad_norm": 1.2109375, "learning_rate": 1.4204934750561275e-05, "loss": 2.0987, "step": 1885 }, { "epoch": 1.2800541821876057, "grad_norm": 1.2109375, "learning_rate": 1.4169138649341422e-05, "loss": 2.0938, "step": 1890 }, { "epoch": 1.2834405689129698, "grad_norm": 1.2265625, "learning_rate": 1.4133277769921192e-05, "loss": 2.1163, "step": 1895 }, { "epoch": 1.2868269556383338, "grad_norm": 1.203125, "learning_rate": 1.4097352669490791e-05, "loss": 2.0889, "step": 1900 }, { "epoch": 1.290213342363698, "grad_norm": 1.171875, "learning_rate": 1.4061363906238268e-05, "loss": 2.103, "step": 1905 }, { "epoch": 1.293599729089062, "grad_norm": 1.1875, "learning_rate": 1.4025312039340829e-05, "loss": 2.104, "step": 1910 }, { "epoch": 1.296986115814426, "grad_norm": 1.1640625, "learning_rate": 1.398919762895616e-05, "loss": 2.1108, "step": 1915 }, { "epoch": 1.30037250253979, "grad_norm": 1.1796875, "learning_rate": 1.395302123621372e-05, "loss": 2.0843, "step": 1920 }, { "epoch": 1.3037588892651542, "grad_norm": 1.1953125, "learning_rate": 1.391678342320602e-05, "loss": 2.0992, "step": 1925 }, { "epoch": 1.3071452759905182, "grad_norm": 1.1796875, "learning_rate": 1.3880484752979898e-05, "loss": 2.1083, "step": 1930 }, { "epoch": 1.3105316627158823, "grad_norm": 1.2265625, "learning_rate": 1.3844125789527755e-05, "loss": 2.09, "step": 1935 }, { "epoch": 1.3139180494412461, "grad_norm": 1.21875, "learning_rate": 1.3807707097778806e-05, "loss": 2.0936, "step": 1940 }, { "epoch": 1.3173044361666102, "grad_norm": 1.203125, "learning_rate": 1.3771229243590296e-05, "loss": 2.1146, "step": 1945 }, { "epoch": 1.3206908228919743, "grad_norm": 1.21875, "learning_rate": 1.3734692793738709e-05, "loss": 2.1093, "step": 1950 }, { "epoch": 1.3240772096173383, "grad_norm": 1.2109375, "learning_rate": 1.3698098315910961e-05, "loss": 2.104, "step": 1955 }, { "epoch": 1.3274635963427024, "grad_norm": 1.21875, "learning_rate": 1.3661446378695588e-05, "loss": 2.0981, "step": 1960 }, { "epoch": 1.3308499830680665, "grad_norm": 1.1796875, "learning_rate": 1.3624737551573896e-05, "loss": 2.1017, "step": 1965 }, { "epoch": 1.3342363697934303, "grad_norm": 1.171875, "learning_rate": 1.358797240491112e-05, "loss": 2.0747, "step": 1970 }, { "epoch": 1.3376227565187944, "grad_norm": 1.15625, "learning_rate": 1.3551151509947572e-05, "loss": 2.0771, "step": 1975 }, { "epoch": 1.3410091432441584, "grad_norm": 1.1875, "learning_rate": 1.3514275438789745e-05, "loss": 2.1104, "step": 1980 }, { "epoch": 1.3443955299695225, "grad_norm": 1.2265625, "learning_rate": 1.3477344764401438e-05, "loss": 2.0789, "step": 1985 }, { "epoch": 1.3477819166948866, "grad_norm": 1.2109375, "learning_rate": 1.344036006059486e-05, "loss": 2.0989, "step": 1990 }, { "epoch": 1.3511683034202506, "grad_norm": 1.1796875, "learning_rate": 1.3403321902021691e-05, "loss": 2.0889, "step": 1995 }, { "epoch": 1.3545546901456147, "grad_norm": 1.2109375, "learning_rate": 1.3366230864164173e-05, "loss": 2.091, "step": 2000 }, { "epoch": 1.3579410768709788, "grad_norm": 1.1640625, "learning_rate": 1.3329087523326168e-05, "loss": 2.0969, "step": 2005 }, { "epoch": 1.3613274635963428, "grad_norm": 1.25, "learning_rate": 1.3291892456624185e-05, "loss": 2.0938, "step": 2010 }, { "epoch": 1.3647138503217067, "grad_norm": 1.234375, "learning_rate": 1.3254646241978439e-05, "loss": 2.1058, "step": 2015 }, { "epoch": 1.3681002370470707, "grad_norm": 1.2109375, "learning_rate": 1.3217349458103855e-05, "loss": 2.0882, "step": 2020 }, { "epoch": 1.3714866237724348, "grad_norm": 1.2421875, "learning_rate": 1.3180002684501075e-05, "loss": 2.1114, "step": 2025 }, { "epoch": 1.3748730104977989, "grad_norm": 1.21875, "learning_rate": 1.3142606501447464e-05, "loss": 2.0993, "step": 2030 }, { "epoch": 1.378259397223163, "grad_norm": 1.2109375, "learning_rate": 1.3105161489988087e-05, "loss": 2.0827, "step": 2035 }, { "epoch": 1.381645783948527, "grad_norm": 1.21875, "learning_rate": 1.3067668231926685e-05, "loss": 2.1031, "step": 2040 }, { "epoch": 1.3850321706738908, "grad_norm": 1.1875, "learning_rate": 1.3030127309816632e-05, "loss": 2.1086, "step": 2045 }, { "epoch": 1.388418557399255, "grad_norm": 1.1953125, "learning_rate": 1.2992539306951878e-05, "loss": 2.0985, "step": 2050 }, { "epoch": 1.391804944124619, "grad_norm": 1.203125, "learning_rate": 1.2954904807357905e-05, "loss": 2.0981, "step": 2055 }, { "epoch": 1.395191330849983, "grad_norm": 1.21875, "learning_rate": 1.2917224395782627e-05, "loss": 2.0995, "step": 2060 }, { "epoch": 1.398577717575347, "grad_norm": 1.1875, "learning_rate": 1.2879498657687335e-05, "loss": 2.1045, "step": 2065 }, { "epoch": 1.4019641043007112, "grad_norm": 1.2109375, "learning_rate": 1.284172817923756e-05, "loss": 2.1057, "step": 2070 }, { "epoch": 1.4053504910260752, "grad_norm": 1.2265625, "learning_rate": 1.280391354729401e-05, "loss": 2.1083, "step": 2075 }, { "epoch": 1.4087368777514393, "grad_norm": 1.1875, "learning_rate": 1.2766055349403415e-05, "loss": 2.0772, "step": 2080 }, { "epoch": 1.4121232644768034, "grad_norm": 1.1796875, "learning_rate": 1.272815417378942e-05, "loss": 2.0991, "step": 2085 }, { "epoch": 1.4155096512021672, "grad_norm": 1.171875, "learning_rate": 1.2690210609343439e-05, "loss": 2.0951, "step": 2090 }, { "epoch": 1.4188960379275313, "grad_norm": 1.1953125, "learning_rate": 1.2652225245615503e-05, "loss": 2.0951, "step": 2095 }, { "epoch": 1.4222824246528953, "grad_norm": 1.1953125, "learning_rate": 1.2614198672805105e-05, "loss": 2.0926, "step": 2100 }, { "epoch": 1.4256688113782594, "grad_norm": 1.171875, "learning_rate": 1.2576131481752018e-05, "loss": 2.0907, "step": 2105 }, { "epoch": 1.4290551981036235, "grad_norm": 1.1875, "learning_rate": 1.2538024263927128e-05, "loss": 2.0992, "step": 2110 }, { "epoch": 1.4324415848289875, "grad_norm": 1.203125, "learning_rate": 1.2499877611423241e-05, "loss": 2.0954, "step": 2115 }, { "epoch": 1.4358279715543514, "grad_norm": 1.21875, "learning_rate": 1.2461692116945878e-05, "loss": 2.1015, "step": 2120 }, { "epoch": 1.4392143582797154, "grad_norm": 1.2421875, "learning_rate": 1.242346837380407e-05, "loss": 2.086, "step": 2125 }, { "epoch": 1.4426007450050795, "grad_norm": 1.1875, "learning_rate": 1.238520697590114e-05, "loss": 2.0897, "step": 2130 }, { "epoch": 1.4459871317304436, "grad_norm": 1.2265625, "learning_rate": 1.2346908517725472e-05, "loss": 2.09, "step": 2135 }, { "epoch": 1.4493735184558076, "grad_norm": 1.171875, "learning_rate": 1.2308573594341275e-05, "loss": 2.0932, "step": 2140 }, { "epoch": 1.4527599051811717, "grad_norm": 1.1875, "learning_rate": 1.227020280137934e-05, "loss": 2.0823, "step": 2145 }, { "epoch": 1.4561462919065358, "grad_norm": 1.2421875, "learning_rate": 1.2231796735027781e-05, "loss": 2.0954, "step": 2150 }, { "epoch": 1.4595326786318998, "grad_norm": 1.2109375, "learning_rate": 1.2193355992022778e-05, "loss": 2.0976, "step": 2155 }, { "epoch": 1.462919065357264, "grad_norm": 1.1953125, "learning_rate": 1.21548811696393e-05, "loss": 2.0933, "step": 2160 }, { "epoch": 1.466305452082628, "grad_norm": 1.1953125, "learning_rate": 1.2116372865681824e-05, "loss": 2.097, "step": 2165 }, { "epoch": 1.4696918388079918, "grad_norm": 1.2421875, "learning_rate": 1.2077831678475047e-05, "loss": 2.1055, "step": 2170 }, { "epoch": 1.4730782255333559, "grad_norm": 1.1875, "learning_rate": 1.2039258206854594e-05, "loss": 2.0978, "step": 2175 }, { "epoch": 1.47646461225872, "grad_norm": 1.203125, "learning_rate": 1.2000653050157713e-05, "loss": 2.0939, "step": 2180 }, { "epoch": 1.479850998984084, "grad_norm": 1.2109375, "learning_rate": 1.1962016808213958e-05, "loss": 2.0923, "step": 2185 }, { "epoch": 1.483237385709448, "grad_norm": 1.203125, "learning_rate": 1.1923350081335871e-05, "loss": 2.0995, "step": 2190 }, { "epoch": 1.486623772434812, "grad_norm": 1.203125, "learning_rate": 1.1884653470309655e-05, "loss": 2.0904, "step": 2195 }, { "epoch": 1.490010159160176, "grad_norm": 1.21875, "learning_rate": 1.1845927576385838e-05, "loss": 2.1105, "step": 2200 }, { "epoch": 1.49339654588554, "grad_norm": 