| { |
| "best_metric": 3.3056321144104004, |
| "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_8397/checkpoint-90000", |
| "epoch": 10.0, |
| "eval_steps": 1000, |
| "global_step": 92910, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.005381552039608223, |
| "grad_norm": 2.765322685241699, |
| "learning_rate": 0.0003, |
| "loss": 8.4843, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.010763104079216447, |
| "grad_norm": 4.5047125816345215, |
| "learning_rate": 0.0006, |
| "loss": 6.9314, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.01614465611882467, |
| "grad_norm": 1.1575511693954468, |
| "learning_rate": 0.0005996767589699385, |
| "loss": 6.5187, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.021526208158432893, |
| "grad_norm": 2.314443826675415, |
| "learning_rate": 0.0005993535179398771, |
| "loss": 6.2703, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.026907760198041114, |
| "grad_norm": 0.848477303981781, |
| "learning_rate": 0.0005990302769098158, |
| "loss": 6.1093, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.03228931223764934, |
| "grad_norm": 1.2320305109024048, |
| "learning_rate": 0.0005987070358797543, |
| "loss": 5.9673, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.03767086427725756, |
| "grad_norm": 1.2428854703903198, |
| "learning_rate": 0.0005983837948496929, |
| "loss": 5.8876, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.04305241631686579, |
| "grad_norm": 1.5372645854949951, |
| "learning_rate": 0.0005980605538196314, |
| "loss": 5.7962, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.048433968356474004, |
| "grad_norm": 1.0138875246047974, |
| "learning_rate": 0.0005977373127895701, |
| "loss": 5.7055, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.05381552039608223, |
| "grad_norm": 1.1828784942626953, |
| "learning_rate": 0.0005974140717595086, |
| "loss": 5.6615, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.05919707243569045, |
| "grad_norm": 1.8574672937393188, |
| "learning_rate": 0.0005970908307294472, |
| "loss": 5.564, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.06457862447529868, |
| "grad_norm": 1.2299022674560547, |
| "learning_rate": 0.0005967675896993858, |
| "loss": 5.4985, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.0699601765149069, |
| "grad_norm": 1.4707765579223633, |
| "learning_rate": 0.0005964443486693243, |
| "loss": 5.4515, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.07534172855451512, |
| "grad_norm": 1.6264588832855225, |
| "learning_rate": 0.000596121107639263, |
| "loss": 5.3654, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.08072328059412334, |
| "grad_norm": 1.0290075540542603, |
| "learning_rate": 0.0005957978666092015, |
| "loss": 5.3191, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.08610483263373157, |
| "grad_norm": 1.0803887844085693, |
| "learning_rate": 0.0005954746255791401, |
| "loss": 5.2614, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.09148638467333979, |
| "grad_norm": 1.364003300666809, |
| "learning_rate": 0.0005951513845490787, |
| "loss": 5.2103, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.09686793671294801, |
| "grad_norm": 1.0939010381698608, |
| "learning_rate": 0.0005948281435190174, |
| "loss": 5.1724, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.10224948875255624, |
| "grad_norm": 1.1524903774261475, |
| "learning_rate": 0.0005945049024889559, |
| "loss": 5.1273, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.10763104079216446, |
| "grad_norm": 1.0799986124038696, |
| "learning_rate": 0.0005941816614588944, |
| "loss": 5.0654, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.10763104079216446, |
| "eval_accuracy": 0.22742907383952468, |
| "eval_loss": 5.017577648162842, |
| "eval_runtime": 181.071, |
| "eval_samples_per_second": 99.469, |
| "eval_steps_per_second": 6.219, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.11301259283177269, |
| "grad_norm": 1.048494577407837, |
| "learning_rate": 0.000593858420428833, |
| "loss": 5.0541, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.1183941448713809, |
| "grad_norm": 1.2681626081466675, |
| "learning_rate": 0.0005935351793987716, |
| "loss": 5.0134, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.12377569691098914, |
| "grad_norm": 1.3078149557113647, |
| "learning_rate": 0.0005932119383687103, |
| "loss": 5.0027, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.12915724895059735, |
| "grad_norm": 0.8166284561157227, |
| "learning_rate": 0.0005928886973386488, |
| "loss": 4.9447, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.13453880099020557, |
| "grad_norm": 1.0351450443267822, |
| "learning_rate": 0.0005925654563085874, |
| "loss": 4.93, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.1399203530298138, |
| "grad_norm": 1.2266950607299805, |
| "learning_rate": 0.000592242215278526, |
| "loss": 4.8866, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.14530190506942203, |
| "grad_norm": 1.2586219310760498, |
| "learning_rate": 0.0005919189742484645, |
| "loss": 4.8748, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.15068345710903025, |
| "grad_norm": 1.322376012802124, |
| "learning_rate": 0.0005915957332184032, |
| "loss": 4.8629, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.15606500914863847, |
| "grad_norm": 0.9112418293952942, |
| "learning_rate": 0.0005912724921883417, |
| "loss": 4.8164, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.16144656118824668, |
| "grad_norm": 1.3490499258041382, |
| "learning_rate": 0.0005909492511582803, |
| "loss": 4.8008, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.1668281132278549, |
| "grad_norm": 1.0545638799667358, |
| "learning_rate": 0.0005906260101282189, |
| "loss": 4.7662, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.17220966526746315, |
| "grad_norm": 0.9380106925964355, |
| "learning_rate": 0.0005903027690981575, |
| "loss": 4.7539, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.17759121730707136, |
| "grad_norm": 0.9979204535484314, |
| "learning_rate": 0.000589979528068096, |
| "loss": 4.7435, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.18297276934667958, |
| "grad_norm": 0.7762619256973267, |
| "learning_rate": 0.0005896562870380347, |
| "loss": 4.7269, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.1883543213862878, |
| "grad_norm": 1.121840000152588, |
| "learning_rate": 0.0005893330460079732, |
| "loss": 4.6665, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.19373587342589602, |
| "grad_norm": 0.9376957416534424, |
| "learning_rate": 0.0005890098049779118, |
| "loss": 4.6738, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.19911742546550426, |
| "grad_norm": 1.1272518634796143, |
| "learning_rate": 0.0005886865639478504, |
| "loss": 4.6606, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.20449897750511248, |
| "grad_norm": 0.9599424600601196, |
| "learning_rate": 0.0005883633229177889, |
| "loss": 4.6454, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.2098805295447207, |
| "grad_norm": 1.1271470785140991, |
| "learning_rate": 0.0005880400818877276, |
| "loss": 4.5919, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.2152620815843289, |
| "grad_norm": 1.019126534461975, |
| "learning_rate": 0.0005877168408576662, |
| "loss": 4.5787, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.2152620815843289, |
| "eval_accuracy": 0.2694480791853554, |
| "eval_loss": 4.517198085784912, |
| "eval_runtime": 179.8348, |
| "eval_samples_per_second": 100.153, |
| "eval_steps_per_second": 6.261, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.22064363362393713, |
| "grad_norm": 0.9893919825553894, |
| "learning_rate": 0.0005873935998276048, |
| "loss": 4.5569, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.22602518566354537, |
| "grad_norm": 0.9964181184768677, |
| "learning_rate": 0.0005870703587975433, |
| "loss": 4.5256, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.2314067377031536, |
| "grad_norm": 0.8266022801399231, |
| "learning_rate": 0.0005867471177674818, |
| "loss": 4.5383, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.2367882897427618, |
| "grad_norm": 1.0192140340805054, |
| "learning_rate": 0.0005864238767374205, |
| "loss": 4.4971, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.24216984178237003, |
| "grad_norm": 1.1359943151474, |
| "learning_rate": 0.0005861006357073591, |
| "loss": 4.4988, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.24755139382197827, |
| "grad_norm": 0.9294039011001587, |
| "learning_rate": 0.0005857773946772977, |
| "loss": 4.4732, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.2529329458615865, |
| "grad_norm": 0.8425009846687317, |
| "learning_rate": 0.0005854541536472362, |
| "loss": 4.4396, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.2583144979011947, |
| "grad_norm": 0.6667693853378296, |
| "learning_rate": 0.0005851309126171749, |
| "loss": 4.4376, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.2636960499408029, |
| "grad_norm": 0.9736728668212891, |
| "learning_rate": 0.0005848076715871134, |
| "loss": 4.4274, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.26907760198041114, |
| "grad_norm": 0.9293293356895447, |
| "learning_rate": 0.000584484430557052, |
| "loss": 4.4234, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.27445915402001936, |
| "grad_norm": 0.9411072134971619, |
| "learning_rate": 0.0005841611895269906, |
| "loss": 4.4028, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.2798407060596276, |
| "grad_norm": 0.9999932646751404, |
| "learning_rate": 0.0005838379484969291, |
| "loss": 4.4162, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.2852222580992358, |
| "grad_norm": 0.8416957855224609, |
| "learning_rate": 0.0005835147074668678, |
| "loss": 4.3881, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.29060381013884407, |
| "grad_norm": 0.708859384059906, |
| "learning_rate": 0.0005831914664368063, |
| "loss": 4.3881, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.2959853621784523, |
| "grad_norm": 0.7405576109886169, |
| "learning_rate": 0.0005828682254067449, |
| "loss": 4.3532, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.3013669142180605, |
| "grad_norm": 0.8557761907577515, |
| "learning_rate": 0.0005825449843766835, |
| "loss": 4.3477, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.3067484662576687, |
| "grad_norm": 0.8067885637283325, |
| "learning_rate": 0.0005822217433466221, |
| "loss": 4.3309, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.31213001829727693, |
| "grad_norm": 1.0514090061187744, |
| "learning_rate": 0.0005818985023165607, |
| "loss": 4.3391, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.31751157033688515, |
| "grad_norm": 0.6951010227203369, |
| "learning_rate": 0.0005815752612864992, |
| "loss": 4.2988, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.32289312237649337, |
| "grad_norm": 0.7815271615982056, |
| "learning_rate": 0.0005812520202564378, |
| "loss": 4.3112, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.32289312237649337, |
| "eval_accuracy": 0.29842786877034594, |
| "eval_loss": 4.237146854400635, |
| "eval_runtime": 179.5413, |
| "eval_samples_per_second": 100.317, |
| "eval_steps_per_second": 6.272, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.3282746744161016, |
| "grad_norm": 0.7033036351203918, |
| "learning_rate": 0.0005809287792263764, |
| "loss": 4.3185, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.3336562264557098, |
| "grad_norm": 0.8528683185577393, |
| "learning_rate": 0.0005806055381963151, |
| "loss": 4.2874, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.3390377784953181, |
| "grad_norm": 0.7960010766983032, |
| "learning_rate": 0.0005802822971662536, |
| "loss": 4.2738, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.3444193305349263, |
| "grad_norm": 0.638249397277832, |
| "learning_rate": 0.0005799590561361922, |
| "loss": 4.2911, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.3498008825745345, |
| "grad_norm": 0.8658631443977356, |
| "learning_rate": 0.0005796358151061307, |
| "loss": 4.2494, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.35518243461414273, |
| "grad_norm": 0.7108155488967896, |
| "learning_rate": 0.0005793125740760694, |
| "loss": 4.27, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.36056398665375095, |
| "grad_norm": 0.7243547439575195, |
| "learning_rate": 0.0005789893330460079, |
| "loss": 4.2532, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.36594553869335916, |
| "grad_norm": 0.7133619785308838, |
| "learning_rate": 0.0005786660920159465, |
| "loss": 4.2336, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.3713270907329674, |
| "grad_norm": 0.7307054400444031, |
| "learning_rate": 0.0005783428509858851, |
| "loss": 4.2161, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.3767086427725756, |
| "grad_norm": 0.783027708530426, |
| "learning_rate": 0.0005780196099558237, |
| "loss": 4.2252, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.3820901948121838, |
| "grad_norm": 0.774836540222168, |
| "learning_rate": 0.0005776963689257623, |
| "loss": 4.2057, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.38747174685179203, |
| "grad_norm": 0.7645639777183533, |
| "learning_rate": 0.0005773731278957008, |