1.21875, "learning_rate": 1.1807173001269938e-05, "loss": 2.1011, "step": 2205 }, { "epoch": 1.4967829326109041, "grad_norm": 1.1484375, "learning_rate": 1.17683903471131e-05, "loss": 2.1078, "step": 2210 }, { "epoch": 1.5001693193362682, "grad_norm": 1.203125, "learning_rate": 1.1729580216502765e-05, "loss": 2.0945, "step": 2215 }, { "epoch": 1.5035557060616322, "grad_norm": 1.1796875, "learning_rate": 1.1690743212453265e-05, "loss": 2.0874, "step": 2220 }, { "epoch": 1.5069420927869963, "grad_norm": 1.21875, "learning_rate": 1.1651879938396505e-05, "loss": 2.1066, "step": 2225 }, { "epoch": 1.5103284795123604, "grad_norm": 1.234375, "learning_rate": 1.1612990998172544e-05, "loss": 2.0968, "step": 2230 }, { "epoch": 1.5137148662377244, "grad_norm": 1.203125, "learning_rate": 1.1574076996020238e-05, "loss": 2.0896, "step": 2235 }, { "epoch": 1.5171012529630885, "grad_norm": 1.203125, "learning_rate": 1.1535138536567845e-05, "loss": 2.0805, "step": 2240 }, { "epoch": 1.5204876396884526, "grad_norm": 1.1953125, "learning_rate": 1.1496176224823623e-05, "loss": 2.105, "step": 2245 }, { "epoch": 1.5238740264138164, "grad_norm": 1.21875, "learning_rate": 1.1457190666166443e-05, "loss": 2.076, "step": 2250 }, { "epoch": 1.5272604131391805, "grad_norm": 1.1875, "learning_rate": 1.1418182466336372e-05, "loss": 2.0935, "step": 2255 }, { "epoch": 1.5306467998645445, "grad_norm": 1.203125, "learning_rate": 1.1379152231425268e-05, "loss": 2.064, "step": 2260 }, { "epoch": 1.5340331865899086, "grad_norm": 1.203125, "learning_rate": 1.1340100567867358e-05, "loss": 2.1166, "step": 2265 }, { "epoch": 1.5374195733152725, "grad_norm": 1.2109375, "learning_rate": 1.1301028082429823e-05, "loss": 2.1109, "step": 2270 }, { "epoch": 1.5408059600406365, "grad_norm": 1.1953125, "learning_rate": 1.1261935382203353e-05, "loss": 2.0978, "step": 2275 }, { "epoch": 1.5441923467660006, "grad_norm": 1.234375, "learning_rate": 1.1222823074592737e-05, "loss": 2.0733, "step": 2280 }, { "epoch": 1.5475787334913647, "grad_norm": 1.2109375, "learning_rate": 1.1183691767307412e-05, "loss": 2.1078, "step": 2285 }, { "epoch": 1.5509651202167287, "grad_norm": 1.2421875, "learning_rate": 1.1144542068352026e-05, "loss": 2.0953, "step": 2290 }, { "epoch": 1.5543515069420928, "grad_norm": 1.234375, "learning_rate": 1.1105374586016984e-05, "loss": 2.1037, "step": 2295 }, { "epoch": 1.5577378936674569, "grad_norm": 1.25, "learning_rate": 1.1066189928868996e-05, "loss": 2.1059, "step": 2300 }, { "epoch": 1.561124280392821, "grad_norm": 1.171875, "learning_rate": 1.102698870574164e-05, "loss": 2.0744, "step": 2305 }, { "epoch": 1.564510667118185, "grad_norm": 1.2109375, "learning_rate": 1.0987771525725882e-05, "loss": 2.0996, "step": 2310 }, { "epoch": 1.567897053843549, "grad_norm": 1.1796875, "learning_rate": 1.0948538998160614e-05, "loss": 2.0866, "step": 2315 }, { "epoch": 1.5712834405689131, "grad_norm": 1.1953125, "learning_rate": 1.0909291732623202e-05, "loss": 2.0929, "step": 2320 }, { "epoch": 1.574669827294277, "grad_norm": 1.203125, "learning_rate": 1.0870030338919995e-05, "loss": 2.097, "step": 2325 }, { "epoch": 1.578056214019641, "grad_norm": 1.203125, "learning_rate": 1.0830755427076865e-05, "loss": 2.0992, "step": 2330 }, { "epoch": 1.581442600745005, "grad_norm": 1.2109375, "learning_rate": 1.0791467607329713e-05, "loss": 2.0996, "step": 2335 }, { "epoch": 1.5848289874703692, "grad_norm": 1.1640625, "learning_rate": 1.0752167490115012e-05, "loss": 2.1022, "step": 2340 }, { "epoch": 1.588215374195733, "grad_norm": 1.1875, "learning_rate": 1.0712855686060293e-05, "loss": 2.0803, "step": 2345 }, { "epoch": 1.591601760921097, "grad_norm": 1.2265625, "learning_rate": 1.0673532805974687e-05, "loss": 2.0929, "step": 2350 }, { "epoch": 1.5949881476464611, "grad_norm": 1.203125, "learning_rate": 1.0634199460839403e-05, "loss": 2.0993, "step": 2355 }, { "epoch": 1.5983745343718252, "grad_norm": 1.2265625, "learning_rate": 1.0594856261798264e-05, "loss": 2.1005, "step": 2360 }, { "epoch": 1.6017609210971893, "grad_norm": 1.1953125, "learning_rate": 1.055550382014819e-05, "loss": 2.074, "step": 2365 }, { "epoch": 1.6051473078225533, "grad_norm": 1.2265625, "learning_rate": 1.0516142747329715e-05, "loss": 2.0856, "step": 2370 }, { "epoch": 1.6085336945479174, "grad_norm": 1.2109375, "learning_rate": 1.0476773654917476e-05, "loss": 2.0722, "step": 2375 }, { "epoch": 1.6119200812732815, "grad_norm": 1.2109375, "learning_rate": 1.0437397154610717e-05, "loss": 2.106, "step": 2380 }, { "epoch": 1.6153064679986455, "grad_norm": 1.203125, "learning_rate": 1.0398013858223784e-05, "loss": 2.0995, "step": 2385 }, { "epoch": 1.6186928547240096, "grad_norm": 1.21875, "learning_rate": 1.035862437767661e-05, "loss": 2.114, "step": 2390 }, { "epoch": 1.6220792414493737, "grad_norm": 1.1953125, "learning_rate": 1.0319229324985228e-05, "loss": 2.118, "step": 2395 }, { "epoch": 1.6254656281747377, "grad_norm": 1.1953125, "learning_rate": 1.027982931225223e-05, "loss": 2.0772, "step": 2400 }, { "epoch": 1.6288520149001016, "grad_norm": 1.2109375, "learning_rate": 1.0240424951657295e-05, "loss": 2.1139, "step": 2405 }, { "epoch": 1.6322384016254656, "grad_norm": 1.21875, "learning_rate": 1.020101685544764e-05, "loss": 2.0946, "step": 2410 }, { "epoch": 1.6356247883508297, "grad_norm": 1.2109375, "learning_rate": 1.0161605635928538e-05, "loss": 2.1095, "step": 2415 }, { "epoch": 1.6390111750761935, "grad_norm": 1.2265625, "learning_rate": 1.012219190545378e-05, "loss": 2.1012, "step": 2420 }, { "epoch": 1.6423975618015576, "grad_norm": 1.171875, "learning_rate": 1.0082776276416177e-05, "loss": 2.1011, "step": 2425 }, { "epoch": 1.6457839485269217, "grad_norm": 1.2265625, "learning_rate": 1.0043359361238036e-05, "loss": 2.0969, "step": 2430 }, { "epoch": 1.6491703352522857, "grad_norm": 1.203125, "learning_rate": 1.0003941772361651e-05, "loss": 2.0733, "step": 2435 }, { "epoch": 1.6525567219776498, "grad_norm": 1.1953125, "learning_rate": 9.96452412223978e-06, "loss": 2.095, "step": 2440 }, { "epoch": 1.6559431087030139, "grad_norm": 1.1953125, "learning_rate": 9.925107023326137e-06, "loss": 2.0997, "step": 2445 }, { "epoch": 1.659329495428378, "grad_norm": 1.1796875, "learning_rate": 9.885691088065866e-06, "loss": 2.1065, "step": 2450 }, { "epoch": 1.662715882153742, "grad_norm": 1.265625, "learning_rate": 9.846276928886032e-06, "loss": 2.1, "step": 2455 }, { "epoch": 1.666102268879106, "grad_norm": 1.1796875, "learning_rate": 9.80686515818611e-06, "loss": 2.0979, "step": 2460 }, { "epoch": 1.6694886556044701, "grad_norm": 1.2109375, "learning_rate": 9.767456388328454e-06, "loss": 2.0939, "step": 2465 }, { "epoch": 1.6728750423298342, "grad_norm": 1.1796875, "learning_rate": 9.728051231628802e-06, "loss": 2.0998, "step": 2470 }, { "epoch": 1.6762614290551983, "grad_norm": 1.25, "learning_rate": 9.688650300346749e-06, "loss": 2.0978, "step": 2475 }, { "epoch": 1.679647815780562, "grad_norm": 1.234375, "learning_rate": 9.649254206676226e-06, "loss": 2.0911, "step": 2480 }, { "epoch": 1.6830342025059262, "grad_norm": 1.2109375, "learning_rate": 9.609863562736023e-06, "loss": 2.0819, "step": 2485 }, { "epoch": 1.6864205892312902, "grad_norm": 1.1953125, "learning_rate": 9.570478980560233e-06, "loss": 2.1249, "step": 2490 }, { "epoch": 1.6898069759566543, "grad_norm": 1.1640625, "learning_rate": 9.531101072088779e-06, "loss": 2.0797, "step": 2495 }, { "epoch": 1.6931933626820181, "grad_norm": 1.1875, "learning_rate": 9.491730449157878e-06, "loss": 2.0914, "step": 2500 }, { "epoch": 1.6965797494073822, "grad_norm": 1.203125, "learning_rate": 9.452367723490553e-06, "loss": 2.1069, "step": 2505 }, { "epoch": 1.6999661361327463, "grad_norm": 1.1875, "learning_rate": 9.41301350668713e-06, "loss": 2.1065, "step": 2510 }, { "epoch": 1.7033525228581103, "grad_norm": 1.234375, "learning_rate": 9.373668410215717e-06, "loss": 2.0948, "step": 2515 }, { "epoch": 1.7067389095834744, "grad_norm": 1.2109375, "learning_rate": 9.334333045402721e-06, "loss": 2.1243, "step": 2520 }, { "epoch": 1.7101252963088385, "grad_norm": 1.25, "learning_rate": 