| "loss": 4.2043, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.3928532988914003, |
| "grad_norm": 0.6912294626235962, |
| "learning_rate": 0.0005770498868656394, |
| "loss": 4.176, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.3982348509310085, |
| "grad_norm": 1.0210829973220825, |
| "learning_rate": 0.000576726645835578, |
| "loss": 4.1892, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.40361640297061674, |
| "grad_norm": 0.7821992635726929, |
| "learning_rate": 0.0005764034048055167, |
| "loss": 4.1735, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.40899795501022496, |
| "grad_norm": 0.8775643110275269, |
| "learning_rate": 0.0005760801637754552, |
| "loss": 4.1808, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.4143795070498332, |
| "grad_norm": 0.787700891494751, |
| "learning_rate": 0.0005757569227453937, |
| "loss": 4.1648, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.4197610590894414, |
| "grad_norm": 0.8575156927108765, |
| "learning_rate": 0.0005754336817153324, |
| "loss": 4.1757, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.4251426111290496, |
| "grad_norm": 0.7350611090660095, |
| "learning_rate": 0.0005751104406852709, |
| "loss": 4.1602, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.4305241631686578, |
| "grad_norm": 0.6890688538551331, |
| "learning_rate": 0.0005747871996552096, |
| "loss": 4.164, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.4305241631686578, |
| "eval_accuracy": 0.31287316155239336, |
| "eval_loss": 4.085758209228516, |
| "eval_runtime": 179.588, |
| "eval_samples_per_second": 100.291, |
| "eval_steps_per_second": 6.27, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.43590571520826604, |
| "grad_norm": 0.7492932677268982, |
| "learning_rate": 0.0005744639586251481, |
| "loss": 4.1546, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.44128726724787426, |
| "grad_norm": 0.6462448239326477, |
| "learning_rate": 0.0005741407175950867, |
| "loss": 4.159, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.44666881928748253, |
| "grad_norm": 0.7062949538230896, |
| "learning_rate": 0.0005738174765650253, |
| "loss": 4.1492, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.45205037132709075, |
| "grad_norm": 0.8296352624893188, |
| "learning_rate": 0.0005734942355349638, |
| "loss": 4.1268, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.45743192336669897, |
| "grad_norm": 0.7677702307701111, |
| "learning_rate": 0.0005731709945049025, |
| "loss": 4.1527, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.4628134754063072, |
| "grad_norm": 0.5997741222381592, |
| "learning_rate": 0.000572847753474841, |
| "loss": 4.1248, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.4681950274459154, |
| "grad_norm": 0.7054913640022278, |
| "learning_rate": 0.0005725245124447796, |
| "loss": 4.1242, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.4735765794855236, |
| "grad_norm": 0.6141555309295654, |
| "learning_rate": 0.0005722012714147182, |
| "loss": 4.102, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.47895813152513184, |
| "grad_norm": 0.9202010035514832, |
| "learning_rate": 0.0005718780303846568, |
| "loss": 4.1078, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.48433968356474005, |
| "grad_norm": 0.7407594919204712, |
| "learning_rate": 0.0005715547893545953, |
| "loss": 4.1116, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.48972123560434827, |
| "grad_norm": 0.7345452308654785, |
| "learning_rate": 0.000571231548324534, |
| "loss": 4.108, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.49510278764395654, |
| "grad_norm": 0.6181378364562988, |
| "learning_rate": 0.0005709083072944725, |
| "loss": 4.0897, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.5004843396835648, |
| "grad_norm": 0.5611840486526489, |
| "learning_rate": 0.0005705850662644111, |
| "loss": 4.0879, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.505865891723173, |
| "grad_norm": 0.7019966840744019, |
| "learning_rate": 0.0005702618252343497, |
| "loss": 4.1121, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.5112474437627812, |
| "grad_norm": 0.5246710777282715, |
| "learning_rate": 0.0005699385842042882, |
| "loss": 4.0842, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.5166289958023894, |
| "grad_norm": 0.7353329062461853, |
| "learning_rate": 0.0005696153431742269, |
| "loss": 4.0774, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.5220105478419976, |
| "grad_norm": 0.674192488193512, |
| "learning_rate": 0.0005692921021441655, |
| "loss": 4.0947, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.5273920998816058, |
| "grad_norm": 0.7477933168411255, |
| "learning_rate": 0.0005689688611141041, |
| "loss": 4.0597, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.5327736519212141, |
| "grad_norm": 0.8160755634307861, |
| "learning_rate": 0.0005686456200840426, |
| "loss": 4.0517, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.5381552039608223, |
| "grad_norm": 0.5745596885681152, |
| "learning_rate": 0.0005683223790539811, |
| "loss": 4.0509, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5381552039608223, |
| "eval_accuracy": 0.32170837977791567, |
| "eval_loss": 3.98921537399292, |
| "eval_runtime": 179.5969, |
| "eval_samples_per_second": 100.286, |
| "eval_steps_per_second": 6.27, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5435367560004305, |
| "grad_norm": 0.701214611530304, |
| "learning_rate": 0.0005679991380239198, |
| "loss": 4.0438, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.5489183080400387, |
| "grad_norm": 0.7315069437026978, |
| "learning_rate": 0.0005676758969938584, |
| "loss": 4.0379, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.5542998600796469, |
| "grad_norm": 0.7453345656394958, |
| "learning_rate": 0.000567352655963797, |
| "loss": 4.0327, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.5596814121192552, |
| "grad_norm": 0.7549764513969421, |
| "learning_rate": 0.0005670294149337355, |
| "loss": 4.0571, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.5650629641588634, |
| "grad_norm": 0.6196472644805908, |
| "learning_rate": 0.0005667061739036742, |
| "loss": 4.0388, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.5704445161984716, |
| "grad_norm": 0.6065132021903992, |
| "learning_rate": 0.0005663829328736127, |
| "loss": 4.023, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.5758260682380799, |
| "grad_norm": 0.7559676170349121, |
| "learning_rate": 0.0005660596918435512, |
| "loss": 4.0224, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.5812076202776881, |
| "grad_norm": 0.5514401793479919, |
| "learning_rate": 0.0005657364508134899, |
| "loss": 4.0343, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.5865891723172963, |
| "grad_norm": 0.6436402201652527, |
| "learning_rate": 0.0005654132097834284, |
| "loss": 4.0141, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.5919707243569046, |
| "grad_norm": 0.6794455647468567, |
| "learning_rate": 0.0005650899687533671, |
| "loss": 4.0005, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.5973522763965128, |
| "grad_norm": 0.6911609768867493, |
| "learning_rate": 0.0005647667277233056, |
| "loss": 4.0005, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.602733828436121, |
| "grad_norm": 0.6370590329170227, |
| "learning_rate": 0.0005644434866932442, |
| "loss": 4.0197, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.6081153804757292, |
| "grad_norm": 0.5897583365440369, |
| "learning_rate": 0.0005641202456631828, |
| "loss": 4.0107, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.6134969325153374, |
| "grad_norm": 0.6555122137069702, |
| "learning_rate": 0.0005637970046331214, |
| "loss": 4.0186, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.6188784845549457, |
| "grad_norm": 0.7258750200271606, |
| "learning_rate": 0.00056347376360306, |
| "loss": 4.0089, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.6242600365945539, |
| "grad_norm": 0.7162750363349915, |
| "learning_rate": 0.0005631505225729985, |
| "loss": 4.0215, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.6296415886341621, |
| "grad_norm": 0.5979080200195312, |
| "learning_rate": 0.0005628272815429371, |
| "loss": 3.991, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.6350231406737703, |
| "grad_norm": 0.5397379398345947, |
| "learning_rate": 0.0005625040405128757, |
| "loss": 4.0112, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.6404046927133785, |
| "grad_norm": 0.8027802109718323, |
| "learning_rate": 0.0005621807994828143, |
| "loss": 3.9821, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.6457862447529867, |
| "grad_norm": 0.7149248719215393, |
| "learning_rate": 0.0005618575584527529, |
| "loss": 3.9733, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.6457862447529867, |
| "eval_accuracy": 0.3285034227289455, |
| "eval_loss": 3.9209306240081787, |
| "eval_runtime": 179.6293, |
| "eval_samples_per_second": 100.268, |
| "eval_steps_per_second": 6.268, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.651167796792595, |
| "grad_norm": 0.5952973365783691, |
| "learning_rate": 0.0005615343174226915, |
| "loss": 3.9764, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.6565493488322032, |
| "grad_norm": 0.5565899014472961, |
| "learning_rate": 0.00056121107639263, |
| "loss": 3.9602, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.6619309008718114, |
| "grad_norm": 0.48603519797325134, |
| "learning_rate": 0.0005608878353625687, |
| "loss": 3.9707, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.6673124529114196, |
| "grad_norm": 0.6118786334991455, |
| "learning_rate": 0.0005605645943325072, |
| "loss": 3.9832, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.6726940049510278, |
| "grad_norm": 0.6439682245254517, |
| "learning_rate": 0.0005602413533024458, |
| "loss": 3.9652, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.6780755569906362, |
| "grad_norm": 0.5692785978317261, |
| "learning_rate": 0.0005599181122723844, |
| "loss": 3.9682, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.6834571090302444, |
| "grad_norm": 0.5279796719551086, |
| "learning_rate": 0.000559594871242323, |
| "loss": 3.9575, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.6888386610698526, |
| "grad_norm": 0.5329293012619019, |
| "learning_rate": 0.0005592716302122616, |
| "loss": 3.9661, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.6942202131094608, |
| "grad_norm": 0.59842848777771, |
| "learning_rate": 0.0005589483891822001, |
| "loss": 3.945, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.699601765149069, |
| "grad_norm": 0.6459380984306335, |
| "learning_rate": 0.0005586251481521387, |
| "loss": 3.9708, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.7049833171886772, |
| "grad_norm": 0.6028556227684021, |
| "learning_rate": 0.0005583083719426786, |
| "loss": 3.9382, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.7103648692282855, |
| "grad_norm": 0.6786797046661377, |
| "learning_rate": 0.0005579851309126171, |
| "loss": 3.9489, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.7157464212678937, |
| "grad_norm": 0.5982097387313843, |
| "learning_rate": 0.0005576618898825558, |
| "loss": 3.9588, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.7211279733075019, |
| "grad_norm": 0.668205738067627, |
| "learning_rate": 0.0005573386488524943, |
| "loss": 3.951, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.7265095253471101, |
| "grad_norm": 0.5475454330444336, |
| "learning_rate": 0.0005570154078224328, |
| "loss": 3.9315, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.7318910773867183, |
| "grad_norm": 0.5718148946762085, |
| "learning_rate": 0.0005566921667923715, |
| "loss": 3.9428, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.7372726294263265, |
| "grad_norm": 0.604880690574646, |
| "learning_rate": 0.00055636892576231, |
| "loss": 3.9552, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.7426541814659348, |
| "grad_norm": 0.5904679298400879, |
| "learning_rate": 0.0005560456847322487, |
| "loss": 3.9205, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.748035733505543, |
| "grad_norm": 0.6117699146270752, |
| "learning_rate": 0.0005557224437021872, |
| "loss": 3.943, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.7534172855451512, |
| "grad_norm": 0.5624613761901855, |
| "learning_rate": 0.0005553992026721258, |
| "loss": 3.9243, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7534172855451512, |
| "eval_accuracy": 0.33315843840158127, |
| "eval_loss": 3.8613877296447754, |
| "eval_runtime": 179.8814, |
| "eval_samples_per_second": 100.127, |
| "eval_steps_per_second": 6.26, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7587988375847594, |
| "grad_norm": 0.6362348794937134, |
| "learning_rate": 0.0005550759616420644, |
| "loss": 3.9187, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.7641803896243676, |
| "grad_norm": 0.6065320372581482, |
| "learning_rate": 0.000554752720612003, |
| "loss": 3.9133, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.7695619416639758, |
| "grad_norm": 0.6070377826690674, |
| "learning_rate": 0.0005544294795819415, |
| "loss": 3.9235, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.7749434937035841, |
| "grad_norm": 0.6300147771835327, |