9.29500802342334e-06, "loss": 2.09, "step": 2525 }, { "epoch": 1.7135116830342025, "grad_norm": 1.203125, "learning_rate": 9.255693955292072e-06, "loss": 2.0846, "step": 2530 }, { "epoch": 1.7168980697595666, "grad_norm": 1.1796875, "learning_rate": 9.216391451853218e-06, "loss": 2.0937, "step": 2535 }, { "epoch": 1.7202844564849307, "grad_norm": 1.2109375, "learning_rate": 9.177101123771387e-06, "loss": 2.0945, "step": 2540 }, { "epoch": 1.7236708432102947, "grad_norm": 1.21875, "learning_rate": 9.137823581522024e-06, "loss": 2.0997, "step": 2545 }, { "epoch": 1.7270572299356588, "grad_norm": 1.21875, "learning_rate": 9.098559435381903e-06, "loss": 2.0878, "step": 2550 }, { "epoch": 1.7304436166610226, "grad_norm": 1.203125, "learning_rate": 9.059309295419665e-06, "loss": 2.0783, "step": 2555 }, { "epoch": 1.7338300033863867, "grad_norm": 1.203125, "learning_rate": 9.020073771486319e-06, "loss": 2.0842, "step": 2560 }, { "epoch": 1.7372163901117508, "grad_norm": 1.2109375, "learning_rate": 8.980853473205776e-06, "loss": 2.1148, "step": 2565 }, { "epoch": 1.7406027768371148, "grad_norm": 1.2109375, "learning_rate": 8.941649009965393e-06, "loss": 2.1094, "step": 2570 }, { "epoch": 1.7439891635624787, "grad_norm": 1.1875, "learning_rate": 8.902460990906474e-06, "loss": 2.0912, "step": 2575 }, { "epoch": 1.7473755502878427, "grad_norm": 1.2265625, "learning_rate": 8.863290024914828e-06, "loss": 2.0935, "step": 2580 }, { "epoch": 1.7507619370132068, "grad_norm": 1.2109375, "learning_rate": 8.8241367206113e-06, "loss": 2.0992, "step": 2585 }, { "epoch": 1.7541483237385709, "grad_norm": 1.21875, "learning_rate": 8.78500168634231e-06, "loss": 2.0929, "step": 2590 }, { "epoch": 1.757534710463935, "grad_norm": 1.1796875, "learning_rate": 8.745885530170418e-06, "loss": 2.1136, "step": 2595 }, { "epoch": 1.760921097189299, "grad_norm": 1.1640625, "learning_rate": 8.70678885986485e-06, "loss": 2.0984, "step": 2600 }, { "epoch": 1.764307483914663, "grad_norm": 1.203125, "learning_rate": 8.667712282892084e-06, "loss": 2.089, "step": 2605 }, { "epoch": 1.7676938706400271, "grad_norm": 1.1796875, "learning_rate": 8.628656406406388e-06, "loss": 2.0945, "step": 2610 }, { "epoch": 1.7710802573653912, "grad_norm": 1.1875, "learning_rate": 8.589621837240393e-06, "loss": 2.0941, "step": 2615 }, { "epoch": 1.7744666440907553, "grad_norm": 1.234375, "learning_rate": 8.550609181895679e-06, "loss": 2.1016, "step": 2620 }, { "epoch": 1.7778530308161193, "grad_norm": 1.1953125, "learning_rate": 8.511619046533319e-06, "loss": 2.082, "step": 2625 }, { "epoch": 1.7812394175414832, "grad_norm": 1.21875, "learning_rate": 8.472652036964504e-06, "loss": 2.0971, "step": 2630 }, { "epoch": 1.7846258042668472, "grad_norm": 1.2109375, "learning_rate": 8.433708758641085e-06, "loss": 2.1035, "step": 2635 }, { "epoch": 1.7880121909922113, "grad_norm": 1.2734375, "learning_rate": 8.394789816646207e-06, "loss": 2.1157, "step": 2640 }, { "epoch": 1.7913985777175754, "grad_norm": 1.1796875, "learning_rate": 8.355895815684875e-06, "loss": 2.0962, "step": 2645 }, { "epoch": 1.7947849644429392, "grad_norm": 1.1953125, "learning_rate": 8.31702736007457e-06, "loss": 2.1044, "step": 2650 }, { "epoch": 1.7981713511683033, "grad_norm": 1.2109375, "learning_rate": 8.278185053735874e-06, "loss": 2.0846, "step": 2655 }, { "epoch": 1.8015577378936674, "grad_norm": 1.1953125, "learning_rate": 8.239369500183057e-06, "loss": 2.0979, "step": 2660 }, { "epoch": 1.8049441246190314, "grad_norm": 1.1953125, "learning_rate": 8.200581302514733e-06, "loss": 2.1079, "step": 2665 }, { "epoch": 1.8083305113443955, "grad_norm": 1.265625, "learning_rate": 8.161821063404458e-06, "loss": 2.1269, "step": 2670 }, { "epoch": 1.8117168980697596, "grad_norm": 1.2265625, "learning_rate": 8.123089385091388e-06, "loss": 2.1159, "step": 2675 }, { "epoch": 1.8151032847951236, "grad_norm": 1.21875, "learning_rate": 8.084386869370917e-06, "loss": 2.1139, "step": 2680 }, { "epoch": 1.8184896715204877, "grad_norm": 1.1953125, "learning_rate": 8.04571411758531e-06, "loss": 2.105, "step": 2685 }, { "epoch": 1.8218760582458517, "grad_norm": 1.1875, "learning_rate": 8.007071730614395e-06, "loss": 2.0957, "step": 2690 }, { "epoch": 1.8252624449712158, "grad_norm": 1.1796875, "learning_rate": 7.968460308866187e-06, "loss": 2.1123, "step": 2695 }, { "epoch": 1.8286488316965799, "grad_norm": 1.2109375, "learning_rate": 7.92988045226758e-06, "loss": 2.1086, "step": 2700 }, { "epoch": 1.832035218421944, "grad_norm": 1.1796875, "learning_rate": 7.891332760255033e-06, "loss": 2.0779, "step": 2705 }, { "epoch": 1.8354216051473078, "grad_norm": 1.234375, "learning_rate": 7.852817831765235e-06, "loss": 2.0877, "step": 2710 }, { "epoch": 1.8388079918726719, "grad_norm": 1.1875, "learning_rate": 7.814336265225819e-06, "loss": 2.0794, "step": 2715 }, { "epoch": 1.842194378598036, "grad_norm": 1.1640625, "learning_rate": 7.775888658546046e-06, "loss": 2.098, "step": 2720 }, { "epoch": 1.8455807653234, "grad_norm": 1.1796875, "learning_rate": 7.737475609107528e-06, "loss": 2.0884, "step": 2725 }, { "epoch": 1.8489671520487638, "grad_norm": 1.1953125, "learning_rate": 7.69909771375495e-06, "loss": 2.1047, "step": 2730 }, { "epoch": 1.852353538774128, "grad_norm": 1.1796875, "learning_rate": 7.660755568786771e-06, "loss": 2.0949, "step": 2735 }, { "epoch": 1.855739925499492, "grad_norm": 1.1953125, "learning_rate": 7.6224497699459965e-06, "loss": 2.0708, "step": 2740 }, { "epoch": 1.859126312224856, "grad_norm": 1.265625, "learning_rate": 7.584180912410888e-06, "loss": 2.0953, "step": 2745 }, { "epoch": 1.86251269895022, "grad_norm": 1.1796875, "learning_rate": 7.545949590785737e-06, "loss": 2.0715, "step": 2750 }, { "epoch": 1.8658990856755842, "grad_norm": 1.21875, "learning_rate": 7.507756399091614e-06, "loss": 2.0796, "step": 2755 }, { "epoch": 1.8692854724009482, "grad_norm": 1.2109375, "learning_rate": 7.469601930757142e-06, "loss": 2.0941, "step": 2760 }, { "epoch": 1.8726718591263123, "grad_norm": 1.203125, "learning_rate": 7.431486778609291e-06, "loss": 2.0884, "step": 2765 }, { "epoch": 1.8760582458516764, "grad_norm": 1.2265625, "learning_rate": 7.393411534864139e-06, "loss": 2.1029, "step": 2770 }, { "epoch": 1.8794446325770404, "grad_norm": 1.1875, "learning_rate": 7.355376791117692e-06, "loss": 2.0997, "step": 2775 }, { "epoch": 1.8828310193024045, "grad_norm": 1.1953125, "learning_rate": 7.3173831383366855e-06, "loss": 2.0989, "step": 2780 }, { "epoch": 1.8862174060277683, "grad_norm": 1.2109375, "learning_rate": 7.279431166849394e-06, "loss": 2.1009, "step": 2785 }, { "epoch": 1.8896037927531324, "grad_norm": 1.1875, "learning_rate": 7.241521466336485e-06, "loss": 2.1078, "step": 2790 }, { "epoch": 1.8929901794784965, "grad_norm": 1.203125, "learning_rate": 7.203654625821815e-06, "loss": 2.0946, "step": 2795 }, { "epoch": 1.8963765662038605, "grad_norm": 1.15625, "learning_rate": 7.165831233663324e-06, "loss": 2.0832, "step": 2800 }, { "epoch": 1.8997629529292244, "grad_norm": 1.21875, "learning_rate": 7.1280518775438555e-06, "loss": 2.0737, "step": 2805 }, { "epoch": 1.9031493396545884, "grad_norm": 1.203125, "learning_rate": 7.090317144462045e-06, "loss": 2.1022, "step": 2810 }, { "epoch": 1.9065357263799525, "grad_norm": 1.1875, "learning_rate": 7.052627620723196e-06, "loss": 2.1055, "step": 2815 }, { "epoch": 1.9099221131053166, "grad_norm": 1.234375, "learning_rate": 7.014983891930167e-06, "loss": 2.1058, "step": 2820 }, { "epoch": 1.9133084998306806, "grad_norm": 1.1953125, "learning_rate": 6.977386542974285e-06, "loss": 2.0923, "step": 2825 }, { "epoch": 1.9166948865560447, "grad_norm": 1.2265625, "learning_rate": 6.939836158026236e-06, "loss": 2.1055, "step": 2830 }, { "epoch": 1.9200812732814088, "grad_norm": 1.1953125, "learning_rate": 6.9023333205270024e-06, "loss": 2.113, "step": 2835 }, { "epoch": 1.9234676600067728, "grad_norm": 1.2265625, "learning_rate": 6.864878613178805e-06, "loss": 2.103, "step": 2840 }, { "epoch": 1.926854046732137, "grad_norm": 1.2109375, "learning_rate": 6.8274726179360285e-06, "loss": 2.0934, "step": 