| "learning_rate": 0.0005541062385518801, |
| "loss": 3.9364, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.7803250457431924, |
| "grad_norm": 0.6292157769203186, |
| "learning_rate": 0.0005537829975218188, |
| "loss": 3.9266, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.7857065977828006, |
| "grad_norm": 0.6281834840774536, |
| "learning_rate": 0.0005534597564917573, |
| "loss": 3.9215, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.7910881498224088, |
| "grad_norm": 0.6987627148628235, |
| "learning_rate": 0.0005531365154616959, |
| "loss": 3.932, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.796469701862017, |
| "grad_norm": 0.5451225638389587, |
| "learning_rate": 0.0005528132744316344, |
| "loss": 3.91, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.8018512539016253, |
| "grad_norm": 0.610156238079071, |
| "learning_rate": 0.0005524900334015731, |
| "loss": 3.9156, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.8072328059412335, |
| "grad_norm": 0.5881654620170593, |
| "learning_rate": 0.0005521667923715117, |
| "loss": 3.8981, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.8126143579808417, |
| "grad_norm": 0.6110484004020691, |
| "learning_rate": 0.0005518435513414502, |
| "loss": 3.9237, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.8179959100204499, |
| "grad_norm": 0.6429305076599121, |
| "learning_rate": 0.0005515203103113888, |
| "loss": 3.8959, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.8233774620600581, |
| "grad_norm": 0.6395870447158813, |
| "learning_rate": 0.0005511970692813274, |
| "loss": 3.9084, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.8287590140996663, |
| "grad_norm": 0.6113373041152954, |
| "learning_rate": 0.000550873828251266, |
| "loss": 3.9029, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.8341405661392746, |
| "grad_norm": 0.641963005065918, |
| "learning_rate": 0.0005505505872212045, |
| "loss": 3.883, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.8395221181788828, |
| "grad_norm": 0.6410965323448181, |
| "learning_rate": 0.0005502273461911432, |
| "loss": 3.8844, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.844903670218491, |
| "grad_norm": 0.604951024055481, |
| "learning_rate": 0.0005499041051610817, |
| "loss": 3.8943, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.8502852222580992, |
| "grad_norm": 0.6073367595672607, |
| "learning_rate": 0.0005495808641310204, |
| "loss": 3.9, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.8556667742977074, |
| "grad_norm": 0.5822165012359619, |
| "learning_rate": 0.0005492576231009589, |
| "loss": 3.898, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.8610483263373157, |
| "grad_norm": 0.618316650390625, |
| "learning_rate": 0.0005489343820708974, |
| "loss": 3.8774, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8610483263373157, |
| "eval_accuracy": 0.3376425430816849, |
| "eval_loss": 3.820213556289673, |
| "eval_runtime": 179.9142, |
| "eval_samples_per_second": 100.109, |
| "eval_steps_per_second": 6.259, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8664298783769239, |
| "grad_norm": 0.5604105591773987, |
| "learning_rate": 0.0005486111410408361, |
| "loss": 3.8929, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.8718114304165321, |
| "grad_norm": 0.6018927097320557, |
| "learning_rate": 0.0005482879000107746, |
| "loss": 3.8857, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.8771929824561403, |
| "grad_norm": 0.5374700427055359, |
| "learning_rate": 0.0005479646589807133, |
| "loss": 3.8744, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.8825745344957485, |
| "grad_norm": 0.5941512584686279, |
| "learning_rate": 0.0005476414179506518, |
| "loss": 3.8599, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.8879560865353568, |
| "grad_norm": 0.5862547159194946, |
| "learning_rate": 0.0005473181769205904, |
| "loss": 3.8706, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.8933376385749651, |
| "grad_norm": 0.5117703080177307, |
| "learning_rate": 0.000546994935890529, |
| "loss": 3.8804, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.8987191906145733, |
| "grad_norm": 0.5411306023597717, |
| "learning_rate": 0.0005466716948604677, |
| "loss": 3.8732, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.9041007426541815, |
| "grad_norm": 0.624138355255127, |
| "learning_rate": 0.0005463484538304062, |
| "loss": 3.8787, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.9094822946937897, |
| "grad_norm": 0.5695147514343262, |
| "learning_rate": 0.0005460252128003447, |
| "loss": 3.8548, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.9148638467333979, |
| "grad_norm": 0.648750901222229, |
| "learning_rate": 0.0005457019717702833, |
| "loss": 3.8585, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.9202453987730062, |
| "grad_norm": 0.6337014436721802, |
| "learning_rate": 0.0005453787307402219, |
| "loss": 3.8518, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.9256269508126144, |
| "grad_norm": 0.5613633990287781, |
| "learning_rate": 0.0005450619545307617, |
| "loss": 3.8631, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.9310085028522226, |
| "grad_norm": 0.6017595529556274, |
| "learning_rate": 0.0005447387135007003, |
| "loss": 3.8429, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.9363900548918308, |
| "grad_norm": 0.6081689596176147, |
| "learning_rate": 0.0005444154724706389, |
| "loss": 3.854, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.941771606931439, |
| "grad_norm": 0.5572118163108826, |
| "learning_rate": 0.0005440922314405775, |
| "loss": 3.8723, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.9471531589710472, |
| "grad_norm": 0.5897894501686096, |
| "learning_rate": 0.0005437689904105161, |
| "loss": 3.8473, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.9525347110106555, |
| "grad_norm": 0.5687997341156006, |
| "learning_rate": 0.0005434457493804546, |
| "loss": 3.8436, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.9579162630502637, |
| "grad_norm": 0.624033510684967, |
| "learning_rate": 0.0005431225083503932, |
| "loss": 3.84, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.9632978150898719, |
| "grad_norm": 0.5875102877616882, |
| "learning_rate": 0.0005427992673203318, |
| "loss": 3.866, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.9686793671294801, |
| "grad_norm": 0.6378977298736572, |
| "learning_rate": 0.0005424760262902704, |
| "loss": 3.8444, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9686793671294801, |
| "eval_accuracy": 0.3414037800991588, |
| "eval_loss": 3.7785165309906006, |
| "eval_runtime": 179.7193, |
| "eval_samples_per_second": 100.217, |
| "eval_steps_per_second": 6.265, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9740609191690883, |
| "grad_norm": 0.6483079195022583, |
| "learning_rate": 0.000542152785260209, |
| "loss": 3.8263, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.9794424712086965, |
| "grad_norm": 0.575869083404541, |
| "learning_rate": 0.0005418295442301476, |
| "loss": 3.8347, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.9848240232483048, |
| "grad_norm": 0.6090970635414124, |
| "learning_rate": 0.0005415063032000861, |
| "loss": 3.8418, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.9902055752879131, |
| "grad_norm": 0.5478109121322632, |
| "learning_rate": 0.0005411830621700248, |
| "loss": 3.8362, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.9955871273275213, |
| "grad_norm": 0.5703595280647278, |
| "learning_rate": 0.0005408598211399633, |
| "loss": 3.8135, |
| "step": 9250 |
| }, |
| { |
| "epoch": 1.0009686793671295, |
| "grad_norm": 0.6738930940628052, |
| "learning_rate": 0.0005405365801099019, |
| "loss": 3.8282, |
| "step": 9300 |
| }, |
| { |
| "epoch": 1.0063502314067376, |
| "grad_norm": 0.5832762718200684, |
| "learning_rate": 0.0005402133390798405, |
| "loss": 3.7677, |
| "step": 9350 |
| }, |
| { |
| "epoch": 1.011731783446346, |
| "grad_norm": 0.6429535150527954, |
| "learning_rate": 0.000539890098049779, |
| "loss": 3.7784, |
| "step": 9400 |
| }, |
| { |
| "epoch": 1.017113335485954, |
| "grad_norm": 0.5857776999473572, |
| "learning_rate": 0.0005395668570197177, |
| "loss": 3.753, |
| "step": 9450 |
| }, |
| { |
| "epoch": 1.0224948875255624, |
| "grad_norm": 0.553745448589325, |
| "learning_rate": 0.0005392436159896562, |
| "loss": 3.7717, |
| "step": 9500 |
| }, |
| { |
| "epoch": 1.0278764395651705, |
| "grad_norm": 0.592886209487915, |
| "learning_rate": 0.0005389203749595948, |
| "loss": 3.7701, |
| "step": 9550 |
| }, |
| { |
| "epoch": 1.0332579916047788, |
| "grad_norm": 0.5704110264778137, |
| "learning_rate": 0.0005385971339295334, |
| "loss": 3.7672, |
| "step": 9600 |
| }, |
| { |
| "epoch": 1.0386395436443872, |
| "grad_norm": 0.5219415426254272, |
| "learning_rate": 0.000538273892899472, |
| "loss": 3.7743, |
| "step": 9650 |
| }, |
| { |
| "epoch": 1.0440210956839953, |
| "grad_norm": 0.5501973628997803, |
| "learning_rate": 0.0005379506518694106, |
| "loss": 3.7665, |
| "step": 9700 |
| }, |
| { |
| "epoch": 1.0494026477236036, |
| "grad_norm": 0.5235220789909363, |
| "learning_rate": 0.0005376274108393491, |
| "loss": 3.772, |
| "step": 9750 |
| }, |
| { |
| "epoch": 1.0547841997632117, |
| "grad_norm": 0.6010895371437073, |
| "learning_rate": 0.0005373041698092877, |
| "loss": 3.7833, |
| "step": 9800 |
| }, |
| { |
| "epoch": 1.06016575180282, |
| "grad_norm": 0.5482034087181091, |
| "learning_rate": 0.0005369809287792263, |
| "loss": 3.7599, |
| "step": 9850 |
| }, |
| { |
| "epoch": 1.0655473038424281, |
| "grad_norm": 0.6096392869949341, |
| "learning_rate": 0.000536657687749165, |
| "loss": 3.7603, |
| "step": 9900 |
| }, |
| { |
| "epoch": 1.0709288558820365, |
| "grad_norm": 0.5938239693641663, |
| "learning_rate": 0.0005363344467191035, |
| "loss": 3.7484, |
| "step": 9950 |
| }, |
| { |
| "epoch": 1.0763104079216446, |
| "grad_norm": 0.6193915605545044, |
| "learning_rate": 0.000536011205689042, |
| "loss": 3.7673, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0763104079216446, |
| "eval_accuracy": 0.3440351357362499, |
| "eval_loss": 3.752418279647827, |
| "eval_runtime": 179.8923, |
| "eval_samples_per_second": 100.121, |
| "eval_steps_per_second": 6.259, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.081691959961253, |
| "grad_norm": 0.6629969477653503, |
| "learning_rate": 0.0005356879646589807, |
| "loss": 3.773, |
| "step": 10050 |
| }, |
| { |
| "epoch": 1.087073512000861, |
| "grad_norm": 0.6844997406005859, |
| "learning_rate": 0.0005353647236289192, |
| "loss": 3.7536, |
| "step": 10100 |
| }, |
| { |
| "epoch": 1.0924550640404693, |
| "grad_norm": 0.5414447784423828, |
| "learning_rate": 0.0005350414825988579, |
| "loss": 3.7628, |
| "step": 10150 |
| }, |
| { |
| "epoch": 1.0978366160800774, |
| "grad_norm": 0.6185393929481506, |
| "learning_rate": 0.0005347182415687964, |
| "loss": 3.7544, |
| "step": 10200 |
| }, |
| { |
| "epoch": 1.1032181681196858, |
| "grad_norm": 0.6166486740112305, |
| "learning_rate": 0.000534395000538735, |
| "loss": 3.7621, |
| "step": 10250 |
| }, |
| { |
| "epoch": 1.1085997201592939, |
| "grad_norm": 0.5102859735488892, |
| "learning_rate": 0.0005340717595086736, |
| "loss": 3.7568, |
| "step": 10300 |
| }, |
| { |
| "epoch": 1.1139812721989022, |
| "grad_norm": 0.5167948007583618, |
| "learning_rate": 0.0005337485184786122, |
| "loss": 3.762, |
| "step": 10350 |
| }, |
| { |
| "epoch": 1.1193628242385103, |
| "grad_norm": 0.5715023279190063, |
| "learning_rate": 0.0005334252774485507, |
| "loss": 3.7745, |
| "step": 10400 |
| }, |
| { |
| "epoch": 1.1247443762781186, |
| "grad_norm": 0.6422619223594666, |
| "learning_rate": 0.0005331020364184894, |
| "loss": 3.7554, |
| "step": 10450 |
| }, |
| { |
| "epoch": 1.1301259283177267, |
| "grad_norm": 0.5923563838005066, |
| "learning_rate": 0.0005327787953884279, |
| "loss": 3.7443, |
| "step": 10500 |
| }, |
| { |
| "epoch": 1.135507480357335, |
| "grad_norm": 0.6288694143295288, |
| "learning_rate": 0.0005324555543583665, |
| "loss": 3.7642, |
| "step": 10550 |
| }, |
| { |
| "epoch": 1.1408890323969434, |
| "grad_norm": 0.626955509185791, |
| "learning_rate": 0.0005321323133283051, |
| "loss": 3.7581, |
| "step": 10600 |
| }, |
| { |
| "epoch": 1.1462705844365515, |
| "grad_norm": 0.6043279767036438, |
| "learning_rate": 0.0005318090722982436, |
| "loss": 3.7467, |
| "step": 10650 |
| }, |
| { |
| "epoch": 1.1516521364761596, |
| "grad_norm": 0.5194068551063538, |
| "learning_rate": 0.0005314922960887834, |
| "loss": 3.747, |
| "step": 10700 |
| }, |
| { |
| "epoch": 1.157033688515768, |
| "grad_norm": 0.5931900143623352, |
| "learning_rate": 0.0005311690550587221, |
| "loss": 3.7423, |
| "step": 10750 |
| }, |
| { |
| "epoch": 1.1624152405553763, |
| "grad_norm": 0.6204814910888672, |
| "learning_rate": 0.0005308458140286607, |
| "loss": 3.7472, |
| "step": 10800 |
| }, |
| { |
| "epoch": 1.1677967925949844, |
| "grad_norm": 0.5869725346565247, |
| "learning_rate": 0.0005305225729985993, |
| "loss": 3.7455, |
| "step": 10850 |
| }, |
| { |
| "epoch": 1.1731783446345927, |
| "grad_norm": 0.616137683391571, |
| "learning_rate": 0.0005301993319685378, |
| "loss": 3.7662, |
| "step": 10900 |
| }, |
| { |
| "epoch": 1.1785598966742008, |
| "grad_norm": 0.5727664232254028, |
| "learning_rate": 0.0005298760909384765, |
| "loss": 3.7566, |
| "step": 10950 |
| }, |
| { |
| "epoch": 1.1839414487138091, |
| "grad_norm": 0.5144981145858765, |
| "learning_rate": 0.000529552849908415, |
| "loss": 3.753, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1839414487138091, |
| "eval_accuracy": 0.34677949037666805, |
| "eval_loss": 3.7220773696899414, |
| "eval_runtime": 179.7359, |
| "eval_samples_per_second": 100.208, |
| "eval_steps_per_second": 6.265, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1893230007534172, |
| "grad_norm": 0.5506604909896851, |
| "learning_rate": 0.0005292296088783535, |
| "loss": 3.7472, |
| "step": 11050 |
| }, |
| { |
| "epoch": 1.1947045527930256, |
| "grad_norm": 0.5824575424194336, |
| "learning_rate": 0.0005289063678482922, |
| "loss": 3.753, |
| "step": 11100 |
| }, |
| { |
| "epoch": 1.2000861048326337, |
| "grad_norm": 0.5471564531326294, |
| "learning_rate": 0.0005285831268182307, |
| "loss": 3.7553, |
| "step": 11150 |
| }, |
| { |
| "epoch": 1.205467656872242, |
| "grad_norm": 0.5185365080833435, |
| "learning_rate": 0.0005282598857881694, |
| "loss": 3.7437, |
| "step": 11200 |
| }, |
| { |
| "epoch": 1.21084920891185, |
| "grad_norm": 0.595317006111145, |
| "learning_rate": 0.0005279366447581079, |
| "loss": 3.7284, |
| "step": 11250 |
| }, |
| { |
| "epoch": 1.2162307609514584, |
| "grad_norm": 0.53959721326828, |
| "learning_rate": 0.0005276134037280465, |
| "loss": 3.7602, |
| "step": 11300 |
| }, |
| { |
| "epoch": 1.2216123129910665, |
| "grad_norm": 0.5472400784492493, |
| "learning_rate": 0.0005272901626979851, |
| "loss": 3.7386, |
| "step": 11350 |
| }, |
| { |
| "epoch": 1.2269938650306749, |
| "grad_norm": 0.5732322335243225, |
| "learning_rate": 0.0005269669216679236, |
| "loss": 3.7517, |
| "step": 11400 |
| }, |
| { |
| "epoch": 1.232375417070283, |
| "grad_norm": 0.5511186122894287, |
| "learning_rate": 0.0005266436806378623, |
| "loss": 3.739, |
| "step": 11450 |
| }, |
| { |
| "epoch": 1.2377569691098913, |
| "grad_norm": 0.5673669576644897, |
| "learning_rate": 0.0005263204396078008, |
| "loss": 3.732, |
| "step": 11500 |
| }, |
| { |
| "epoch": 1.2431385211494996, |
| "grad_norm": 0.5002085566520691, |
| "learning_rate": 0.0005259971985777394, |
| "loss": 3.7578, |
| "step": 11550 |
| }, |
| { |
| "epoch": 1.2485200731891077, |
| "grad_norm": 0.5736905336380005, |
| "learning_rate": 0.000525673957547678, |
| "loss": 3.7448, |
| "step": 11600 |
| }, |
| { |
| "epoch": 1.2539016252287158, |
| "grad_norm": 0.5424861907958984, |
| "learning_rate": 0.0005253507165176167, |
| "loss": 3.7202, |
| "step": 11650 |
| }, |
| { |
| "epoch": 1.2592831772683242, |
| "grad_norm": 0.542267382144928, |
| "learning_rate": 0.0005250274754875552, |
| "loss": 3.7241, |
| "step": 11700 |
| }, |
| { |
| "epoch": 1.2646647293079325, |
| "grad_norm": 0.5920253396034241, |
| "learning_rate": 0.0005247042344574938, |
| "loss": 3.7427, |
| "step": 11750 |
| }, |
| { |
| "epoch": 1.2700462813475406, |
| "grad_norm": 0.5953553915023804, |
| "learning_rate": 0.0005243809934274323, |
| "loss": 3.7356, |
| "step": 11800 |
| }, |
| { |
| "epoch": 1.275427833387149, |
| "grad_norm": 0.574566662311554, |
| "learning_rate": 0.0005240577523973709, |
| "loss": 3.7172, |
| "step": 11850 |
| }, |
| { |
| "epoch": 1.280809385426757, |
| "grad_norm": 0.6085739731788635, |
| "learning_rate": 0.0005237345113673095, |
| "loss": 3.7379, |
| "step": 11900 |
| }, |
| { |
| "epoch": 1.2861909374663654, |
| "grad_norm": 0.5501704812049866, |
| "learning_rate": 0.0005234112703372481, |
| "loss": 3.7723, |
| "step": 11950 |
| }, |
| { |
| "epoch": 1.2915724895059735, |
| "grad_norm": 0.5961480140686035, |
| "learning_rate": 0.0005230880293071867, |
| "loss": 3.7378, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2915724895059735, |
| "eval_accuracy": 0.34941084601375916, |
| "eval_loss": 3.695894241333008, |
| "eval_runtime": 179.7781, |
| "eval_samples_per_second": 100.185, |
| "eval_steps_per_second": 6.263, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2969540415455818, |
| "grad_norm": 0.587378740310669, |
| "learning_rate": 0.0005227647882771253, |
| "loss": 3.7458, |
| "step": 12050 |
| }, |
| { |
| "epoch": 1.30233559358519, |
| "grad_norm": 0.5950103402137756, |
| "learning_rate": 0.0005224415472470639, |
| "loss": 3.7297, |
| "step": 12100 |
| }, |
| { |
| "epoch": 1.3077171456247982, |
| "grad_norm": 0.5634168386459351, |
| "learning_rate": 0.0005221183062170024, |
| "loss": 3.7124, |
| "step": 12150 |
| }, |
| { |
| "epoch": 1.3130986976644063, |
| "grad_norm": 0.5455742478370667, |
| "learning_rate": 0.0005217950651869409, |
| "loss": 3.7115, |
| "step": 12200 |
| }, |
| { |
| "epoch": 1.3184802497040147, |
| "grad_norm": 0.5473319888114929, |
| "learning_rate": 0.0005214718241568796, |
| "loss": 3.7046, |
| "step": 12250 |
| }, |
| { |
| "epoch": 1.3238618017436228, |
| "grad_norm": 0.5351397395133972, |
| "learning_rate": 0.0005211485831268182, |
| "loss": 3.7516, |
| "step": 12300 |
| }, |
| { |
| "epoch": 1.329243353783231, |
| "grad_norm": 0.5263974070549011, |
| "learning_rate": 0.0005208253420967568, |
| "loss": 3.7259, |
| "step": 12350 |
| }, |
| { |
| "epoch": 1.3346249058228392, |
| "grad_norm": 0.544121503829956, |
| "learning_rate": 0.0005205021010666953, |
| "loss": 3.7135, |
| "step": 12400 |
| }, |
| { |
| "epoch": 1.3400064578624475, |
| "grad_norm": 0.5584757924079895, |
| "learning_rate": 0.0005201788600366339, |
| "loss": 3.7257, |
| "step": 12450 |
| }, |
| { |
| "epoch": 1.3453880099020559, |
| "grad_norm": 0.5996146202087402, |
| "learning_rate": 0.0005198556190065725, |
| "loss": 3.7364, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.350769561941664, |
| "grad_norm": 0.5599164366722107, |
| "learning_rate": 0.0005195323779765112, |
| "loss": 3.7402, |
| "step": 12550 |
| }, |
| { |
| "epoch": 1.356151113981272, |
| "grad_norm": 0.6167038083076477, |
| "learning_rate": 0.0005192091369464497, |
| "loss": 3.7381, |
| "step": 12600 |
| }, |
| { |
| "epoch": 1.3615326660208804, |
| "grad_norm": 0.5943508744239807, |
| "learning_rate": 0.0005188858959163882, |
| "loss": 3.7255, |
| "step": 12650 |
| }, |
| { |
| "epoch": 1.3669142180604887, |
| "grad_norm": 0.5350813269615173, |
| "learning_rate": 0.0005185626548863269, |
| "loss": 3.7235, |
| "step": 12700 |
| }, |
| { |
| "epoch": 1.3722957701000968, |
| "grad_norm": 0.5343997478485107, |
| "learning_rate": 0.0005182458786768667, |
| "loss": 3.7154, |
| "step": 12750 |
| }, |
| { |
| "epoch": 1.3776773221397052, |
| "grad_norm": 0.6642977595329285, |
| "learning_rate": 0.0005179226376468052, |
| "loss": 3.7167, |
| "step": 12800 |
| }, |
| { |
| "epoch": 1.3830588741793133, |
| "grad_norm": 0.56437748670578, |
| "learning_rate": 0.0005175993966167438, |
| "loss": 3.7007, |
| "step": 12850 |
| }, |
| { |
| "epoch": 1.3884404262189216, |
| "grad_norm": 0.5799341797828674, |
| "learning_rate": 0.0005172761555866824, |
| "loss": 3.7025, |
| "step": 12900 |
| }, |
| { |
| "epoch": 1.3938219782585297, |
| "grad_norm": 0.6053862571716309, |
| "learning_rate": 0.0005169529145566211, |
| "loss": 3.6903, |
| "step": 12950 |
| }, |
| { |
| "epoch": 1.399203530298138, |
| "grad_norm": 0.5302448868751526, |
| "learning_rate": 0.0005166296735265596, |
| "loss": 3.711, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.399203530298138, |
| "eval_accuracy": 0.3507369545095349, |
| "eval_loss": 3.6787874698638916, |
| "eval_runtime": 179.7451, |
| "eval_samples_per_second": 100.203, |
| "eval_steps_per_second": 6.264, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.4045850823377461, |
| "grad_norm": 0.5383056402206421, |
| "learning_rate": 0.0005163064324964982, |
| "loss": 3.7115, |
| "step": 13050 |
| }, |
| { |
| "epoch": 1.4099666343773545, |
| "grad_norm": 0.5403004288673401, |
| "learning_rate": 0.0005159831914664367, |
| "loss": 3.7064, |
| "step": 13100 |
| }, |
| { |
| "epoch": 1.4153481864169626, |
| "grad_norm": 0.5427159070968628, |
| "learning_rate": 0.0005156599504363753, |
| "loss": 3.7023, |
| "step": 13150 |
| }, |
| { |
| "epoch": 1.420729738456571, |
| "grad_norm": 0.5563752055168152, |
| "learning_rate": 0.000515336709406314, |
| "loss": 3.7225, |
| "step": 13200 |
| }, |
| { |
| "epoch": 1.426111290496179, |
| "grad_norm": 0.6402376890182495, |
| "learning_rate": 0.0005150134683762525, |
| "loss": 3.7011, |
| "step": 13250 |
| }, |
| { |
| "epoch": 1.4314928425357873, |
| "grad_norm": 0.5857889652252197, |
| "learning_rate": 0.0005146902273461911, |
| "loss": 3.7031, |
| "step": 13300 |
| }, |
| { |
| "epoch": 1.4368743945753955, |
| "grad_norm": 0.6158170104026794, |
| "learning_rate": 0.0005143669863161297, |
| "loss": 3.6897, |
| "step": 13350 |
| }, |
| { |
| "epoch": 1.4422559466150038, |
| "grad_norm": 0.6112419962882996, |
| "learning_rate": 0.0005140437452860683, |
| "loss": 3.7072, |
| "step": 13400 |
| }, |
| { |
| "epoch": 1.447637498654612, |
| "grad_norm": 0.5744062662124634, |
| "learning_rate": 0.0005137205042560069, |
| "loss": 3.7056, |
| "step": 13450 |
| }, |
| { |
| "epoch": 1.4530190506942202, |
| "grad_norm": 0.6232595443725586, |
| "learning_rate": 0.0005133972632259455, |
| "loss": 3.7002, |
| "step": 13500 |
| }, |
| { |
| "epoch": 1.4584006027338283, |
| "grad_norm": 0.567087709903717, |
| "learning_rate": 0.000513074022195884, |
| "loss": 3.704, |
| "step": 13550 |
| }, |
| { |
| "epoch": 1.4637821547734367, |
| "grad_norm": 0.5451653003692627, |
| "learning_rate": 0.0005127507811658226, |
| "loss": 3.7109, |
| "step": 13600 |
| }, |
| { |
| "epoch": 1.469163706813045, |
| "grad_norm": 0.671532154083252, |
| "learning_rate": 0.0005124275401357612, |
| "loss": 3.6943, |
| "step": 13650 |
| }, |
| { |
| "epoch": 1.474545258852653, |
| "grad_norm": 0.542718768119812, |
| "learning_rate": 0.0005121042991056997, |
| "loss": 3.6789, |
| "step": 13700 |
| }, |
| { |
| "epoch": 1.4799268108922612, |
| "grad_norm": 0.5489877462387085, |
| "learning_rate": 0.0005117810580756384, |
| "loss": 3.714, |
| "step": 13750 |
| }, |
| { |
| "epoch": 1.4853083629318695, |
| "grad_norm": 0.5520055890083313, |
| "learning_rate": 0.0005114578170455769, |
| "loss": 3.6945, |
| "step": 13800 |
| }, |
| { |
| "epoch": 1.4906899149714778, |
| "grad_norm": 0.5908713340759277, |
| "learning_rate": 0.0005111345760155156, |
| "loss": 3.6954, |
| "step": 13850 |
| }, |
| { |
| "epoch": 1.496071467011086, |
| "grad_norm": 0.5579512119293213, |
| "learning_rate": 0.0005108113349854541, |
| "loss": 3.7161, |
| "step": 13900 |
| }, |
| { |
| "epoch": 1.501453019050694, |
| "grad_norm": 0.5487799048423767, |
| "learning_rate": 0.0005104880939553926, |
| "loss": 3.7058, |
| "step": 13950 |
| }, |
| { |
| "epoch": 1.5068345710903024, |
| "grad_norm": 0.5599663257598877, |
| "learning_rate": 0.0005101648529253313, |
| "loss": 3.7049, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5068345710903024, |
| "eval_accuracy": 0.35338612922022755, |
| "eval_loss": 3.6561355590820312, |
| "eval_runtime": 179.7601, |
| "eval_samples_per_second": 100.195, |
| "eval_steps_per_second": 6.264, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5122161231299107, |
| "grad_norm": 0.5696370601654053, |
| "learning_rate": 0.0005098416118952699, |
| "loss": 3.6653, |
| "step": 14050 |
| }, |
| { |
| "epoch": 1.5175976751695188, |
| "grad_norm": 0.7141689658164978, |
| "learning_rate": 0.0005095183708652085, |
| "loss": 3.6826, |
| "step": 14100 |
| }, |
| { |
| "epoch": 1.5229792272091272, |
| "grad_norm": 0.6532941460609436, |
| "learning_rate": 0.000509195129835147, |
| "loss": 3.7016, |
| "step": 14150 |
| }, |
| { |
| "epoch": 1.5283607792487355, |
| "grad_norm": 0.5246692299842834, |
| "learning_rate": 0.0005088718888050856, |
| "loss": 3.6981, |
| "step": 14200 |
| }, |
| { |
| "epoch": 1.5337423312883436, |
| "grad_norm": 0.5713950991630554, |
| "learning_rate": 0.0005085486477750242, |
| "loss": 3.6811, |
| "step": 14250 |
| }, |
| { |
| "epoch": 1.5391238833279517, |
| "grad_norm": 0.5982765555381775, |
| "learning_rate": 0.0005082254067449629, |
| "loss": 3.7083, |
| "step": 14300 |
| }, |
| { |
| "epoch": 1.54450543536756, |
| "grad_norm": 0.5973095893859863, |
| "learning_rate": 0.0005079021657149014, |
| "loss": 3.6941, |
| "step": 14350 |
| }, |
| { |
| "epoch": 1.5498869874071683, |
| "grad_norm": 0.555167555809021, |
| "learning_rate": 0.0005075789246848399, |
| "loss": 3.6879, |
| "step": 14400 |
| }, |
| { |
| "epoch": 1.5552685394467765, |
| "grad_norm": 0.5746867060661316, |
| "learning_rate": 0.0005072556836547785, |
| "loss": 3.6938, |
| "step": 14450 |
| }, |
| { |
| "epoch": 1.5606500914863846, |
| "grad_norm": 0.5191563367843628, |
| "learning_rate": 0.0005069324426247171, |
| "loss": 3.6978, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.566031643525993, |
| "grad_norm": 0.5576121807098389, |
| "learning_rate": 0.0005066092015946557, |
| "loss": 3.7036, |
| "step": 14550 |
| }, |
| { |
| "epoch": 1.5714131955656012, |
| "grad_norm": 0.5687995553016663, |
| "learning_rate": 0.0005062859605645943, |
| "loss": 3.6899, |
| "step": 14600 |
| }, |
| { |
| "epoch": 1.5767947476052093, |
| "grad_norm": 0.6132740378379822, |
| "learning_rate": 0.0005059627195345329, |