2845 }, { "epoch": 1.930240433457501, "grad_norm": 1.2109375, "learning_rate": 6.790115915996199e-06, "loss": 2.0839, "step": 2850 }, { "epoch": 1.933626820182865, "grad_norm": 1.234375, "learning_rate": 6.752809087790934e-06, "loss": 2.0942, "step": 2855 }, { "epoch": 1.9370132069082289, "grad_norm": 1.2265625, "learning_rate": 6.715552712976956e-06, "loss": 2.0843, "step": 2860 }, { "epoch": 1.940399593633593, "grad_norm": 1.203125, "learning_rate": 6.678347370427047e-06, "loss": 2.0934, "step": 2865 }, { "epoch": 1.943785980358957, "grad_norm": 1.1796875, "learning_rate": 6.641193638221075e-06, "loss": 2.0906, "step": 2870 }, { "epoch": 1.947172367084321, "grad_norm": 1.2109375, "learning_rate": 6.604092093637018e-06, "loss": 2.1081, "step": 2875 }, { "epoch": 1.950558753809685, "grad_norm": 1.203125, "learning_rate": 6.567043313141976e-06, "loss": 2.0758, "step": 2880 }, { "epoch": 1.953945140535049, "grad_norm": 1.1953125, "learning_rate": 6.5300478723832385e-06, "loss": 2.0903, "step": 2885 }, { "epoch": 1.957331527260413, "grad_norm": 1.15625, "learning_rate": 6.493106346179312e-06, "loss": 2.1073, "step": 2890 }, { "epoch": 1.960717913985777, "grad_norm": 1.203125, "learning_rate": 6.456219308511005e-06, "loss": 2.088, "step": 2895 }, { "epoch": 1.9641043007111412, "grad_norm": 1.1875, "learning_rate": 6.4193873325125186e-06, "loss": 2.0892, "step": 2900 }, { "epoch": 1.9674906874365052, "grad_norm": 1.234375, "learning_rate": 6.382610990462519e-06, "loss": 2.1027, "step": 2905 }, { "epoch": 1.9708770741618693, "grad_norm": 1.2421875, "learning_rate": 6.3458908537752616e-06, "loss": 2.0747, "step": 2910 }, { "epoch": 1.9742634608872334, "grad_norm": 1.203125, "learning_rate": 6.309227492991708e-06, "loss": 2.1045, "step": 2915 }, { "epoch": 1.9776498476125974, "grad_norm": 1.203125, "learning_rate": 6.272621477770655e-06, "loss": 2.1038, "step": 2920 }, { "epoch": 1.9810362343379615, "grad_norm": 1.1875, "learning_rate": 6.236073376879907e-06, "loss": 2.0945, "step": 2925 }, { "epoch": 1.9844226210633256, "grad_norm": 1.1875, "learning_rate": 6.1995837581874e-06, "loss": 2.0771, "step": 2930 }, { "epoch": 1.9878090077886896, "grad_norm": 1.203125, "learning_rate": 6.1631531886524175e-06, "loss": 2.0948, "step": 2935 }, { "epoch": 1.9911953945140535, "grad_norm": 1.1875, "learning_rate": 6.126782234316752e-06, "loss": 2.0919, "step": 2940 }, { "epoch": 1.9945817812394175, "grad_norm": 1.1953125, "learning_rate": 6.090471460295928e-06, "loss": 2.0973, "step": 2945 }, { "epoch": 1.9979681679647816, "grad_norm": 1.1640625, "learning_rate": 6.054221430770416e-06, "loss": 2.0925, "step": 2950 }, { "epoch": 2.0, "eval_loss": 2.1174328327178955, "eval_runtime": 85.4558, "eval_samples_per_second": 15.271, "eval_steps_per_second": 1.919, "step": 2953 }, { "epoch": 2.0013545546901454, "grad_norm": 1.2265625, "learning_rate": 6.018032708976857e-06, "loss": 2.0953, "step": 2955 }, { "epoch": 2.0047409414155095, "grad_norm": 1.171875, "learning_rate": 5.981905857199335e-06, "loss": 2.0947, "step": 2960 }, { "epoch": 2.0081273281408736, "grad_norm": 1.21875, "learning_rate": 5.945841436760612e-06, "loss": 2.1016, "step": 2965 }, { "epoch": 2.0115137148662376, "grad_norm": 1.1953125, "learning_rate": 5.909840008013428e-06, "loss": 2.0855, "step": 2970 }, { "epoch": 2.0149001015916017, "grad_norm": 1.2109375, "learning_rate": 5.8739021303317825e-06, "loss": 2.0817, "step": 2975 }, { "epoch": 2.0182864883169658, "grad_norm": 1.2109375, "learning_rate": 5.838028362102246e-06, "loss": 2.0723, "step": 2980 }, { "epoch": 2.02167287504233, "grad_norm": 1.1875, "learning_rate": 5.80221926071529e-06, "loss": 2.0643, "step": 2985 }, { "epoch": 2.025059261767694, "grad_norm": 1.203125, "learning_rate": 5.7664753825566155e-06, "loss": 2.0711, "step": 2990 }, { "epoch": 2.028445648493058, "grad_norm": 1.2265625, "learning_rate": 5.7307972829985234e-06, "loss": 2.1014, "step": 2995 }, { "epoch": 2.031832035218422, "grad_norm": 1.1953125, "learning_rate": 5.69518551639127e-06, "loss": 2.0799, "step": 3000 }, { "epoch": 2.035218421943786, "grad_norm": 1.171875, "learning_rate": 5.659640636054453e-06, "loss": 2.0754, "step": 3005 }, { "epoch": 2.03860480866915, "grad_norm": 1.203125, "learning_rate": 5.624163194268441e-06, "loss": 2.0847, "step": 3010 }, { "epoch": 2.0419911953945142, "grad_norm": 1.1640625, "learning_rate": 5.588753742265751e-06, "loss": 2.082, "step": 3015 }, { "epoch": 2.0453775821198783, "grad_norm": 1.203125, "learning_rate": 5.553412830222523e-06, "loss": 2.0866, "step": 3020 }, { "epoch": 2.048763968845242, "grad_norm": 1.2109375, "learning_rate": 5.518141007249944e-06, "loss": 2.1161, "step": 3025 }, { "epoch": 2.052150355570606, "grad_norm": 1.1953125, "learning_rate": 5.482938821385722e-06, "loss": 2.073, "step": 3030 }, { "epoch": 2.05553674229597, "grad_norm": 1.25, "learning_rate": 5.44780681958559e-06, "loss": 2.0779, "step": 3035 }, { "epoch": 2.058923129021334, "grad_norm": 1.2109375, "learning_rate": 5.412745547714777e-06, "loss": 2.0771, "step": 3040 }, { "epoch": 2.062309515746698, "grad_norm": 1.1953125, "learning_rate": 5.377755550539559e-06, "loss": 2.0847, "step": 3045 }, { "epoch": 2.0656959024720623, "grad_norm": 1.2265625, "learning_rate": 5.3428373717187575e-06, "loss": 2.1026, "step": 3050 }, { "epoch": 2.0690822891974263, "grad_norm": 1.2109375, "learning_rate": 5.307991553795328e-06, "loss": 2.1031, "step": 3055 }, { "epoch": 2.0724686759227904, "grad_norm": 1.1875, "learning_rate": 5.2732186381879115e-06, "loss": 2.0797, "step": 3060 }, { "epoch": 2.0758550626481544, "grad_norm": 1.21875, "learning_rate": 5.238519165182415e-06, "loss": 2.0848, "step": 3065 }, { "epoch": 2.0792414493735185, "grad_norm": 1.203125, "learning_rate": 5.203893673923649e-06, "loss": 2.0678, "step": 3070 }, { "epoch": 2.0826278360988826, "grad_norm": 1.21875, "learning_rate": 5.169342702406907e-06, "loss": 2.0913, "step": 3075 }, { "epoch": 2.0860142228242466, "grad_norm": 1.171875, "learning_rate": 5.1348667874696455e-06, "loss": 2.0952, "step": 3080 }, { "epoch": 2.0894006095496107, "grad_norm": 1.2421875, "learning_rate": 5.100466464783113e-06, "loss": 2.0562, "step": 3085 }, { "epoch": 2.0927869962749748, "grad_norm": 1.1953125, "learning_rate": 5.066142268844043e-06, "loss": 2.0847, "step": 3090 }, { "epoch": 2.096173383000339, "grad_norm": 1.21875, "learning_rate": 5.0318947329663545e-06, "loss": 2.1097, "step": 3095 }, { "epoch": 2.0995597697257025, "grad_norm": 1.1796875, "learning_rate": 4.997724389272841e-06, "loss": 2.0795, "step": 3100 }, { "epoch": 2.1029461564510665, "grad_norm": 1.1953125, "learning_rate": 4.963631768686937e-06, "loss": 2.0798, "step": 3105 }, { "epoch": 2.1063325431764306, "grad_norm": 1.203125, "learning_rate": 4.929617400924435e-06, "loss": 2.1103, "step": 3110 }, { "epoch": 2.1097189299017947, "grad_norm": 1.2265625, "learning_rate": 4.895681814485277e-06, "loss": 2.0683, "step": 3115 }, { "epoch": 2.1131053166271587, "grad_norm": 1.203125, "learning_rate": 4.861825536645335e-06, "loss": 2.0845, "step": 3120 }, { "epoch": 2.116491703352523, "grad_norm": 1.1875, "learning_rate": 4.828049093448216e-06, "loss": 2.0827, "step": 3125 }, { "epoch": 2.119878090077887, "grad_norm": 1.2265625, "learning_rate": 4.794353009697105e-06, "loss": 2.0843, "step": 3130 }, { "epoch": 2.123264476803251, "grad_norm": 1.21875, "learning_rate": 4.760737808946585e-06, "loss": 2.0966, "step": 3135 }, { "epoch": 2.126650863528615, "grad_norm": 1.234375, "learning_rate": 4.727204013494515e-06, "loss": 2.1099, "step": 3140 }, { "epoch": 2.130037250253979, "grad_norm": 1.1875, "learning_rate": 4.693752144373929e-06, "loss": 2.0737, "step": 3145 }, { "epoch": 2.133423636979343, "grad_norm": 1.2265625, "learning_rate": 4.660382721344909e-06, "loss": 2.0899, "step": 3150 }, { "epoch": 2.136810023704707, "grad_norm": 1.234375, "learning_rate": 4.627096262886542e-06, "loss": 2.0803, "step": 3155 }, { "epoch": 2.1401964104300712, "grad_norm": 1.234375, "learning_rate": 4.59389328618884e-06, "loss": 2.0913, "step": 3160 }, { "epoch": 