| "loss": 3.6664, |
| "step": 14650 |
| }, |
| { |
| "epoch": 1.5821762996448174, |
| "grad_norm": 0.5791540145874023, |
| "learning_rate": 0.0005056394785044715, |
| "loss": 3.6893, |
| "step": 14700 |
| }, |
| { |
| "epoch": 1.5875578516844258, |
| "grad_norm": 0.5374217629432678, |
| "learning_rate": 0.00050531623747441, |
| "loss": 3.7027, |
| "step": 14750 |
| }, |
| { |
| "epoch": 1.592939403724034, |
| "grad_norm": 0.5353657603263855, |
| "learning_rate": 0.0005049994612649499, |
| "loss": 3.6779, |
| "step": 14800 |
| }, |
| { |
| "epoch": 1.5983209557636422, |
| "grad_norm": 0.6020593047142029, |
| "learning_rate": 0.0005046762202348884, |
| "loss": 3.6988, |
| "step": 14850 |
| }, |
| { |
| "epoch": 1.6037025078032503, |
| "grad_norm": 0.6164742708206177, |
| "learning_rate": 0.000504352979204827, |
| "loss": 3.711, |
| "step": 14900 |
| }, |
| { |
| "epoch": 1.6090840598428586, |
| "grad_norm": 0.5817080140113831, |
| "learning_rate": 0.0005040297381747656, |
| "loss": 3.6941, |
| "step": 14950 |
| }, |
| { |
| "epoch": 1.614465611882467, |
| "grad_norm": 0.5781374573707581, |
| "learning_rate": 0.0005037064971447042, |
| "loss": 3.6648, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.614465611882467, |
| "eval_accuracy": 0.35519965457073904, |
| "eval_loss": 3.6373958587646484, |
| "eval_runtime": 179.8191, |
| "eval_samples_per_second": 100.162, |
| "eval_steps_per_second": 6.262, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.619847163922075, |
| "grad_norm": 0.5674581527709961, |
| "learning_rate": 0.0005033832561146428, |
| "loss": 3.7027, |
| "step": 15050 |
| }, |
| { |
| "epoch": 1.6252287159616834, |
| "grad_norm": 0.5419674515724182, |
| "learning_rate": 0.0005030600150845813, |
| "loss": 3.6874, |
| "step": 15100 |
| }, |
| { |
| "epoch": 1.6306102680012917, |
| "grad_norm": 0.5471088886260986, |
| "learning_rate": 0.00050273677405452, |
| "loss": 3.6781, |
| "step": 15150 |
| }, |
| { |
| "epoch": 1.6359918200408998, |
| "grad_norm": 0.5623396039009094, |
| "learning_rate": 0.0005024135330244585, |
| "loss": 3.6821, |
| "step": 15200 |
| }, |
| { |
| "epoch": 1.641373372080508, |
| "grad_norm": 0.5248699188232422, |
| "learning_rate": 0.0005020902919943972, |
| "loss": 3.6966, |
| "step": 15250 |
| }, |
| { |
| "epoch": 1.6467549241201163, |
| "grad_norm": 0.5599279403686523, |
| "learning_rate": 0.0005017670509643357, |
| "loss": 3.6825, |
| "step": 15300 |
| }, |
| { |
| "epoch": 1.6521364761597246, |
| "grad_norm": 0.5801636576652527, |
| "learning_rate": 0.0005014438099342743, |
| "loss": 3.6487, |
| "step": 15350 |
| }, |
| { |
| "epoch": 1.6575180281993327, |
| "grad_norm": 0.5466539263725281, |
| "learning_rate": 0.0005011205689042129, |
| "loss": 3.6799, |
| "step": 15400 |
| }, |
| { |
| "epoch": 1.6628995802389408, |
| "grad_norm": 0.5191233158111572, |
| "learning_rate": 0.0005007973278741514, |
| "loss": 3.6652, |
| "step": 15450 |
| }, |
| { |
| "epoch": 1.6682811322785491, |
| "grad_norm": 0.538023829460144, |
| "learning_rate": 0.00050047408684409, |
| "loss": 3.6783, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.6736626843181575, |
| "grad_norm": 0.5874162316322327, |
| "learning_rate": 0.0005001508458140286, |
| "loss": 3.6757, |
| "step": 15550 |
| }, |
| { |
| "epoch": 1.6790442363577656, |
| "grad_norm": 0.5796945691108704, |
| "learning_rate": 0.0004998276047839673, |
| "loss": 3.6699, |
| "step": 15600 |
| }, |
| { |
| "epoch": 1.6844257883973737, |
| "grad_norm": 0.6127719879150391, |
| "learning_rate": 0.0004995043637539058, |
| "loss": 3.6761, |
| "step": 15650 |
| }, |
| { |
| "epoch": 1.689807340436982, |
| "grad_norm": 0.5545514822006226, |
| "learning_rate": 0.0004991811227238443, |
| "loss": 3.6704, |
| "step": 15700 |
| }, |
| { |
| "epoch": 1.6951888924765903, |
| "grad_norm": 0.59245765209198, |
| "learning_rate": 0.0004988578816937829, |
| "loss": 3.6674, |
| "step": 15750 |
| }, |
| { |
| "epoch": 1.7005704445161984, |
| "grad_norm": 0.5616726279258728, |
| "learning_rate": 0.0004985346406637215, |
| "loss": 3.6786, |
| "step": 15800 |
| }, |
| { |
| "epoch": 1.7059519965558065, |
| "grad_norm": 0.5251116156578064, |
| "learning_rate": 0.0004982113996336602, |
| "loss": 3.6491, |
| "step": 15850 |
| }, |
| { |
| "epoch": 1.7113335485954149, |
| "grad_norm": 0.5580370426177979, |
| "learning_rate": 0.0004978881586035987, |
| "loss": 3.6599, |
| "step": 15900 |
| }, |
| { |
| "epoch": 1.7167151006350232, |
| "grad_norm": 0.5888770818710327, |
| "learning_rate": 0.0004975649175735373, |
| "loss": 3.6759, |
| "step": 15950 |
| }, |
| { |
| "epoch": 1.7220966526746313, |
| "grad_norm": 0.561681866645813, |
| "learning_rate": 0.0004972416765434759, |
| "loss": 3.6552, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7220966526746313, |
| "eval_accuracy": 0.3567330727764648, |
| "eval_loss": 3.6214287281036377, |
| "eval_runtime": 179.9334, |
| "eval_samples_per_second": 100.098, |
| "eval_steps_per_second": 6.258, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7274782047142396, |
| "grad_norm": 0.5320652723312378, |
| "learning_rate": 0.0004969184355134145, |
| "loss": 3.6572, |
| "step": 16050 |
| }, |
| { |
| "epoch": 1.732859756753848, |
| "grad_norm": 0.5540884733200073, |
| "learning_rate": 0.0004965951944833531, |
| "loss": 3.6749, |
| "step": 16100 |
| }, |
| { |
| "epoch": 1.738241308793456, |
| "grad_norm": 0.6164997220039368, |
| "learning_rate": 0.0004962719534532916, |
| "loss": 3.6743, |
| "step": 16150 |
| }, |
| { |
| "epoch": 1.7436228608330642, |
| "grad_norm": 0.5549599528312683, |
| "learning_rate": 0.0004959487124232302, |
| "loss": 3.6486, |
| "step": 16200 |
| }, |
| { |
| "epoch": 1.7490044128726725, |
| "grad_norm": 0.5784186720848083, |
| "learning_rate": 0.0004956254713931688, |
| "loss": 3.655, |
| "step": 16250 |
| }, |
| { |
| "epoch": 1.7543859649122808, |
| "grad_norm": 0.6285885572433472, |
| "learning_rate": 0.0004953022303631074, |
| "loss": 3.6737, |
| "step": 16300 |
| }, |
| { |
| "epoch": 1.759767516951889, |
| "grad_norm": 0.5278353095054626, |
| "learning_rate": 0.0004949789893330459, |
| "loss": 3.6691, |
| "step": 16350 |
| }, |
| { |
| "epoch": 1.765149068991497, |
| "grad_norm": 0.5532869100570679, |
| "learning_rate": 0.0004946557483029846, |
| "loss": 3.6645, |
| "step": 16400 |
| }, |
| { |
| "epoch": 1.7705306210311054, |
| "grad_norm": 0.5599241852760315, |
| "learning_rate": 0.0004943325072729231, |
| "loss": 3.6569, |
| "step": 16450 |
| }, |
| { |
| "epoch": 1.7759121730707137, |
| "grad_norm": 0.5885684490203857, |
| "learning_rate": 0.0004940092662428617, |
| "loss": 3.6504, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.7812937251103218, |
| "grad_norm": 0.5403637290000916, |
| "learning_rate": 0.0004936860252128003, |
| "loss": 3.667, |
| "step": 16550 |
| }, |
| { |
| "epoch": 1.78667527714993, |
| "grad_norm": 0.5659014582633972, |
| "learning_rate": 0.0004933627841827388, |
| "loss": 3.6669, |
| "step": 16600 |
| }, |
| { |
| "epoch": 1.7920568291895382, |
| "grad_norm": 0.5648633241653442, |
| "learning_rate": 0.0004930395431526775, |
| "loss": 3.6579, |
| "step": 16650 |
| }, |
| { |
| "epoch": 1.7974383812291466, |
| "grad_norm": 0.5897268056869507, |
| "learning_rate": 0.0004927163021226161, |
| "loss": 3.6617, |
| "step": 16700 |
| }, |
| { |
| "epoch": 1.8028199332687547, |
| "grad_norm": 0.5169820189476013, |
| "learning_rate": 0.0004923930610925547, |
| "loss": 3.6545, |
| "step": 16750 |
| }, |
| { |
| "epoch": 1.8082014853083628, |
| "grad_norm": 0.5676159858703613, |
| "learning_rate": 0.0004920698200624932, |
| "loss": 3.6502, |
| "step": 16800 |
| }, |
| { |
| "epoch": 1.813583037347971, |
| "grad_norm": 0.629549503326416, |
| "learning_rate": 0.000491753043853033, |
| "loss": 3.6371, |
| "step": 16850 |
| }, |
| { |
| "epoch": 1.8189645893875794, |
| "grad_norm": 0.5634384155273438, |
| "learning_rate": 0.0004914298028229717, |
| "loss": 3.6592, |
| "step": 16900 |
| }, |
| { |
| "epoch": 1.8243461414271875, |
| "grad_norm": 0.5446241497993469, |
| "learning_rate": 0.0004911065617929102, |
| "loss": 3.6693, |
| "step": 16950 |
| }, |
| { |
| "epoch": 1.8297276934667959, |
| "grad_norm": 0.5624287128448486, |
| "learning_rate": 0.0004907833207628487, |
| "loss": 3.6525, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8297276934667959, |
| "eval_accuracy": 0.3582982176254324, |
| "eval_loss": 3.606470823287964, |
| "eval_runtime": 179.8801, |
| "eval_samples_per_second": 100.128, |
| "eval_steps_per_second": 6.26, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8351092455064042, |
| "grad_norm": 0.5610666275024414, |
| "learning_rate": 0.0004904600797327874, |
| "loss": 3.6374, |
| "step": 17050 |
| }, |
| { |
| "epoch": 1.8404907975460123, |
| "grad_norm": 0.5537661910057068, |
| "learning_rate": 0.0004901368387027259, |
| "loss": 3.6434, |
| "step": 17100 |
| }, |
| { |
| "epoch": 1.8458723495856204, |
| "grad_norm": 0.5830965638160706, |
| "learning_rate": 0.0004898135976726646, |
| "loss": 3.6273, |
| "step": 17150 |
| }, |
| { |
| "epoch": 1.8512539016252287, |
| "grad_norm": 0.635771632194519, |
| "learning_rate": 0.0004894903566426031, |
| "loss": 3.6567, |
| "step": 17200 |
| }, |
| { |
| "epoch": 1.856635453664837, |
| "grad_norm": 0.6146743297576904, |
| "learning_rate": 0.0004891671156125417, |
| "loss": 3.6323, |
| "step": 17250 |
| }, |
| { |
| "epoch": 1.8620170057044452, |
| "grad_norm": 0.5971278548240662, |
| "learning_rate": 0.0004888438745824803, |
| "loss": 3.6194, |
| "step": 17300 |
| }, |
| { |
| "epoch": 1.8673985577440533, |
| "grad_norm": 0.4955149292945862, |
| "learning_rate": 0.0004885206335524189, |
| "loss": 3.6616, |
| "step": 17350 |
| }, |
| { |
| "epoch": 1.8727801097836616, |
| "grad_norm": 0.5917460322380066, |
| "learning_rate": 0.0004881973925223575, |
| "loss": 3.6489, |
| "step": 17400 |
| }, |
| { |
| "epoch": 1.87816166182327, |
| "grad_norm": 0.5226415991783142, |
| "learning_rate": 0.0004878806163128973, |
| "loss": 3.6379, |
| "step": 17450 |
| }, |
| { |
| "epoch": 1.883543213862878, |
| "grad_norm": 0.5437645316123962, |
| "learning_rate": 0.00048755737528283585, |
| "loss": 3.6462, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.8889247659024861, |
| "grad_norm": 0.5848841667175293, |
| "learning_rate": 0.00048723413425277444, |
| "loss": 3.6338, |
| "step": 17550 |
| }, |
| { |
| "epoch": 1.8943063179420945, |
| "grad_norm": 0.5598640441894531, |
| "learning_rate": 0.00048691089322271304, |
| "loss": 3.6487, |
| "step": 17600 |
| }, |
| { |
| "epoch": 1.8996878699817028, |
| "grad_norm": 0.626156747341156, |
| "learning_rate": 0.0004865876521926516, |
| "loss": 3.6352, |
| "step": 17650 |
| }, |
| { |
| "epoch": 1.905069422021311, |
| "grad_norm": 0.589815080165863, |
| "learning_rate": 0.0004862644111625902, |
| "loss": 3.6364, |
| "step": 17700 |
| }, |
| { |
| "epoch": 1.910450974060919, |
| "grad_norm": 0.6275328993797302, |
| "learning_rate": 0.0004859411701325288, |
| "loss": 3.6526, |
| "step": 17750 |
| }, |
| { |
| "epoch": 1.9158325261005273, |
| "grad_norm": 0.5926043391227722, |
| "learning_rate": 0.00048561792910246736, |
| "loss": 3.6384, |
| "step": 17800 |
| }, |
| { |
| "epoch": 1.9212140781401357, |
| "grad_norm": 0.5851974487304688, |
| "learning_rate": 0.00048529468807240596, |
| "loss": 3.6396, |
| "step": 17850 |
| }, |
| { |
| "epoch": 1.9265956301797438, |
| "grad_norm": 0.5288086533546448, |
| "learning_rate": 0.0004849714470423445, |
| "loss": 3.6497, |
| "step": 17900 |
| }, |
| { |
| "epoch": 1.931977182219352, |
| "grad_norm": 0.5444283485412598, |
| "learning_rate": 0.0004846482060122831, |
| "loss": 3.6326, |
| "step": 17950 |
| }, |
| { |
| "epoch": 1.9373587342589604, |
| "grad_norm": 0.507646918296814, |
| "learning_rate": 0.00048432496498222174, |
| "loss": 3.6271, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9373587342589604, |
| "eval_accuracy": 0.35979230348576935, |
| "eval_loss": 3.591550827026367, |
| "eval_runtime": 179.8657, |
| "eval_samples_per_second": 100.136, |
| "eval_steps_per_second": 6.26, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9427402862985685, |
| "grad_norm": 0.5543516278266907, |
| "learning_rate": 0.0004840017239521603, |
| "loss": 3.6373, |
| "step": 18050 |
| }, |
| { |
| "epoch": 1.9481218383381766, |
| "grad_norm": 0.5450665950775146, |
| "learning_rate": 0.0004836784829220989, |
| "loss": 3.6373, |
| "step": 18100 |
| }, |
| { |
| "epoch": 1.953503390377785, |
| "grad_norm": 0.5807346701622009, |
| "learning_rate": 0.00048335524189203747, |
| "loss": 3.6481, |
| "step": 18150 |
| }, |
| { |
| "epoch": 1.9588849424173933, |
| "grad_norm": 0.5595925450325012, |
| "learning_rate": 0.000483032000861976, |
| "loss": 3.6223, |
| "step": 18200 |
| }, |
| { |
| "epoch": 1.9642664944570014, |
| "grad_norm": 0.6111944913864136, |
| "learning_rate": 0.0004827087598319146, |
| "loss": 3.6431, |
| "step": 18250 |
| }, |
| { |
| "epoch": 1.9696480464966095, |
| "grad_norm": 0.5740875601768494, |
| "learning_rate": 0.00048238551880185325, |
| "loss": 3.6397, |
| "step": 18300 |
| }, |
| { |
| "epoch": 1.9750295985362178, |
| "grad_norm": 0.5665279626846313, |
| "learning_rate": 0.0004820622777717918, |
| "loss": 3.6398, |
| "step": 18350 |
| }, |
| { |
| "epoch": 1.9804111505758262, |
| "grad_norm": 0.5787563323974609, |
| "learning_rate": 0.0004817390367417304, |
| "loss": 3.6241, |
| "step": 18400 |
| }, |
| { |
| "epoch": 1.9857927026154343, |
| "grad_norm": 0.5683515071868896, |
| "learning_rate": 0.00048141579571166893, |
| "loss": 3.6398, |
| "step": 18450 |
| }, |
| { |
| "epoch": 1.9911742546550424, |
| "grad_norm": 0.5649462938308716, |
| "learning_rate": 0.0004810925546816075, |
| "loss": 3.6431, |
| "step": 18500 |
| }, |
| { |
| "epoch": 1.9965558066946507, |
| "grad_norm": 0.5490776300430298, |
| "learning_rate": 0.0004807693136515461, |
| "loss": 3.6415, |
| "step": 18550 |
| }, |
| { |
| "epoch": 2.001937358734259, |
| "grad_norm": 0.5904884934425354, |
| "learning_rate": 0.0004804460726214847, |
| "loss": 3.5982, |
| "step": 18600 |
| }, |
| { |
| "epoch": 2.007318910773867, |
| "grad_norm": 0.5800067186355591, |
| "learning_rate": 0.0004801228315914233, |
| "loss": 3.5538, |
| "step": 18650 |
| }, |
| { |
| "epoch": 2.0127004628134753, |
| "grad_norm": 0.5690127611160278, |
| "learning_rate": 0.0004797995905613619, |
| "loss": 3.5412, |
| "step": 18700 |
| }, |
| { |
| "epoch": 2.018082014853084, |
| "grad_norm": 0.5800209641456604, |
| "learning_rate": 0.00047947634953130044, |
| "loss": 3.554, |
| "step": 18750 |
| }, |
| { |
| "epoch": 2.023463566892692, |
| "grad_norm": 0.595440685749054, |
| "learning_rate": 0.00047915310850123904, |
| "loss": 3.538, |
| "step": 18800 |
| }, |
| { |
| "epoch": 2.0288451189323, |
| "grad_norm": 0.5423743724822998, |
| "learning_rate": 0.0004788298674711777, |
| "loss": 3.544, |
| "step": 18850 |
| }, |
| { |
| "epoch": 2.034226670971908, |
| "grad_norm": 0.5956675410270691, |
| "learning_rate": 0.00047850662644111623, |
| "loss": 3.5531, |
| "step": 18900 |
| }, |
| { |
| "epoch": 2.0396082230115167, |
| "grad_norm": 0.531262993812561, |
| "learning_rate": 0.0004781833854110548, |
| "loss": 3.5422, |
| "step": 18950 |
| }, |
| { |
| "epoch": 2.044989775051125, |
| "grad_norm": 0.6084029674530029, |
| "learning_rate": 0.00047786014438099336, |
| "loss": 3.5643, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.044989775051125, |
| "eval_accuracy": 0.3612721558178026, |
| "eval_loss": 3.5829977989196777, |
| "eval_runtime": 179.5266, |
| "eval_samples_per_second": 100.325, |
| "eval_steps_per_second": 6.272, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.050371327090733, |
| "grad_norm": 0.5787180662155151, |
| "learning_rate": 0.00047753690335093196, |
| "loss": 3.5347, |
| "step": 19050 |
| }, |
| { |
| "epoch": 2.055752879130341, |
| "grad_norm": 0.5521721839904785, |
| "learning_rate": 0.00047721366232087055, |
| "loss": 3.555, |
| "step": 19100 |
| }, |
| { |
| "epoch": 2.0611344311699495, |
| "grad_norm": 0.584023118019104, |
| "learning_rate": 0.00047689042129080915, |
| "loss": 3.556, |
| "step": 19150 |
| }, |
| { |
| "epoch": 2.0665159832095576, |
| "grad_norm": 0.6237092018127441, |
| "learning_rate": 0.00047656718026074774, |
| "loss": 3.5654, |
| "step": 19200 |
| }, |
| { |
| "epoch": 2.0718975352491658, |
| "grad_norm": 0.5408397316932678, |
| "learning_rate": 0.00047624393923068634, |
| "loss": 3.5346, |
| "step": 19250 |
| }, |
| { |
| "epoch": 2.0772790872887743, |
| "grad_norm": 0.6254966855049133, |
| "learning_rate": 0.0004759206982006249, |
| "loss": 3.5376, |
| "step": 19300 |
| }, |
| { |
| "epoch": 2.0826606393283824, |
| "grad_norm": 0.5704235434532166, |
| "learning_rate": 0.00047559745717056347, |
| "loss": 3.5709, |
| "step": 19350 |
| }, |
| { |
| "epoch": 2.0880421913679905, |
| "grad_norm": 0.5291028022766113, |
| "learning_rate": 0.000475274216140502, |
| "loss": 3.5693, |
| "step": 19400 |
| }, |
| { |
| "epoch": 2.0934237434075986, |
| "grad_norm": 0.5555849671363831, |
| "learning_rate": 0.00047495097511044066, |
| "loss": 3.552, |
| "step": 19450 |
| }, |
| { |
| "epoch": 2.098805295447207, |
| "grad_norm": 0.5503198504447937, |
| "learning_rate": 0.00047462773408037925, |
| "loss": 3.5421, |
| "step": 19500 |
| }, |
| { |
| "epoch": 2.1041868474868153, |
| "grad_norm": 0.5942191481590271, |
| "learning_rate": 0.0004743044930503178, |
| "loss": 3.5594, |
| "step": 19550 |
| }, |
| { |
| "epoch": 2.1095683995264234, |
| "grad_norm": 0.5651226043701172, |
| "learning_rate": 0.0004739812520202564, |
| "loss": 3.554, |
| "step": 19600 |
| }, |
| { |
| "epoch": 2.1149499515660315, |
| "grad_norm": 0.5572126507759094, |
| "learning_rate": 0.000473658010990195, |
| "loss": 3.5501, |
| "step": 19650 |
| }, |
| { |
| "epoch": 2.12033150360564, |
| "grad_norm": 0.5130614042282104, |
| "learning_rate": 0.0004733347699601336, |
| "loss": 3.5539, |
| "step": 19700 |
| }, |
| { |
| "epoch": 2.125713055645248, |
| "grad_norm": 0.5788480639457703, |
| "learning_rate": 0.0004730115289300722, |
| "loss": 3.5763, |
| "step": 19750 |
| }, |
| { |
| "epoch": 2.1310946076848563, |
| "grad_norm": 0.5898392796516418, |
| "learning_rate": 0.00047268828790001077, |
| "loss": 3.563, |
| "step": 19800 |
| }, |
| { |
| "epoch": 2.1364761597244644, |
| "grad_norm": 0.5292854309082031, |
| "learning_rate": 0.0004723650468699493, |
| "loss": 3.5473, |
| "step": 19850 |
| }, |
| { |
| "epoch": 2.141857711764073, |
| "grad_norm": 0.5742781758308411, |
| "learning_rate": 0.0004720418058398879, |
| "loss": 3.5564, |
| "step": 19900 |
| }, |
| { |
| "epoch": 2.147239263803681, |
| "grad_norm": 0.5774244666099548, |
| "learning_rate": 0.00047171856480982644, |
| "loss": 3.565, |
| "step": 19950 |
| }, |
| { |
| "epoch": 2.152620815843289, |
| "grad_norm": 0.652957558631897, |
| "learning_rate": 0.0004713953237797651, |
| "loss": 3.5546, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.152620815843289, |
| "eval_accuracy": 0.3621433346722991, |
| "eval_loss": 3.573176860809326, |
| "eval_runtime": 180.022, |
| "eval_samples_per_second": 100.049, |
| "eval_steps_per_second": 6.255, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.1580023678828972, |
| "grad_norm": 0.5954522490501404, |
| "learning_rate": 0.0004710720827497037, |
| "loss": 3.5618, |
| "step": 20050 |
| }, |
| { |
| "epoch": 2.163383919922506, |
| "grad_norm": 0.5455233454704285, |
| "learning_rate": 0.00047074884171964223, |
| "loss": 3.5538, |
| "step": 20100 |
| }, |
| { |
| "epoch": 2.168765471962114, |
| "grad_norm": 0.5485193133354187, |
| "learning_rate": 0.0004704256006895808, |
| "loss": 3.5532, |
| "step": 20150 |
| }, |
| { |
| "epoch": 2.174147024001722, |
| "grad_norm": 0.5491883158683777, |
| "learning_rate": 0.00047010235965951936, |
| "loss": 3.546, |
| "step": 20200 |
| }, |
| { |
| "epoch": 2.1795285760413305, |
| "grad_norm": 0.627436101436615, |
| "learning_rate": 0.00046977911862945796, |
| "loss": 3.5433, |
| "step": 20250 |
| }, |
| { |
| "epoch": 2.1849101280809387, |
| "grad_norm": 0.5892840027809143, |
| "learning_rate": 0.0004694558775993966, |
| "loss": 3.565, |
| "step": 20300 |
| }, |
| { |
| "epoch": 2.1902916801205468, |
| "grad_norm": 0.594659686088562, |
| "learning_rate": 0.00046913263656933515, |
| "loss": 3.5551, |
| "step": 20350 |
| }, |
| { |
| "epoch": 2.195673232160155, |
| "grad_norm": 0.578382670879364, |
| "learning_rate": 0.00046880939553927374, |
| "loss": 3.5707, |
| "step": 20400 |
| }, |
| { |
| "epoch": 2.2010547841997634, |
| "grad_norm": 0.5580980181694031, |
| "learning_rate": 0.00046848615450921234, |
| "loss": 3.5548, |
| "step": 20450 |
| }, |
| { |
| "epoch": 2.2064363362393715, |
| "grad_norm": 0.5704784989356995, |
| "learning_rate": 0.0004681629134791509, |
| "loss": 3.5635, |
| "step": 20500 |
| }, |
| { |
| "epoch": 2.2118178882789796, |
| "grad_norm": 0.5783650279045105, |
| "learning_rate": 0.00046783967244908947, |
| "loss": 3.5705, |
| "step": 20550 |
| }, |
| { |
| "epoch": 2.2171994403185877, |
| "grad_norm": 0.5566691160202026, |
| "learning_rate": 0.0004675164314190281, |
| "loss": 3.5542, |
| "step": 20600 |
| }, |
| { |
| "epoch": 2.2225809923581963, |
| "grad_norm": 0.5815993547439575, |
| "learning_rate": 0.00046719319038896666, |
| "loss": 3.5566, |
| "step": 20650 |
| }, |
| { |
| "epoch": 2.2279625443978044, |
| "grad_norm": 0.5823930501937866, |
| "learning_rate": 0.00046686994935890526, |
| "loss": 3.5397, |
| "step": 20700 |
| }, |
| { |
| "epoch": 2.2333440964374125, |
| "grad_norm": 0.624121904373169, |
| "learning_rate": 0.0004665467083288438, |
| "loss": 3.5692, |
| "step": 20750 |
| }, |
| { |
| "epoch": 2.2387256484770206, |
| "grad_norm": 0.6141852140426636, |
| "learning_rate": 0.0004662234672987824, |
| "loss": 3.5432, |
| "step": 20800 |
| }, |
| { |
| "epoch": 2.244107200516629, |
| "grad_norm": 0.5725633502006531, |
| "learning_rate": 0.00046590022626872104, |
| "loss": 3.5536, |
| "step": 20850 |
| }, |
| { |
| "epoch": 2.2494887525562373, |
| "grad_norm": 0.6109887957572937, |
| "learning_rate": 0.0004655769852386596, |
| "loss": 3.5639, |
| "step": 20900 |
| }, |
| { |
| "epoch": 2.2548703045958454, |
| "grad_norm": 0.6038004159927368, |
| "learning_rate": 0.0004652537442085982, |
| "loss": 3.5691, |
| "step": 20950 |
| }, |
| { |
| "epoch": 2.2602518566354535, |
| "grad_norm": 0.5455430150032043, |
| "learning_rate": 0.00046493050317853677, |
| "loss": 3.5598, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.2602518566354535, |
| "eval_accuracy": 0.3633288463312429, |
| "eval_loss": 3.563311815261841, |
| "eval_runtime": 179.7973, |
| "eval_samples_per_second": 100.174, |
| "eval_steps_per_second": 6.263, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.265633408675062, |
| "grad_norm": 0.5887195467948914, |
| "learning_rate": 0.0004646072621484753, |
| "loss": 3.5452, |
| "step": 21050 |
| }, |
| { |
| "epoch": 2.27101496071467, |
| "grad_norm": 0.5390068888664246, |
| "learning_rate": 0.0004642840211184139, |
| "loss": 3.5495, |
| "step": 21100 |
| }, |
| { |
| "epoch": 2.2763965127542782, |
| "grad_norm": 0.5845677852630615, |
| "learning_rate": 0.00046396078008835255, |
| "loss": 3.5614, |
| "step": 21150 |
| }, |
| { |
| "epoch": 2.281778064793887, |
| "grad_norm": 0.5742790102958679, |
| "learning_rate": 0.0004636375390582911, |
| "loss": 3.569, |
| "step": 21200 |
| }, |
| { |
| "epoch": 2.287159616833495, |
| "grad_norm": 0.5437269806861877, |
| "learning_rate": 0.0004633142980282297, |
| "loss": 3.565, |
| "step": 21250 |
| }, |
| { |
| "epoch": 2.292541168873103, |
| "grad_norm": 0.6048343181610107, |
| "learning_rate": 0.00046299105699816823, |
| "loss": 3.5613, |
| "step": 21300 |
| }, |
| { |
| "epoch": 2.297922720912711, |
| "grad_norm": 0.6674041748046875, |
| "learning_rate": 0.0004626678159681068, |
| "loss": 3.5655, |
| "step": 21350 |
| }, |
| { |
| "epoch": 2.303304272952319, |
| "grad_norm": 0.6118209958076477, |
| "learning_rate": 0.0004623445749380454, |
| "loss": 3.5589, |
| "step": 21400 |
| }, |
| { |
| "epoch": 2.3086858249919278, |
| "grad_norm": 0.5320875644683838, |
| "learning_rate": 0.000462021333907984, |
| "loss": 3.5578, |
| "step": 21450 |
| }, |
| { |
| "epoch": 2.314067377031536, |
| "grad_norm": 0.5597154498100281, |
| "learning_rate": 0.0004616980928779226, |
| "loss": 3.5686, |
| "step": 21500 |
| }, |
| { |
| "epoch": 2.319448929071144, |
| "grad_norm": 0.6195567846298218, |
| "learning_rate": 0.0004613748518478612, |
| "loss": 3.5259, |
| "step": 21550 |
| }, |
| { |
| "epoch": 2.3248304811107525, |
| "grad_norm": 0.5434510111808777, |
| "learning_rate": 0.000461058075638401, |
| "loss": 3.5511, |
| "step": 21600 |
| }, |
| { |
| "epoch": 2.3302120331503606, |
| "grad_norm": 0.5677668452262878, |
| "learning_rate": 0.00046073483460833955, |
| "loss": 3.5598, |
| "step": 21650 |
| }, |
| { |
| "epoch": 2.3355935851899687, |
| "grad_norm": 0.5577279925346375, |
| "learning_rate": 0.00046041159357827814, |
| "loss": 3.551, |
| "step": 21700 |
| }, |
| { |
| "epoch": 2.340975137229577, |
| "grad_norm": 0.5741985440254211, |
| "learning_rate": 0.00046008835254821674, |
| "loss": 3.5495, |
| "step": 21750 |
| }, |
| { |
| "epoch": 2.3463566892691854, |
| "grad_norm": 0.5373184680938721, |
| "learning_rate": 0.0004597651115181554, |
| "loss": 3.559, |
| "step": 21800 |
| }, |
| { |
| "epoch": 2.3517382413087935, |
| "grad_norm": 0.6047030091285706, |
| "learning_rate": 0.00045944187048809393, |
| "loss": 3.5594, |
| "step": 21850 |
| }, |
| { |
| "epoch": 2.3571197933484016, |
| "grad_norm": 0.5527496933937073, |
| "learning_rate": 0.0004591186294580325, |
| "loss": 3.5631, |
| "step": 21900 |
| }, |
| { |
| "epoch": 2.3625013453880097, |
| "grad_norm": 0.6777981519699097, |
| "learning_rate": 0.00045879538842797106, |
| "loss": 3.5531, |
| "step": 21950 |
| }, |
| { |
| "epoch": 2.3678828974276183, |
| "grad_norm": 0.5738788843154907, |
| "learning_rate": 0.00045847214739790966, |
| "loss": 3.5472, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3678828974276183, |
| "eval_accuracy": 0.3643889725576488, |
| "eval_loss": 3.552778720855713, |
| "eval_runtime": 179.721, |
| "eval_samples_per_second": 100.216, |
| "eval_steps_per_second": 6.265, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3732644494672264, |
| "grad_norm": 0.5943243503570557, |
| "learning_rate": 0.0004581489063678482, |
| "loss": 3.5544, |
| "step": 22050 |
| }, |
| { |
| "epoch": 2.3786460015068345, |
| "grad_norm": 0.5966238379478455, |
| "learning_rate": 0.00045782566533778685, |