2.1435827971554353, "grad_norm": 1.203125, "learning_rate": 4.5607743071447085e-06, "loss": 2.0812, "step": 3165 }, { "epoch": 2.1469691838807994, "grad_norm": 1.2109375, "learning_rate": 4.52773984034195e-06, "loss": 2.0924, "step": 3170 }, { "epoch": 2.1503555706061634, "grad_norm": 1.2421875, "learning_rate": 4.494790399055242e-06, "loss": 2.1026, "step": 3175 }, { "epoch": 2.153741957331527, "grad_norm": 1.203125, "learning_rate": 4.461926495238178e-06, "loss": 2.0782, "step": 3180 }, { "epoch": 2.157128344056891, "grad_norm": 1.2265625, "learning_rate": 4.429148639515304e-06, "loss": 2.0978, "step": 3185 }, { "epoch": 2.160514730782255, "grad_norm": 1.2109375, "learning_rate": 4.3964573411741994e-06, "loss": 2.0553, "step": 3190 }, { "epoch": 2.1639011175076193, "grad_norm": 1.234375, "learning_rate": 4.363853108157545e-06, "loss": 2.0731, "step": 3195 }, { "epoch": 2.1672875042329833, "grad_norm": 1.1875, "learning_rate": 4.33133644705524e-06, "loss": 2.0734, "step": 3200 }, { "epoch": 2.1706738909583474, "grad_norm": 1.234375, "learning_rate": 4.29890786309654e-06, "loss": 2.1203, "step": 3205 }, { "epoch": 2.1740602776837115, "grad_norm": 1.25, "learning_rate": 4.266567860142184e-06, "loss": 2.0883, "step": 3210 }, { "epoch": 2.1774466644090755, "grad_norm": 1.2109375, "learning_rate": 4.234316940676592e-06, "loss": 2.0996, "step": 3215 }, { "epoch": 2.1808330511344396, "grad_norm": 1.21875, "learning_rate": 4.202155605800035e-06, "loss": 2.0974, "step": 3220 }, { "epoch": 2.1842194378598037, "grad_norm": 1.21875, "learning_rate": 4.170084355220862e-06, "loss": 2.0774, "step": 3225 }, { "epoch": 2.1876058245851677, "grad_norm": 1.25, "learning_rate": 4.13810368724774e-06, "loss": 2.0841, "step": 3230 }, { "epoch": 2.190992211310532, "grad_norm": 1.2109375, "learning_rate": 4.1062140987818934e-06, "loss": 2.1023, "step": 3235 }, { "epoch": 2.194378598035896, "grad_norm": 1.21875, "learning_rate": 4.074416085309399e-06, "loss": 2.0651, "step": 3240 }, { "epoch": 2.19776498476126, "grad_norm": 1.1875, "learning_rate": 4.042710140893482e-06, "loss": 2.0787, "step": 3245 }, { "epoch": 2.201151371486624, "grad_norm": 1.203125, "learning_rate": 4.011096758166837e-06, "loss": 2.066, "step": 3250 }, { "epoch": 2.2045377582119876, "grad_norm": 1.21875, "learning_rate": 3.979576428323988e-06, "loss": 2.1087, "step": 3255 }, { "epoch": 2.2079241449373517, "grad_norm": 1.203125, "learning_rate": 3.948149641113628e-06, "loss": 2.0741, "step": 3260 }, { "epoch": 2.2113105316627157, "grad_norm": 1.203125, "learning_rate": 3.916816884831044e-06, "loss": 2.0685, "step": 3265 }, { "epoch": 2.21469691838808, "grad_norm": 1.203125, "learning_rate": 3.8855786463105015e-06, "loss": 2.0564, "step": 3270 }, { "epoch": 2.218083305113444, "grad_norm": 1.203125, "learning_rate": 3.854435410917689e-06, "loss": 2.0938, "step": 3275 }, { "epoch": 2.221469691838808, "grad_norm": 1.234375, "learning_rate": 3.823387662542196e-06, "loss": 2.08, "step": 3280 }, { "epoch": 2.224856078564172, "grad_norm": 1.234375, "learning_rate": 3.7924358835899556e-06, "loss": 2.0755, "step": 3285 }, { "epoch": 2.228242465289536, "grad_norm": 1.2265625, "learning_rate": 3.76158055497579e-06, "loss": 2.0729, "step": 3290 }, { "epoch": 2.2316288520149, "grad_norm": 1.2421875, "learning_rate": 3.730822156115904e-06, "loss": 2.0929, "step": 3295 }, { "epoch": 2.235015238740264, "grad_norm": 1.25, "learning_rate": 3.7001611649204685e-06, "loss": 2.1104, "step": 3300 }, { "epoch": 2.2384016254656283, "grad_norm": 1.234375, "learning_rate": 3.6695980577861614e-06, "loss": 2.0777, "step": 3305 }, { "epoch": 2.2417880121909923, "grad_norm": 1.234375, "learning_rate": 3.6391333095887917e-06, "loss": 2.0583, "step": 3310 }, { "epoch": 2.2451743989163564, "grad_norm": 1.203125, "learning_rate": 3.6087673936759084e-06, "loss": 2.0685, "step": 3315 }, { "epoch": 2.2485607856417205, "grad_norm": 1.2109375, "learning_rate": 3.578500781859444e-06, "loss": 2.0925, "step": 3320 }, { "epoch": 2.2519471723670845, "grad_norm": 1.203125, "learning_rate": 3.548333944408404e-06, "loss": 2.0788, "step": 3325 }, { "epoch": 2.255333559092448, "grad_norm": 1.21875, "learning_rate": 3.518267350041529e-06, "loss": 2.0847, "step": 3330 }, { "epoch": 2.258719945817812, "grad_norm": 1.2109375, "learning_rate": 3.488301465920031e-06, "loss": 2.0857, "step": 3335 }, { "epoch": 2.2621063325431763, "grad_norm": 1.234375, "learning_rate": 3.458436757640341e-06, "loss": 2.0855, "step": 3340 }, { "epoch": 2.2654927192685403, "grad_norm": 1.2421875, "learning_rate": 3.428673689226851e-06, "loss": 2.1017, "step": 3345 }, { "epoch": 2.2688791059939044, "grad_norm": 1.21875, "learning_rate": 3.3990127231247337e-06, "loss": 2.0951, "step": 3350 }, { "epoch": 2.2722654927192685, "grad_norm": 1.234375, "learning_rate": 3.3694543201927288e-06, "loss": 2.0723, "step": 3355 }, { "epoch": 2.2756518794446325, "grad_norm": 1.2421875, "learning_rate": 3.3399989396959988e-06, "loss": 2.071, "step": 3360 }, { "epoch": 2.2790382661699966, "grad_norm": 1.2109375, "learning_rate": 3.3106470392989977e-06, "loss": 2.0926, "step": 3365 }, { "epoch": 2.2824246528953607, "grad_norm": 1.1953125, "learning_rate": 3.2813990750583434e-06, "loss": 2.0823, "step": 3370 }, { "epoch": 2.2858110396207247, "grad_norm": 1.21875, "learning_rate": 3.2522555014157454e-06, "loss": 2.0888, "step": 3375 }, { "epoch": 2.289197426346089, "grad_norm": 1.1796875, "learning_rate": 3.2232167711909368e-06, "loss": 2.0763, "step": 3380 }, { "epoch": 2.292583813071453, "grad_norm": 1.1875, "learning_rate": 3.194283335574637e-06, "loss": 2.0701, "step": 3385 }, { "epoch": 2.295970199796817, "grad_norm": 1.2265625, "learning_rate": 3.1654556441215577e-06, "loss": 2.0877, "step": 3390 }, { "epoch": 2.299356586522181, "grad_norm": 1.234375, "learning_rate": 3.1367341447433897e-06, "loss": 2.0859, "step": 3395 }, { "epoch": 2.302742973247545, "grad_norm": 1.2265625, "learning_rate": 3.1081192837018736e-06, "loss": 2.0796, "step": 3400 }, { "epoch": 2.3061293599729087, "grad_norm": 1.1953125, "learning_rate": 3.0796115056018383e-06, "loss": 2.0888, "step": 3405 }, { "epoch": 2.3095157466982728, "grad_norm": 1.203125, "learning_rate": 3.0512112533843185e-06, "loss": 2.0851, "step": 3410 }, { "epoch": 2.312902133423637, "grad_norm": 1.21875, "learning_rate": 3.0229189683196513e-06, "loss": 2.0888, "step": 3415 }, { "epoch": 2.316288520149001, "grad_norm": 1.1875, "learning_rate": 2.994735090000629e-06, "loss": 2.0817, "step": 3420 }, { "epoch": 2.319674906874365, "grad_norm": 1.2265625, "learning_rate": 2.9666600563356773e-06, "loss": 2.0871, "step": 3425 }, { "epoch": 2.323061293599729, "grad_norm": 1.203125, "learning_rate": 2.938694303542032e-06, "loss": 2.0758, "step": 3430 }, { "epoch": 2.326447680325093, "grad_norm": 1.25, "learning_rate": 2.910838266138979e-06, "loss": 2.08, "step": 3435 }, { "epoch": 2.329834067050457, "grad_norm": 1.2109375, "learning_rate": 2.8830923769410903e-06, "loss": 2.0799, "step": 3440 }, { "epoch": 2.333220453775821, "grad_norm": 1.2421875, "learning_rate": 2.855457067051507e-06, "loss": 2.0702, "step": 3445 }, { "epoch": 2.3366068405011853, "grad_norm": 1.21875, "learning_rate": 2.8279327658552447e-06, "loss": 2.0697, "step": 3450 }, { "epoch": 2.3399932272265493, "grad_norm": 1.203125, "learning_rate": 2.800519901012504e-06, "loss": 2.083, "step": 3455 }, { "epoch": 2.3433796139519134, "grad_norm": 1.1953125, "learning_rate": 2.77321889845205e-06, "loss": 2.0727, "step": 3460 }, { "epoch": 2.3467660006772775, "grad_norm": 1.171875, "learning_rate": 2.746030182364574e-06, "loss": 2.0728, "step": 3465 }, { "epoch": 2.3501523874026415, "grad_norm": 1.203125, "learning_rate": 2.718954175196109e-06, "loss": 2.0772, "step": 3470 }, { "epoch": 2.3535387741280056, "grad_norm": 1.25, "learning_rate": 2.6919912976414787e-06, "loss": 2.0827, "step": 3475 }, { "epoch": 2.3569251608533692, "grad_norm": 1.2109375, "learning_rate": 2.6651419686377335e-06, "loss": 2.0773, "step": 3480 }, { "epoch": 2.3603115475787337, "grad_norm": 1.171875, "learning_rate": 2.6384066053576742e-06, "loss": 