| "loss": 3.5685, |
| "step": 22100 |
| }, |
| { |
| "epoch": 2.384027553546443, |
| "grad_norm": 0.6006495952606201, |
| "learning_rate": 0.00045750242430772544, |
| "loss": 3.5509, |
| "step": 22150 |
| }, |
| { |
| "epoch": 2.389409105586051, |
| "grad_norm": 0.5800380706787109, |
| "learning_rate": 0.000457179183277664, |
| "loss": 3.5442, |
| "step": 22200 |
| }, |
| { |
| "epoch": 2.3947906576256592, |
| "grad_norm": 0.6179288029670715, |
| "learning_rate": 0.0004568559422476026, |
| "loss": 3.5578, |
| "step": 22250 |
| }, |
| { |
| "epoch": 2.4001722096652673, |
| "grad_norm": 0.5599116683006287, |
| "learning_rate": 0.00045653270121754117, |
| "loss": 3.5604, |
| "step": 22300 |
| }, |
| { |
| "epoch": 2.4055537617048754, |
| "grad_norm": 0.5807077884674072, |
| "learning_rate": 0.0004562094601874797, |
| "loss": 3.5322, |
| "step": 22350 |
| }, |
| { |
| "epoch": 2.410935313744484, |
| "grad_norm": 0.6106342077255249, |
| "learning_rate": 0.00045588621915741836, |
| "loss": 3.5576, |
| "step": 22400 |
| }, |
| { |
| "epoch": 2.416316865784092, |
| "grad_norm": 0.5440487861633301, |
| "learning_rate": 0.00045556297812735696, |
| "loss": 3.5573, |
| "step": 22450 |
| }, |
| { |
| "epoch": 2.4216984178237, |
| "grad_norm": 0.5883488655090332, |
| "learning_rate": 0.0004552397370972955, |
| "loss": 3.5533, |
| "step": 22500 |
| }, |
| { |
| "epoch": 2.4270799698633088, |
| "grad_norm": 0.5989765524864197, |
| "learning_rate": 0.0004549164960672341, |
| "loss": 3.5697, |
| "step": 22550 |
| }, |
| { |
| "epoch": 2.432461521902917, |
| "grad_norm": 0.5756821632385254, |
| "learning_rate": 0.00045459325503717263, |
| "loss": 3.5493, |
| "step": 22600 |
| }, |
| { |
| "epoch": 2.437843073942525, |
| "grad_norm": 0.5626195669174194, |
| "learning_rate": 0.0004542700140071113, |
| "loss": 3.5562, |
| "step": 22650 |
| }, |
| { |
| "epoch": 2.443224625982133, |
| "grad_norm": 0.5613085627555847, |
| "learning_rate": 0.0004539467729770499, |
| "loss": 3.5262, |
| "step": 22700 |
| }, |
| { |
| "epoch": 2.4486061780217416, |
| "grad_norm": 0.6236090064048767, |
| "learning_rate": 0.0004536235319469884, |
| "loss": 3.5649, |
| "step": 22750 |
| }, |
| { |
| "epoch": 2.4539877300613497, |
| "grad_norm": 0.5902277827262878, |
| "learning_rate": 0.000453300290916927, |
| "loss": 3.553, |
| "step": 22800 |
| }, |
| { |
| "epoch": 2.459369282100958, |
| "grad_norm": 0.6016800403594971, |
| "learning_rate": 0.0004529770498868656, |
| "loss": 3.5478, |
| "step": 22850 |
| }, |
| { |
| "epoch": 2.464750834140566, |
| "grad_norm": 0.6465705633163452, |
| "learning_rate": 0.00045265380885680414, |
| "loss": 3.5654, |
| "step": 22900 |
| }, |
| { |
| "epoch": 2.4701323861801745, |
| "grad_norm": 0.5794358849525452, |
| "learning_rate": 0.0004523305678267428, |
| "loss": 3.5521, |
| "step": 22950 |
| }, |
| { |
| "epoch": 2.4755139382197826, |
| "grad_norm": 0.5933077335357666, |
| "learning_rate": 0.0004520073267966814, |
| "loss": 3.5516, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4755139382197826, |
| "eval_accuracy": 0.3657163848880783, |
| "eval_loss": 3.5414011478424072, |
| "eval_runtime": 180.0044, |
| "eval_samples_per_second": 100.059, |
| "eval_steps_per_second": 6.255, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4808954902593907, |
| "grad_norm": 0.5735173225402832, |
| "learning_rate": 0.00045168408576661993, |
| "loss": 3.5382, |
| "step": 23050 |
| }, |
| { |
| "epoch": 2.4862770422989993, |
| "grad_norm": 0.5812858939170837, |
| "learning_rate": 0.0004513608447365585, |
| "loss": 3.5504, |
| "step": 23100 |
| }, |
| { |
| "epoch": 2.4916585943386074, |
| "grad_norm": 0.5932419300079346, |
| "learning_rate": 0.00045103760370649706, |
| "loss": 3.5628, |
| "step": 23150 |
| }, |
| { |
| "epoch": 2.4970401463782155, |
| "grad_norm": 0.5909202098846436, |
| "learning_rate": 0.00045071436267643566, |
| "loss": 3.5538, |
| "step": 23200 |
| }, |
| { |
| "epoch": 2.5024216984178236, |
| "grad_norm": 0.5628249645233154, |
| "learning_rate": 0.0004503911216463743, |
| "loss": 3.5681, |
| "step": 23250 |
| }, |
| { |
| "epoch": 2.5078032504574317, |
| "grad_norm": 0.5927748680114746, |
| "learning_rate": 0.00045006788061631285, |
| "loss": 3.5467, |
| "step": 23300 |
| }, |
| { |
| "epoch": 2.5131848024970402, |
| "grad_norm": 0.6498246192932129, |
| "learning_rate": 0.00044974463958625144, |
| "loss": 3.5468, |
| "step": 23350 |
| }, |
| { |
| "epoch": 2.5185663545366483, |
| "grad_norm": 0.5863044857978821, |
| "learning_rate": 0.00044942139855619004, |
| "loss": 3.5537, |
| "step": 23400 |
| }, |
| { |
| "epoch": 2.5239479065762565, |
| "grad_norm": 0.5949398279190063, |
| "learning_rate": 0.0004490981575261286, |
| "loss": 3.5465, |
| "step": 23450 |
| }, |
| { |
| "epoch": 2.529329458615865, |
| "grad_norm": 0.6202422380447388, |
| "learning_rate": 0.0004487749164960672, |
| "loss": 3.5471, |
| "step": 23500 |
| }, |
| { |
| "epoch": 2.534711010655473, |
| "grad_norm": 0.5920185446739197, |
| "learning_rate": 0.0004484516754660058, |
| "loss": 3.5452, |
| "step": 23550 |
| }, |
| { |
| "epoch": 2.540092562695081, |
| "grad_norm": 0.5713174939155579, |
| "learning_rate": 0.00044812843443594436, |
| "loss": 3.5476, |
| "step": 23600 |
| }, |
| { |
| "epoch": 2.5454741147346893, |
| "grad_norm": 0.5799892544746399, |
| "learning_rate": 0.00044781165822648417, |
| "loss": 3.5464, |
| "step": 23650 |
| }, |
| { |
| "epoch": 2.550855666774298, |
| "grad_norm": 0.5944963693618774, |
| "learning_rate": 0.00044748841719642276, |
| "loss": 3.5446, |
| "step": 23700 |
| }, |
| { |
| "epoch": 2.556237218813906, |
| "grad_norm": 0.5628204941749573, |
| "learning_rate": 0.00044716517616636136, |
| "loss": 3.5501, |
| "step": 23750 |
| }, |
| { |
| "epoch": 2.561618770853514, |
| "grad_norm": 0.5771909952163696, |
| "learning_rate": 0.0004468419351362999, |
| "loss": 3.5373, |
| "step": 23800 |
| }, |
| { |
| "epoch": 2.567000322893122, |
| "grad_norm": 0.5741766691207886, |
| "learning_rate": 0.0004465186941062385, |
| "loss": 3.5667, |
| "step": 23850 |
| }, |
| { |
| "epoch": 2.5723818749327307, |
| "grad_norm": 0.5425882339477539, |
| "learning_rate": 0.00044619545307617714, |
| "loss": 3.5707, |
| "step": 23900 |
| }, |
| { |
| "epoch": 2.577763426972339, |
| "grad_norm": 0.6335693001747131, |
| "learning_rate": 0.0004458722120461157, |
| "loss": 3.5424, |
| "step": 23950 |
| }, |
| { |
| "epoch": 2.583144979011947, |
| "grad_norm": 0.6267858743667603, |
| "learning_rate": 0.0004455489710160543, |
| "loss": 3.5437, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.583144979011947, |
| "eval_accuracy": 0.36688157845700076, |
| "eval_loss": 3.532123565673828, |
| "eval_runtime": 179.6516, |
| "eval_samples_per_second": 100.255, |
| "eval_steps_per_second": 6.268, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.5885265310515555, |
| "grad_norm": 0.6489107608795166, |
| "learning_rate": 0.0004452257299859928, |
| "loss": 3.5309, |
| "step": 24050 |
| }, |
| { |
| "epoch": 2.5939080830911636, |
| "grad_norm": 0.6115685105323792, |
| "learning_rate": 0.0004449024889559314, |
| "loss": 3.5566, |
| "step": 24100 |
| }, |
| { |
| "epoch": 2.5992896351307717, |
| "grad_norm": 0.5589924454689026, |
| "learning_rate": 0.00044457924792587, |
| "loss": 3.5505, |
| "step": 24150 |
| }, |
| { |
| "epoch": 2.60467118717038, |
| "grad_norm": 0.6251258254051208, |
| "learning_rate": 0.0004442560068958086, |
| "loss": 3.5464, |
| "step": 24200 |
| }, |
| { |
| "epoch": 2.610052739209988, |
| "grad_norm": 0.5544481873512268, |
| "learning_rate": 0.0004439327658657472, |
| "loss": 3.556, |
| "step": 24250 |
| }, |
| { |
| "epoch": 2.6154342912495965, |
| "grad_norm": 0.5777297616004944, |
| "learning_rate": 0.0004436095248356858, |
| "loss": 3.5515, |
| "step": 24300 |
| }, |
| { |
| "epoch": 2.6208158432892046, |
| "grad_norm": 0.6586969494819641, |
| "learning_rate": 0.00044328628380562433, |
| "loss": 3.5414, |
| "step": 24350 |
| }, |
| { |
| "epoch": 2.6261973953288127, |
| "grad_norm": 0.6207393407821655, |
| "learning_rate": 0.0004429630427755629, |
| "loss": 3.5585, |
| "step": 24400 |
| }, |
| { |
| "epoch": 2.6315789473684212, |
| "grad_norm": 0.5969880819320679, |
| "learning_rate": 0.0004426398017455016, |
| "loss": 3.5571, |
| "step": 24450 |
| }, |
| { |
| "epoch": 2.6369604994080293, |
| "grad_norm": 0.6068384051322937, |
| "learning_rate": 0.0004423165607154401, |
| "loss": 3.5417, |
| "step": 24500 |
| }, |
| { |
| "epoch": 2.6423420514476375, |
| "grad_norm": 0.582072913646698, |
| "learning_rate": 0.0004419933196853787, |
| "loss": 3.5233, |
| "step": 24550 |
| }, |
| { |
| "epoch": 2.6477236034872456, |
| "grad_norm": 0.6101149320602417, |
| "learning_rate": 0.00044167007865531725, |
| "loss": 3.5402, |
| "step": 24600 |
| }, |
| { |
| "epoch": 2.653105155526854, |
| "grad_norm": 0.6337757110595703, |
| "learning_rate": 0.00044134683762525584, |
| "loss": 3.5594, |
| "step": 24650 |
| }, |
| { |
| "epoch": 2.658486707566462, |
| "grad_norm": 0.6083040833473206, |
| "learning_rate": 0.00044102359659519444, |
| "loss": 3.5444, |
| "step": 24700 |
| }, |
| { |
| "epoch": 2.6638682596060703, |
| "grad_norm": 0.6233500242233276, |
| "learning_rate": 0.00044070035556513303, |
| "loss": 3.5584, |
| "step": 24750 |
| }, |
| { |
| "epoch": 2.6692498116456784, |
| "grad_norm": 0.5510074496269226, |
| "learning_rate": 0.00044037711453507163, |
| "loss": 3.5389, |
| "step": 24800 |
| }, |
| { |
| "epoch": 2.674631363685287, |
| "grad_norm": 0.6422367095947266, |
| "learning_rate": 0.0004400538735050102, |
| "loss": 3.5267, |
| "step": 24850 |
| }, |
| { |
| "epoch": 2.680012915724895, |
| "grad_norm": 0.5510942339897156, |
| "learning_rate": 0.00043973063247494876, |
| "loss": 3.5431, |
| "step": 24900 |
| }, |
| { |
| "epoch": 2.685394467764503, |
| "grad_norm": 0.6150233149528503, |
| "learning_rate": 0.00043940739144488736, |
| "loss": 3.5392, |
| "step": 24950 |
| }, |
| { |
| "epoch": 2.6907760198041117, |
| "grad_norm": 0.5818309783935547, |
| "learning_rate": 0.0004390841504148259, |
| "loss": 3.5452, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.6907760198041117, |
| "eval_accuracy": 0.36736758282419496, |
| "eval_loss": 3.5232725143432617, |
| "eval_runtime": 180.1337, |
| "eval_samples_per_second": 99.987, |
| "eval_steps_per_second": 6.251, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.69615757184372, |
| "grad_norm": 0.5925581455230713, |
| "learning_rate": 0.00043876090938476455, |
| "loss": 3.5541, |
| "step": 25050 |
| }, |
| { |
| "epoch": 2.701539123883328, |
| "grad_norm": 0.5792503952980042, |
| "learning_rate": 0.00043843766835470314, |
| "loss": 3.5547, |
| "step": 25100 |
| }, |
| { |
| "epoch": 2.706920675922936, |
| "grad_norm": 0.6317951083183289, |
| "learning_rate": 0.0004381144273246417, |
| "loss": 3.5403, |
| "step": 25150 |
| }, |
| { |
| "epoch": 2.712302227962544, |
| "grad_norm": 0.5946325659751892, |
| "learning_rate": 0.0004377911862945803, |
| "loss": 3.5453, |
| "step": 25200 |
| }, |
| { |
| "epoch": 2.7176837800021527, |
| "grad_norm": 0.59228515625, |
| "learning_rate": 0.00043746794526451887, |
| "loss": 3.5341, |
| "step": 25250 |
| }, |
| { |
| "epoch": 2.723065332041761, |
| "grad_norm": 0.5823553800582886, |
| "learning_rate": 0.00043714470423445747, |
| "loss": 3.5449, |
| "step": 25300 |
| }, |
| { |
| "epoch": 2.728446884081369, |
| "grad_norm": 0.6187898516654968, |
| "learning_rate": 0.00043682146320439606, |
| "loss": 3.5427, |
| "step": 25350 |
| }, |
| { |
| "epoch": 2.7338284361209775, |
| "grad_norm": 0.6080958247184753, |
| "learning_rate": 0.00043649822217433466, |
| "loss": 3.5392, |
| "step": 25400 |
| }, |
| { |
| "epoch": 2.7392099881605856, |
| "grad_norm": 0.5838017463684082, |
| "learning_rate": 0.0004361749811442732, |
| "loss": 3.5302, |
| "step": 25450 |
| }, |
| { |
| "epoch": 2.7445915402001937, |
| "grad_norm": 0.5467104911804199, |
| "learning_rate": 0.0004358517401142118, |
| "loss": 3.5391, |
| "step": 25500 |
| }, |
| { |
| "epoch": 2.749973092239802, |
| "grad_norm": 0.5581143498420715, |
| "learning_rate": 0.00043552849908415033, |
| "loss": 3.5437, |
| "step": 25550 |
| }, |
| { |
| "epoch": 2.7553546442794103, |
| "grad_norm": 0.5552875399589539, |
| "learning_rate": 0.000435205258054089, |
| "loss": 3.5394, |
| "step": 25600 |
| }, |
| { |
| "epoch": 2.7607361963190185, |
| "grad_norm": 0.6552839875221252, |
| "learning_rate": 0.00043488848184462873, |
| "loss": 3.5327, |
| "step": 25650 |
| }, |
| { |
| "epoch": 2.7661177483586266, |
| "grad_norm": 0.648295521736145, |
| "learning_rate": 0.0004345652408145674, |
| "loss": 3.5352, |
| "step": 25700 |
| }, |
| { |
| "epoch": 2.7714993003982347, |
| "grad_norm": 0.6237471699714661, |
| "learning_rate": 0.000434241999784506, |
| "loss": 3.5333, |
| "step": 25750 |
| }, |
| { |
| "epoch": 2.776880852437843, |
| "grad_norm": 0.5941251516342163, |
| "learning_rate": 0.0004339187587544445, |
| "loss": 3.5268, |
| "step": 25800 |
| }, |
| { |
| "epoch": 2.7822624044774513, |
| "grad_norm": 0.6055666208267212, |
| "learning_rate": 0.0004335955177243831, |
| "loss": 3.5436, |
| "step": 25850 |
| }, |
| { |
| "epoch": 2.7876439565170594, |
| "grad_norm": 0.6202614903450012, |
| "learning_rate": 0.00043327227669432165, |
| "loss": 3.5259, |
| "step": 25900 |
| }, |
| { |
| "epoch": 2.793025508556668, |
| "grad_norm": 0.579383909702301, |
| "learning_rate": 0.00043294903566426025, |
| "loss": 3.5166, |
| "step": 25950 |
| }, |
| { |
| "epoch": 2.798407060596276, |
| "grad_norm": 0.5871159434318542, |
| "learning_rate": 0.0004326257946341989, |
| "loss": 3.5385, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.798407060596276, |
| "eval_accuracy": 0.3681571633599428, |
| "eval_loss": 3.51601505279541, |
| "eval_runtime": 179.8655, |
| "eval_samples_per_second": 100.136, |
| "eval_steps_per_second": 6.26, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.803788612635884, |
| "grad_norm": 0.6053968071937561, |
| "learning_rate": 0.00043230255360413744, |
| "loss": 3.5358, |
| "step": 26050 |
| }, |
| { |
| "epoch": 2.8091701646754923, |
| "grad_norm": 0.5837214589118958, |
| "learning_rate": 0.00043197931257407603, |
| "loss": 3.5295, |
| "step": 26100 |
| }, |
| { |
| "epoch": 2.8145517167151004, |
| "grad_norm": 0.6265986561775208, |
| "learning_rate": 0.0004316560715440146, |
| "loss": 3.5275, |
| "step": 26150 |
| }, |
| { |
| "epoch": 2.819933268754709, |
| "grad_norm": 0.6679002642631531, |
| "learning_rate": 0.00043133929533455443, |
| "loss": 3.534, |
| "step": 26200 |
| }, |
| { |
| "epoch": 2.825314820794317, |
| "grad_norm": 0.6008105874061584, |
| "learning_rate": 0.00043101605430449297, |
| "loss": 3.5266, |
| "step": 26250 |
| }, |
| { |
| "epoch": 2.830696372833925, |
| "grad_norm": 0.6022807955741882, |
| "learning_rate": 0.00043069281327443157, |
| "loss": 3.5404, |
| "step": 26300 |
| }, |
| { |
| "epoch": 2.8360779248735337, |
| "grad_norm": 0.5778976678848267, |
| "learning_rate": 0.0004303695722443702, |
| "loss": 3.5387, |
| "step": 26350 |
| }, |
| { |
| "epoch": 2.841459476913142, |
| "grad_norm": 0.6605262160301208, |
| "learning_rate": 0.00043004633121430876, |
| "loss": 3.5403, |
| "step": 26400 |
| }, |
| { |
| "epoch": 2.84684102895275, |
| "grad_norm": 0.5825318098068237, |
| "learning_rate": 0.00042972309018424735, |
| "loss": 3.5128, |
| "step": 26450 |
| }, |
| { |
| "epoch": 2.852222580992358, |
| "grad_norm": 0.612047553062439, |
| "learning_rate": 0.00042939984915418594, |
| "loss": 3.54, |
| "step": 26500 |
| }, |
| { |
| "epoch": 2.857604133031966, |
| "grad_norm": 0.6171521544456482, |
| "learning_rate": 0.0004290766081241245, |
| "loss": 3.5308, |
| "step": 26550 |
| }, |
| { |
| "epoch": 2.8629856850715747, |
| "grad_norm": 0.5376119017601013, |
| "learning_rate": 0.0004287533670940631, |
| "loss": 3.5204, |
| "step": 26600 |
| }, |
| { |
| "epoch": 2.868367237111183, |
| "grad_norm": 0.6116226315498352, |
| "learning_rate": 0.00042843012606400173, |
| "loss": 3.5264, |
| "step": 26650 |
| }, |
| { |
| "epoch": 2.873748789150791, |
| "grad_norm": 0.5715238451957703, |
| "learning_rate": 0.00042810688503394027, |
| "loss": 3.5308, |
| "step": 26700 |
| }, |
| { |
| "epoch": 2.8791303411903995, |
| "grad_norm": 0.5794273018836975, |
| "learning_rate": 0.00042778364400387886, |
| "loss": 3.5252, |
| "step": 26750 |
| }, |
| { |
| "epoch": 2.8845118932300076, |
| "grad_norm": 0.6008096933364868, |
| "learning_rate": 0.0004274604029738174, |
| "loss": 3.5288, |
| "step": 26800 |
| }, |
| { |
| "epoch": 2.8898934452696157, |
| "grad_norm": 0.6101154088973999, |
| "learning_rate": 0.000427137161943756, |
| "loss": 3.5337, |
| "step": 26850 |
| }, |
| { |
| "epoch": 2.895274997309224, |
| "grad_norm": 0.5782980918884277, |
| "learning_rate": 0.00042681392091369465, |
| "loss": 3.531, |
| "step": 26900 |
| }, |
| { |
| "epoch": 2.9006565493488323, |
| "grad_norm": 0.5600022077560425, |
| "learning_rate": 0.0004264906798836332, |
| "loss": 3.509, |
| "step": 26950 |
| }, |
| { |
| "epoch": 2.9060381013884404, |
| "grad_norm": 0.5735082030296326, |
| "learning_rate": 0.0004261674388535718, |
| "loss": 3.5233, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9060381013884404, |
| "eval_accuracy": 0.3694485029316179, |
| "eval_loss": 3.506122589111328, |
| "eval_runtime": 179.9881, |
| "eval_samples_per_second": 100.068, |
| "eval_steps_per_second": 6.256, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9114196534280485, |
| "grad_norm": 0.6081140041351318, |
| "learning_rate": 0.0004258441978235104, |
| "loss": 3.5229, |
| "step": 27050 |
| }, |
| { |
| "epoch": 2.9168012054676566, |
| "grad_norm": 0.591622531414032, |
| "learning_rate": 0.0004255209567934489, |
| "loss": 3.5366, |
| "step": 27100 |
| }, |
| { |
| "epoch": 2.922182757507265, |
| "grad_norm": 0.5882881283760071, |
| "learning_rate": 0.0004251977157633875, |
| "loss": 3.528, |
| "step": 27150 |
| }, |
| { |
| "epoch": 2.9275643095468733, |
| "grad_norm": 0.6217557191848755, |
| "learning_rate": 0.00042487447473332616, |
| "loss": 3.5435, |
| "step": 27200 |
| }, |
| { |
| "epoch": 2.9329458615864814, |
| "grad_norm": 0.5857243537902832, |
| "learning_rate": 0.0004245512337032647, |
| "loss": 3.514, |
| "step": 27250 |
| }, |
| { |
| "epoch": 2.93832741362609, |
| "grad_norm": 0.5875625610351562, |
| "learning_rate": 0.0004242279926732033, |
| "loss": 3.5184, |
| "step": 27300 |
| }, |
| { |
| "epoch": 2.943708965665698, |
| "grad_norm": 0.5514461398124695, |
| "learning_rate": 0.00042390475164314184, |
| "loss": 3.5367, |
| "step": 27350 |
| }, |
| { |
| "epoch": 2.949090517705306, |
| "grad_norm": 0.5746405124664307, |
| "learning_rate": 0.00042358151061308043, |
| "loss": 3.5174, |
| "step": 27400 |
| }, |
| { |
| "epoch": 2.9544720697449143, |
| "grad_norm": 0.5785201191902161, |
| "learning_rate": 0.000423258269583019, |
| "loss": 3.5364, |
| "step": 27450 |
| }, |
| { |
| "epoch": 2.9598536217845224, |
| "grad_norm": 0.6101451516151428, |
| "learning_rate": 0.0004229350285529576, |
| "loss": 3.5128, |
| "step": 27500 |
| }, |
| { |
| "epoch": 2.965235173824131, |
| "grad_norm": 0.607959508895874, |
| "learning_rate": 0.0004226117875228962, |
| "loss": 3.5115, |
| "step": 27550 |
| }, |
| { |
| "epoch": 2.970616725863739, |
| "grad_norm": 0.5360074043273926, |
| "learning_rate": 0.0004222885464928348, |
| "loss": 3.5161, |
| "step": 27600 |
| }, |
| { |
| "epoch": 2.975998277903347, |
| "grad_norm": 0.6007967591285706, |
| "learning_rate": 0.00042196530546277335, |
| "loss": 3.5294, |
| "step": 27650 |
| }, |
| { |
| "epoch": 2.9813798299429557, |
| "grad_norm": 0.679694652557373, |
| "learning_rate": 0.00042164206443271195, |
| "loss": 3.5162, |
| "step": 27700 |
| }, |
| { |
| "epoch": 2.986761381982564, |
| "grad_norm": 0.589306116104126, |
| "learning_rate": 0.0004213188234026505, |
| "loss": 3.5213, |
| "step": 27750 |
| }, |
| { |
| "epoch": 2.992142934022172, |
| "grad_norm": 0.5684702396392822, |
| "learning_rate": 0.00042099558237258914, |
| "loss": 3.5406, |
| "step": 27800 |
| }, |
| { |
| "epoch": 2.9975244860617805, |
| "grad_norm": 0.6052414774894714, |
| "learning_rate": 0.00042067234134252773, |
| "loss": 3.5357, |
| "step": 27850 |
| }, |
| { |
| "epoch": 3.0029060381013886, |
| "grad_norm": 0.5678727030754089, |
| "learning_rate": 0.00042034910031246627, |
| "loss": 3.486, |
| "step": 27900 |
| }, |
| { |
| "epoch": 3.0082875901409967, |
| "grad_norm": 0.6087445020675659, |
| "learning_rate": 0.00042002585928240486, |
| "loss": 3.4427, |
| "step": 27950 |
| }, |
| { |
| "epoch": 3.0136691421806048, |
| "grad_norm": 0.6876854300498962, |
| "learning_rate": 0.00041970261825234346, |
| "loss": 3.4446, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.0136691421806048, |
| "eval_accuracy": 0.37010759134910054, |
| "eval_loss": 3.5012426376342773, |
| "eval_runtime": 179.7175, |
| "eval_samples_per_second": 100.218, |
| "eval_steps_per_second": 6.265, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.0190506942202133, |
| "grad_norm": 0.5857146978378296, |
| "learning_rate": 0.00041937937722228205, |
| "loss": 3.4242, |
| "step": 28050 |
| }, |
| { |
| "epoch": 3.0244322462598214, |
| "grad_norm": 0.6560085415840149, |
| "learning_rate": 0.00041905613619222065, |
| "loss": 3.4337, |
| "step": 28100 |
| }, |
| { |
| "epoch": 3.0298137982994295, |
| "grad_norm": 0.6381294131278992, |
| "learning_rate": 0.00041873289516215924, |
| "loss": 3.4334, |
| "step": 28150 |
| }, |
| { |
| "epoch": 3.0351953503390376, |
| "grad_norm": 0.6040878891944885, |
| "learning_rate": 0.0004184096541320978, |
| "loss": 3.4403, |
| "step": 28200 |
| }, |
| { |
| "epoch": 3.040576902378646, |
| "grad_norm": 0.5884063839912415, |
| "learning_rate": 0.0004180864131020364, |
| "loss": 3.4367, |
| "step": 28250 |
| }, |
| { |
| "epoch": 3.0459584544182543, |
| "grad_norm": 0.574171781539917, |
| "learning_rate": 0.0004177631720719749, |
| "loss": 3.4463, |
| "step": 28300 |
| }, |
| { |
| "epoch": 3.0513400064578624, |
| "grad_norm": 0.6166306734085083, |
| "learning_rate": 0.00041743993104191357, |
| "loss": 3.4219, |
| "step": 28350 |
| }, |
| { |
| "epoch": 3.0567215584974705, |
| "grad_norm": 0.5581924915313721, |
| "learning_rate": 0.00041711669001185216, |
| "loss": 3.4681, |
| "step": 28400 |
| }, |
| { |
| "epoch": 3.062103110537079, |
| "grad_norm": 0.5798004269599915, |
| "learning_rate": 0.0004167934489817907, |
| "loss": 3.402, |
| "step": 28450 |
| }, |
| { |
| "epoch": 3.067484662576687, |
| "grad_norm": 0.5792421102523804, |
| "learning_rate": 0.0004164702079517293, |
| "loss": 3.4292, |
| "step": 28500 |
| }, |
| { |
| "epoch": 3.0728662146162953, |
| "grad_norm": 0.5974885821342468, |
| "learning_rate": 0.0004161469669216679, |
| "loss": 3.4502, |
| "step": 28550 |
| }, |
| { |
| "epoch": 3.0782477666559034, |
| "grad_norm": 0.6063544750213623, |
| "learning_rate": 0.00041582372589160643, |
| "loss": 3.4676, |
| "step": 28600 |
| }, |
| { |
| "epoch": 3.083629318695512, |
| "grad_norm": 0.6233583688735962, |
| "learning_rate": 0.0004155004848615451, |
| "loss": 3.4665, |
| "step": 28650 |
| }, |
| { |
| "epoch": 3.08901087073512, |
| "grad_norm": 0.6059950590133667, |
| "learning_rate": 0.0004151772438314837, |
| "loss": 3.4411, |
| "step": 28700 |
| }, |
| { |
| "epoch": 3.094392422774728, |
| "grad_norm": 0.634493887424469, |
| "learning_rate": 0.0004148540028014222, |
| "loss": 3.4556, |
| "step": 28750 |
| }, |
| { |
| "epoch": 3.0997739748143363, |
| "grad_norm": 0.6094062328338623, |
| "learning_rate": 0.0004145307617713608, |
| "loss": 3.4545, |
| "step": 28800 |
| }, |
| { |
| "epoch": 3.105155526853945, |
| "grad_norm": 0.6499791145324707, |
| "learning_rate": 0.00041420752074129935, |
| "loss": 3.4478, |
| "step": 28850 |
| }, |
| { |
| "epoch": 3.110537078893553, |
| "grad_norm": 0.5969070196151733, |
| "learning_rate": 0.000413884279711238, |
| "loss": 3.4373, |
| "step": 28900 |
| }, |
| { |
| "epoch": 3.115918630933161, |
| "grad_norm": 0.5845366716384888, |
| "learning_rate": 0.0004135610386811766, |
| "loss": 3.4629, |
| "step": 28950 |
| }, |
| { |
| "epoch": 3.121300182972769, |
| "grad_norm": 0.609656035900116, |
| "learning_rate": 0.00041323779765111514, |
| "loss": 3.4549, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.121300182972769, |
| "eval_accuracy": 0.37079699392228344, |
| "eval_loss": 3.4950499534606934, |
| "eval_runtime": 179.9053, |
| "eval_samples_per_second": 100.114, |
| "eval_steps_per_second": 6.259, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.1266817350123777, |
| "grad_norm": 0.5781153440475464, |
| "learning_rate": 0.00041291455662105373, |
| "loss": 3.4505, |
| "step": 29050 |
| }, |
| { |
| "epoch": 3.132063287051986, |
| "grad_norm": 0.6373555064201355, |
| "learning_rate": 0.0004125913155909923, |
| "loss": 3.4634, |
| "step": 29100 |
| }, |
| { |
| "epoch": 3.137444839091594, |
| "grad_norm": 0.6015342473983765, |
| "learning_rate": 0.00041226807456093087, |
| "loss": 3.4469, |
| "step": 29150 |
| }, |
| { |
| "epoch": 3.1428263911312024, |
| "grad_norm": 0.644429624080658, |
| "learning_rate": 0.0004119448335308695, |
| "loss": 3.4523, |
| "step": 29200 |
| }, |
| { |
| "epoch": 3.1482079431708105, |
| "grad_norm": 0.5962895750999451, |
| "learning_rate": 0.0004116215925008081, |
| "loss": 3.4383, |
| "step": 29250 |
| }, |
| { |
| "epoch": 3.1535894952104186, |
| "grad_norm": 0.6676060557365417, |
| "learning_rate": 0.00041129835147074665, |
| "loss": 3.4608, |
| "step": 29300 |
| }, |
| { |
| "epoch": 3.1589710472500268, |
| "grad_norm": 0.6141141057014465, |
| "learning_rate": 0.00041097511044068524, |
| "loss": 3.4517, |
| "step": 29350 |
| }, |
| { |
| "epoch": 3.1643525992896353, |
| "grad_norm": 0.5669366121292114, |
| "learning_rate": 0.0004106518694106238, |
| "loss": 3.4687, |
| "step": 29400 |
| }, |
| { |
| "epoch": 3.1697341513292434, |
| "grad_norm": 0.5981097221374512, |
| "learning_rate": 0.0004103286283805624, |
| "loss": 3.4496, |
| "step": 29450 |
| }, |
| { |
| "epoch": 3.1751157033688515, |
| "grad_norm": 0.5925617814064026, |
| "learning_rate": 0.0004100118521711022, |
| "loss": 3.4593, |
| "step": 29500 |
| }, |
| { |
| "epoch": 3.1804972554084596, |
| "grad_norm": 0.5993756055831909, |
| "learning_rate": 0.0004096886111410408, |
| "loss": 3.4616, |
| "step": 29550 |
| }, |
| { |
| "epoch": 3.185878807448068, |
| "grad_norm": 0.6227186322212219, |
| "learning_rate": 0.00040936537011097943, |
| "loss": 3.46, |
| "step": 29600 |
| }, |
| { |
| "epoch": 3.1912603594876763, |
| "grad_norm": 0.6608