2.0858, "step": 3485 }, { "epoch": 2.3636979343040974, "grad_norm": 1.1796875, "learning_rate": 2.611785623203341e-06, "loss": 2.1062, "step": 3490 }, { "epoch": 2.3670843210294614, "grad_norm": 1.21875, "learning_rate": 2.585279435799578e-06, "loss": 2.0904, "step": 3495 }, { "epoch": 2.3704707077548255, "grad_norm": 1.21875, "learning_rate": 2.558888454987598e-06, "loss": 2.0853, "step": 3500 }, { "epoch": 2.3738570944801896, "grad_norm": 1.2421875, "learning_rate": 2.532613090818582e-06, "loss": 2.0966, "step": 3505 }, { "epoch": 2.3772434812055536, "grad_norm": 1.21875, "learning_rate": 2.506453751547322e-06, "loss": 2.1083, "step": 3510 }, { "epoch": 2.3806298679309177, "grad_norm": 1.21875, "learning_rate": 2.4804108436258545e-06, "loss": 2.0822, "step": 3515 }, { "epoch": 2.3840162546562818, "grad_norm": 1.21875, "learning_rate": 2.454484771697171e-06, "loss": 2.0813, "step": 3520 }, { "epoch": 2.387402641381646, "grad_norm": 1.203125, "learning_rate": 2.428675938588908e-06, "loss": 2.0685, "step": 3525 }, { "epoch": 2.39078902810701, "grad_norm": 1.1796875, "learning_rate": 2.4029847453070987e-06, "loss": 2.0774, "step": 3530 }, { "epoch": 2.394175414832374, "grad_norm": 1.1875, "learning_rate": 2.37741159102995e-06, "loss": 2.0857, "step": 3535 }, { "epoch": 2.397561801557738, "grad_norm": 1.203125, "learning_rate": 2.3519568731016205e-06, "loss": 2.0696, "step": 3540 }, { "epoch": 2.400948188283102, "grad_norm": 1.2109375, "learning_rate": 2.326620987026069e-06, "loss": 2.0851, "step": 3545 }, { "epoch": 2.404334575008466, "grad_norm": 1.1796875, "learning_rate": 2.301404326460891e-06, "loss": 2.1071, "step": 3550 }, { "epoch": 2.4077209617338298, "grad_norm": 1.1875, "learning_rate": 2.276307283211209e-06, "loss": 2.0909, "step": 3555 }, { "epoch": 2.4111073484591943, "grad_norm": 1.171875, "learning_rate": 2.2513302472235922e-06, "loss": 2.09, "step": 3560 }, { "epoch": 2.414493735184558, "grad_norm": 1.2265625, "learning_rate": 2.2264736065799863e-06, "loss": 2.0854, "step": 3565 }, { "epoch": 2.417880121909922, "grad_norm": 1.203125, "learning_rate": 2.201737747491688e-06, "loss": 2.0796, "step": 3570 }, { "epoch": 2.421266508635286, "grad_norm": 1.2109375, "learning_rate": 2.177123054293345e-06, "loss": 2.0866, "step": 3575 }, { "epoch": 2.42465289536065, "grad_norm": 1.2265625, "learning_rate": 2.152629909436984e-06, "loss": 2.0796, "step": 3580 }, { "epoch": 2.428039282086014, "grad_norm": 1.21875, "learning_rate": 2.128258693486073e-06, "loss": 2.092, "step": 3585 }, { "epoch": 2.4314256688113782, "grad_norm": 1.21875, "learning_rate": 2.1040097851095933e-06, "loss": 2.1183, "step": 3590 }, { "epoch": 2.4348120555367423, "grad_norm": 1.234375, "learning_rate": 2.079883561076178e-06, "loss": 2.071, "step": 3595 }, { "epoch": 2.4381984422621064, "grad_norm": 1.2265625, "learning_rate": 2.0558803962482375e-06, "loss": 2.1, "step": 3600 }, { "epoch": 2.4415848289874704, "grad_norm": 1.2421875, "learning_rate": 2.0320006635761435e-06, "loss": 2.0839, "step": 3605 }, { "epoch": 2.4449712157128345, "grad_norm": 1.1875, "learning_rate": 2.008244734092444e-06, "loss": 2.0761, "step": 3610 }, { "epoch": 2.4483576024381986, "grad_norm": 1.203125, "learning_rate": 1.984612976906075e-06, "loss": 2.0696, "step": 3615 }, { "epoch": 2.4517439891635626, "grad_norm": 1.203125, "learning_rate": 1.9611057591966522e-06, "loss": 2.0966, "step": 3620 }, { "epoch": 2.4551303758889267, "grad_norm": 1.2109375, "learning_rate": 1.9377234462087447e-06, "loss": 2.0985, "step": 3625 }, { "epoch": 2.4585167626142903, "grad_norm": 1.203125, "learning_rate": 1.91446640124621e-06, "loss": 2.0828, "step": 3630 }, { "epoch": 2.461903149339655, "grad_norm": 1.2421875, "learning_rate": 1.8913349856665486e-06, "loss": 2.0791, "step": 3635 }, { "epoch": 2.4652895360650184, "grad_norm": 1.234375, "learning_rate": 1.8683295588752847e-06, "loss": 2.0866, "step": 3640 }, { "epoch": 2.4686759227903825, "grad_norm": 1.21875, "learning_rate": 1.845450478320393e-06, "loss": 2.1008, "step": 3645 }, { "epoch": 2.4720623095157466, "grad_norm": 1.1875, "learning_rate": 1.822698099486727e-06, "loss": 2.0725, "step": 3650 }, { "epoch": 2.4754486962411106, "grad_norm": 1.21875, "learning_rate": 1.8000727758905168e-06, "loss": 2.0797, "step": 3655 }, { "epoch": 2.4788350829664747, "grad_norm": 1.1796875, "learning_rate": 1.7775748590738584e-06, "loss": 2.0912, "step": 3660 }, { "epoch": 2.4822214696918388, "grad_norm": 1.203125, "learning_rate": 1.7552046985992555e-06, "loss": 2.0789, "step": 3665 }, { "epoch": 2.485607856417203, "grad_norm": 1.234375, "learning_rate": 1.7329626420442026e-06, "loss": 2.0996, "step": 3670 }, { "epoch": 2.488994243142567, "grad_norm": 1.2421875, "learning_rate": 1.7108490349957607e-06, "loss": 2.0803, "step": 3675 }, { "epoch": 2.492380629867931, "grad_norm": 1.2109375, "learning_rate": 1.6888642210452112e-06, "loss": 2.0879, "step": 3680 }, { "epoch": 2.495767016593295, "grad_norm": 1.203125, "learning_rate": 1.6670085417826975e-06, "loss": 2.0896, "step": 3685 }, { "epoch": 2.499153403318659, "grad_norm": 1.2265625, "learning_rate": 1.6452823367919314e-06, "loss": 2.076, "step": 3690 }, { "epoch": 2.502539790044023, "grad_norm": 1.203125, "learning_rate": 1.6236859436449082e-06, "loss": 2.1018, "step": 3695 }, { "epoch": 2.5059261767693872, "grad_norm": 1.2421875, "learning_rate": 1.6022196978966653e-06, "loss": 2.0709, "step": 3700 }, { "epoch": 2.509312563494751, "grad_norm": 1.1953125, "learning_rate": 1.5808839330800751e-06, "loss": 2.0698, "step": 3705 }, { "epoch": 2.5126989502201154, "grad_norm": 1.1875, "learning_rate": 1.5596789807006496e-06, "loss": 2.0919, "step": 3710 }, { "epoch": 2.516085336945479, "grad_norm": 1.2421875, "learning_rate": 1.5386051702313932e-06, "loss": 2.1077, "step": 3715 }, { "epoch": 2.519471723670843, "grad_norm": 1.2109375, "learning_rate": 1.5176628291076967e-06, "loss": 2.0902, "step": 3720 }, { "epoch": 2.522858110396207, "grad_norm": 1.2265625, "learning_rate": 1.4968522827222277e-06, "loss": 2.0845, "step": 3725 }, { "epoch": 2.526244497121571, "grad_norm": 1.2265625, "learning_rate": 1.4761738544198966e-06, "loss": 2.0843, "step": 3730 }, { "epoch": 2.5296308838469352, "grad_norm": 1.171875, "learning_rate": 1.4556278654928146e-06, "loss": 2.0749, "step": 3735 }, { "epoch": 2.5330172705722993, "grad_norm": 1.2265625, "learning_rate": 1.435214635175316e-06, "loss": 2.0794, "step": 3740 }, { "epoch": 2.5364036572976634, "grad_norm": 1.2578125, "learning_rate": 1.4149344806389863e-06, "loss": 2.0812, "step": 3745 }, { "epoch": 2.5397900440230274, "grad_norm": 1.1953125, "learning_rate": 1.394787716987741e-06, "loss": 2.0753, "step": 3750 }, { "epoch": 2.5431764307483915, "grad_norm": 1.2265625, "learning_rate": 1.3747746572529275e-06, "loss": 2.0816, "step": 3755 }, { "epoch": 2.5465628174737556, "grad_norm": 1.203125, "learning_rate": 1.3548956123884649e-06, "loss": 2.0848, "step": 3760 }, { "epoch": 2.5499492041991196, "grad_norm": 1.1875, "learning_rate": 1.3351508912660084e-06, "loss": 2.0692, "step": 3765 }, { "epoch": 2.5533355909244837, "grad_norm": 1.21875, "learning_rate": 1.3155408006701476e-06, "loss": 2.0906, "step": 3770 }, { "epoch": 2.5567219776498478, "grad_norm": 1.1953125, "learning_rate": 1.296065645293646e-06, "loss": 2.0738, "step": 3775 }, { "epoch": 2.5601083643752114, "grad_norm": 1.25, "learning_rate": 1.2767257277327083e-06, "loss": 2.0964, "step": 3780 }, { "epoch": 2.563494751100576, "grad_norm": 1.21875, "learning_rate": 1.2575213484822669e-06, "loss": 2.0804, "step": 3785 }, { "epoch": 2.5668811378259395, "grad_norm": 1.1796875, "learning_rate": 1.2384528059313306e-06, "loss": 2.0697, "step": 3790 }, { "epoch": 2.570267524551304, "grad_norm": 1.2109375, "learning_rate": 1.21952039635833e-06, "loss": 2.0894, "step": 3795 }, { "epoch": 2.5736539112766676, "grad_norm": 1.21875, "learning_rate": 1.200724413926525e-06, "loss": 2.0812, "step": 3800 }, { "epoch": 2.5770402980020317, "grad_norm": 1.234375, "learning_rate": 1.1820651506794368e-06, "loss": 2.1039, "step": 3805 }, { "epoch": 2.580426684727396, "grad_norm": 1.2109375, "learning_rate": 1.1635428965362982e-06, "loss": 2.0739, "step": 3810 }, { "epoch": 2.58381307145276, "grad_norm": 1.1875, "learning_rate": 1.1451579392875633e-06, "loss": 2.0805, "step": 3815 }, { "epoch": 2.587199458178124, "grad_norm": 1.203125, "learning_rate": 1.1269105645904232e-06, "loss": 2.0979, "step": 3820 }, { "epoch": 2.590585844903488, "grad_norm": 1.2265625, "learning_rate": 1.1088010559643758e-06, "loss": 2.086, "step": 3825 }, { "epoch": 2.593972231628852, "grad_norm": 1.21875, "learning_rate": 1.090829694786818e-06, "loss": 2.0914, "step": 3830 }, { "epoch": 2.597358618354216, "grad_norm": 1.2421875, "learning_rate": 1.0729967602886703e-06, "loss": 2.0993, "step": 3835 }, { "epoch": 2.60074500507958, "grad_norm": 1.2109375, "learning_rate": 1.0553025295500484e-06, "loss": 2.0861, "step": 3840 }, { "epoch": 2.6041313918049442, "grad_norm": 1.234375, "learning_rate": 1.037747277495945e-06, "loss": 2.0837, "step": 3845 }, { "epoch": 2.6075177785303083, "grad_norm": 1.2109375, "learning_rate": 1.0203312768919716e-06, "loss": 2.0849, "step": 3850 }, { "epoch": 2.610904165255672, "grad_norm": 1.21875, "learning_rate": 1.003054798340105e-06, "loss": 2.0779, "step": 3855 }, { "epoch": 2.6142905519810364, "grad_norm": 1.234375, "learning_rate": 9.859181102744963e-07, "loss": 2.0924, "step": 3860 }, { "epoch": 2.6176769387064, "grad_norm": 1.1953125, "learning_rate": 9.689214789572943e-07, "loss": 2.0715, "step": 3865 }, { "epoch": 2.6210633254317646, "grad_norm": 1.1953125, "learning_rate": 9.520651684745064e-07, "loss": 2.0824, "step": 3870 }, { "epoch": 2.624449712157128, "grad_norm": 1.2109375, "learning_rate": 9.353494407319052e-07, "loss": 2.0809, "step": 3875 }, { "epoch": 2.6278360988824923, "grad_norm": 1.203125, "learning_rate": 9.187745554509431e-07, "loss": 2.0817, "step": 3880 }, { "epoch": 2.6312224856078563, "grad_norm": 1.2109375, "learning_rate": 9.023407701647291e-07, "loss": 2.0802, "step": 3885 }, { "epoch": 2.6346088723332204, "grad_norm": 1.21875, "learning_rate": 8.860483402140263e-07, "loss": 2.0896, "step": 3890 }, { "epoch": 2.6379952590585845, "grad_norm": 1.2109375, "learning_rate": 8.698975187432779e-07, "loss": 2.0717, "step": 3895 }, { "epoch": 2.6413816457839485, "grad_norm": 1.1484375, "learning_rate": 8.538885566966837e-07, "loss": 2.0819, "step": 3900 }, { "epoch": 2.6447680325093126, "grad_norm": 1.1953125, "learning_rate": 8.380217028142912e-07, "loss": 2.0907, "step": 3905 }, { "epoch": 2.6481544192346766, "grad_norm": 1.234375, "learning_rate": 8.222972036281351e-07, "loss": 2.0789, "step": 3910 }, { "epoch": 2.6515408059600407, "grad_norm": 1.2109375, "learning_rate": 8.067153034584108e-07, "loss": 2.0913, "step": 3915 }, { "epoch": 2.654927192685405, "grad_norm": 1.171875, "learning_rate": 7.912762444096689e-07, "loss": 2.0671, "step": 3920 }, { "epoch": 2.658313579410769, "grad_norm": 1.1796875, "learning_rate": 7.759802663670657e-07, "loss": 2.0727, "step": 3925 }, { "epoch": 2.661699966136133, "grad_norm": 1.21875, "learning_rate": 7.608276069926224e-07, "loss": 2.0865, "step": 3930 }, { "epoch": 2.665086352861497, "grad_norm": 1.25, "learning_rate": 7.458185017215425e-07, "loss": 2.0759, "step": 3935 }, { "epoch": 2.6684727395868606, "grad_norm": 1.203125, "learning_rate": 7.309531837585526e-07, "loss": 2.103, "step": 3940 }, { "epoch": 2.671859126312225, "grad_norm": 1.1953125, "learning_rate": 7.162318840742732e-07, "loss": 2.1004, "step": 3945 }, { "epoch": 2.6752455130375887, "grad_norm": 1.171875, "learning_rate": 7.016548314016336e-07, "loss": 2.0904, "step": 3950 }, { "epoch": 2.678631899762953, "grad_norm": 1.25, "learning_rate": 6.87222252232318e-07, "loss": 2.1003, "step": 3955 }, { "epoch": 2.682018286488317, "grad_norm": 1.2109375, "learning_rate": 6.729343708132507e-07, "loss": 2.0847, "step": 3960 }, { "epoch": 2.685404673213681, "grad_norm": 1.203125, "learning_rate": 6.587914091431036e-07, "loss": 2.0695, "step": 3965 }, { "epoch": 2.688791059939045, "grad_norm": 1.21875, "learning_rate": 6.447935869688482e-07, "loss": 2.0567, "step": 3970 }, { "epoch": 2.692177446664409, "grad_norm": 1.21875, "learning_rate": 6.309411217823503e-07, "loss": 2.0951, "step": 3975 }, { "epoch": 2.695563833389773, "grad_norm": 1.1953125, "learning_rate": 6.172342288169786e-07, "loss": 2.0876, "step": 3980 }, { "epoch": 2.698950220115137, "grad_norm": 1.1875, "learning_rate": 6.036731210442715e-07, "loss": 2.0966, "step": 3985 }, { "epoch": 2.7023366068405013, "grad_norm": 1.1953125, "learning_rate": 5.902580091706167e-07, "loss": 2.1063, "step": 3990 }, { "epoch": 2.7057229935658653, "grad_norm": 1.21875, "learning_rate": 5.769891016339868e-07, "loss": 2.0848, "step": 3995 }, { "epoch": 2.7091093802912294, "grad_norm": 1.25, "learning_rate": 5.63866604600698e-07, "loss": 2.0966, "step": 4000 }, { "epoch": 2.7124957670165935, "grad_norm": 1.2578125, "learning_rate": 5.508907219622029e-07, "loss": 2.0774, "step": 4005 }, { "epoch": 2.7158821537419575, "grad_norm": 1.1875, "learning_rate": 5.380616553319273e-07, "loss": 2.0928, "step": 4010 }, { "epoch": 2.719268540467321, "grad_norm": 1.203125, "learning_rate": 5.253796040421366e-07, "loss": 2.0798, "step": 4015 }, { "epoch": 2.7226549271926856, "grad_norm": 1.234375, "learning_rate": 5.128447651408386e-07, "loss": 2.0824, "step": 4020 }, { "epoch": 2.7260413139180493, "grad_norm": 1.1875, "learning_rate": 5.004573333887175e-07, "loss": 2.0768, "step": 4025 }, { "epoch": 2.7294277006434133, "grad_norm": 1.171875, "learning_rate": 4.882175012561141e-07, "loss": 2.0848, "step": 4030 }, { "epoch": 2.7328140873687774, "grad_norm": 1.1953125, "learning_rate": 4.761254589200359e-07, "loss": 2.0625, "step": 4035 }, { "epoch": 2.7362004740941415, "grad_norm": 1.21875, "learning_rate": 4.641813942611917e-07, "loss": 2.1149, "step": 4040 }, { "epoch": 2.7395868608195055, "grad_norm": 1.1796875, "learning_rate": 4.523854928610849e-07, "loss": 2.0874, "step": 4045 }, { "epoch": 2.7429732475448696, "grad_norm": 1.2578125, "learning_rate": 4.407379379991261e-07, "loss": 2.0857, "step": 4050 }, { "epoch": 2.7463596342702337, "grad_norm": 1.2265625, "learning_rate": 4.292389106497785e-07, "loss": 2.0922, "step": 4055 }, { "epoch": 2.7497460209955977, "grad_norm": 1.1796875, "learning_rate": 4.178885894797569e-07, "loss": 2.0767, "step": 4060 }, { "epoch": 2.753132407720962, "grad_norm": 1.21875, "learning_rate": 4.066871508452441e-07, "loss": 2.0927, "step": 4065 }, { "epoch": 2.756518794446326, "grad_norm": 1.1796875, "learning_rate": 3.9563476878915374e-07, "loss": 2.0975, "step": 4070 }, { "epoch": 2.75990518117169, "grad_norm": 1.203125, "learning_rate": 3.847316150384239e-07, "loss": 2.0766, "step": 4075 }, { "epoch": 2.763291567897054, "grad_norm": 1.1953125, "learning_rate": 3.7397785900135317e-07, "loss": 2.0697, "step": 4080 }, { "epoch": 2.766677954622418, "grad_norm": 1.203125, "learning_rate": 3.633736677649624e-07, "loss": 2.0817, "step": 4085 }, { "epoch": 2.7700643413477817, "grad_norm": 1.2578125, "learning_rate": 3.529192060924036e-07, "loss": 2.0901, "step": 4090 }, { "epoch": 2.773450728073146, "grad_norm": 1.203125, "learning_rate": 3.426146364203997e-07, "loss": 2.0796, "step": 4095 }, { "epoch": 2.77683711479851, "grad_norm": 1.2109375, "learning_rate": 3.324601188567167e-07, "loss": 2.0835, "step": 4100 }, { "epoch": 2.780223501523874, "grad_norm": 1.234375, "learning_rate": 3.2245581117767875e-07, "loss": 2.1009, "step": 4105 }, { "epoch": 2.783609888249238, "grad_norm": 1.203125, "learning_rate": 3.126018688257182e-07, "loss": 2.0807, "step": 4110 }, { "epoch": 2.786996274974602, "grad_norm": 1.1875, "learning_rate": 3.028984449069561e-07, "loss": 2.0664, "step": 4115 }, { "epoch": 2.790382661699966, "grad_norm": 1.1953125, "learning_rate": 2.9334569018882874e-07, "loss": 2.0727, "step": 4120 }, { "epoch": 2.79376904842533, "grad_norm": 1.2265625, "learning_rate": 2.839437530977385e-07, "loss": 2.0744, "step": 4125 }, { "epoch": 2.797155435150694, "grad_norm": 1.2421875, "learning_rate": 2.7469277971675536e-07, "loss": 2.0825, "step": 4130 }, { "epoch": 2.8005418218760583, "grad_norm": 1.2265625, "learning_rate": 2.6559291378334016e-07, "loss": 2.0826, "step": 4135 }, { "epoch": 2.8039282086014223, "grad_norm": 1.2265625, "learning_rate": 2.566442966871152e-07, "loss": 2.0688, "step": 4140 }, { "epoch": 2.8073145953267864, "grad_norm": 1.1953125, "learning_rate": 2.4784706746766694e-07, "loss": 2.0766, "step": 4145 }, { "epoch": 2.8107009820521505, "grad_norm": 1.21875, "learning_rate": 2.3920136281238347e-07, "loss": 2.0842, "step": 4150 }, { "epoch": 2.8140873687775145, "grad_norm": 1.1875, "learning_rate": 2.3070731705433392e-07, "loss": 2.0715, "step": 4155 }, { "epoch": 2.8174737555028786, "grad_norm": 1.2109375, "learning_rate": 2.2236506217018007e-07, "loss": 2.0785, "step": 4160 }, { "epoch": 2.820860142228242, "grad_norm": 1.21875, "learning_rate": 2.1417472777812365e-07, "loss": 2.0964, "step": 4165 }, { "epoch": 2.8242465289536067, "grad_norm": 1.2421875, "learning_rate": 2.0613644113589571e-07, "loss": 2.096, "step": 4170 }, { "epoch": 2.8276329156789703, "grad_norm": 1.21875, "learning_rate": 1.982503271387759e-07, "loss": 2.0975, "step": 4175 }, { "epoch": 2.8310193024043344, "grad_norm": 1.203125, "learning_rate": 1.9051650831765745e-07, "loss": 2.0841, "step": 4180 }, { "epoch": 2.8344056891296985, "grad_norm": 1.2265625, "learning_rate": 1.829351048371364e-07, "loss": 2.0861, "step": 4185 }, { "epoch": 2.8377920758550625, "grad_norm": 1.234375, "learning_rate": 1.7550623449364645e-07, "loss": 2.098, "step": 4190 }, { "epoch": 2.8411784625804266, "grad_norm": 1.21875, "learning_rate": 1.6823001271363493e-07, "loss": 2.0681, "step": 4195 }, { "epoch": 2.8445648493057907, "grad_norm": 1.2109375, "learning_rate": 1.611065525517619e-07, "loss": 2.1106, "step": 4200 }, { "epoch": 2.8479512360311547, "grad_norm": 1.2265625, "learning_rate": 1.5413596468914384e-07, "loss": 2.0908, "step": 4205 }, { "epoch": 2.851337622756519, "grad_norm": 1.1875, "learning_rate": 1.4731835743164057e-07, "loss": 2.0829, "step": 4210 }, { "epoch": 2.854724009481883, "grad_norm": 1.2265625, "learning_rate": 1.4065383670816446e-07, "loss": 2.084, "step": 4215 }, { "epoch": 2.858110396207247, "grad_norm": 1.2578125, "learning_rate": 1.3414250606904046e-07, "loss": 2.0741, "step": 4220 }, { "epoch": 2.861496782932611, "grad_norm": 1.2109375, "learning_rate": 1.2778446668439415e-07, "loss": 2.0881, "step": 4225 }, { "epoch": 2.864883169657975, "grad_norm": 1.2578125, "learning_rate": 1.2157981734258084e-07, "loss": 2.0815, "step": 4230 }, { "epoch": 2.868269556383339, "grad_norm": 1.203125, "learning_rate": 1.1552865444865002e-07, "loss": 2.0903, "step": 4235 }, { "epoch": 2.8716559431087028, "grad_norm": 1.2109375, "learning_rate": 1.0963107202284773e-07, "loss": 2.1015, "step": 4240 }, { "epoch": 2.8750423298340673, "grad_norm": 1.1953125, "learning_rate": 1.0388716169915547e-07, "loss": 2.0921, "step": 4245 }, { "epoch": 2.878428716559431, "grad_norm": 1.2109375, "learning_rate": 9.829701272386804e-08, "loss": 2.0922, "step": 4250 }, { "epoch": 2.8818151032847954, "grad_norm": 1.2421875, "learning_rate": 9.286071195420576e-08, "loss": 2.0977, "step": 4255 }, { "epoch": 2.885201490010159, "grad_norm": 1.1953125, "learning_rate": 8.757834385696329e-08, "loss": 2.0848, "step": 4260 }, { "epoch": 2.888587876735523, "grad_norm": 1.2265625, "learning_rate": 8.244999050719738e-08, "loss": 2.0834, "step": 4265 }, { "epoch": 2.891974263460887, "grad_norm": 1.1640625, "learning_rate": 7.747573158695676e-08, "loss": 2.0504, "step": 4270 }, { "epoch": 2.895360650186251, "grad_norm": 1.1875, "learning_rate": 7.265564438403872e-08, "loss": 2.0784, "step": 4275 }, { "epoch": 2.8987470369116153, "grad_norm": 1.1875, "learning_rate": 6.798980379078779e-08, "loss": 2.0662, "step": 4280 }, { "epoch": 2.9021334236369793, "grad_norm": 1.25, "learning_rate": 6.347828230293563e-08, "loss": 2.088, "step": 4285 }, { "epoch": 2.9055198103623434, "grad_norm": 1.2265625, "learning_rate": 5.912115001847407e-08, "loss": 2.1124, "step": 4290 }, { "epoch": 2.9089061970877075, "grad_norm": 1.203125, "learning_rate": 5.491847463656497e-08, "loss": 2.0721, "step": 4295 }, { "epoch": 2.9122925838130715, "grad_norm": 1.203125, "learning_rate": 5.08703214564843e-08, "loss": 2.0933, "step": 4300 }, { "epoch": 2.9156789705384356, "grad_norm": 1.2734375, "learning_rate": 4.697675337661856e-08, "loss": 2.0805, "step": 4305 }, { "epoch": 2.9190653572637997, "grad_norm": 1.1953125, "learning_rate": 4.323783089347222e-08, "loss": 2.0628, "step": 4310 }, { "epoch": 2.9224517439891633, "grad_norm": 1.15625, "learning_rate": 3.965361210074514e-08, "loss": 2.0946, "step": 4315 }, { "epoch": 2.925838130714528, "grad_norm": 1.171875, "learning_rate": 3.6224152688412174e-08, "loss": 2.0809, "step": 4320 }, { "epoch": 2.9292245174398914, "grad_norm": 1.1796875, "learning_rate": 3.2949505941872736e-08, "loss": 2.083, "step": 4325 }, { "epoch": 2.932610904165256, "grad_norm": 1.2109375, "learning_rate": 2.9829722741113734e-08, "loss": 2.0766, "step": 4330 }, { "epoch": 2.9359972908906196, "grad_norm": 1.1875, "learning_rate": 2.6864851559923465e-08, "loss": 2.0876, "step": 4335 }, { "epoch": 2.9393836776159836, "grad_norm": 1.203125, "learning_rate": 2.4054938465135626e-08, "loss": 2.0948, "step": 4340 }, { "epoch": 2.9427700643413477, "grad_norm": 1.21875, "learning_rate": 2.140002711591538e-08, "loss": 2.0879, "step": 4345 }, { "epoch": 2.9461564510667118, "grad_norm": 1.2421875, "learning_rate": 1.8900158763081045e-08, "loss": 2.0865, "step": 4350 }, { "epoch": 2.949542837792076, "grad_norm": 1.21875, "learning_rate": 1.6555372248462376e-08, "loss": 2.0778, "step": 4355 }, { "epoch": 2.95292922451744, "grad_norm": 1.21875, "learning_rate": 1.4365704004295489e-08, "loss": 2.0722, "step": 4360 }, { "epoch": 2.956315611242804, "grad_norm": 1.1953125, "learning_rate": 1.2331188052663312e-08, "loss": 2.089, "step": 4365 }, { "epoch": 2.959701997968168, "grad_norm": 1.21875, "learning_rate": 1.0451856004957128e-08, "loss": 2.0723, "step": 4370 }, { "epoch": 2.963088384693532, "grad_norm": 1.203125, "learning_rate": 8.727737061393626e-09, "loss": 2.0725, "step": 4375 }, { "epoch": 2.966474771418896, "grad_norm": 1.1953125, "learning_rate": 7.15885801055638e-09, "loss": 2.0768, "step": 4380 }, { "epoch": 2.96986115814426, "grad_norm": 1.203125, "learning_rate": 5.745243228982844e-09, "loss": 2.0641, "step": 4385 }, { "epoch": 2.973247544869624, "grad_norm": 1.2109375, "learning_rate": 4.4869146807824396e-09, "loss": 2.0904, "step": 4390 }, { "epoch": 2.9766339315949883, "grad_norm": 1.203125, "learning_rate": 3.3838919172990426e-09, "loss": 2.1051, "step": 4395 }, { "epoch": 2.980020318320352, "grad_norm": 1.1875, "learning_rate": 2.4361920768045668e-09, "loss": 2.0645, "step": 4400 }, { "epoch": 2.9834067050457165, "grad_norm": 1.2265625, "learning_rate": 1.6438298842302858e-09, "loss": 2.0866, "step": 4405 }, { "epoch": 2.98679309177108, "grad_norm": 1.2265625, "learning_rate": 1.0068176509447913e-09, "loss": 2.0859, "step": 4410 }, { "epoch": 2.990179478496444, "grad_norm": 1.1875, "learning_rate": 5.251652745585922e-10, "loss": 2.0636, "step": 4415 }, { "epoch": 2.9935658652218082, "grad_norm": 1.1953125, "learning_rate": 1.9888023876757368e-10, "loss": 2.0931, "step": 4420 }, { "epoch": 2.9969522519471723, "grad_norm": 1.1953125, "learning_rate": 2.796761324419528e-11, "loss": 2.0766, "step": 4425 }, { "epoch": 2.998984083982391, "eval_loss": 2.1175546646118164, "eval_runtime": 85.4567, "eval_samples_per_second": 15.271, "eval_steps_per_second": 1.919, "step": 4428 }, { "epoch": 2.998984083982391, "step": 4428, "total_flos": 9.004989021620797e+17, "train_loss": 2.11554208852842, "train_runtime": 19880.5774, "train_samples_per_second": 3.564, "train_steps_per_second": 0.223 } ], "logging_steps": 5, "max_steps": 4428, "num_input_tokens_seen": 0, "num_train_epochs": 3, "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": 9.004989021620797e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }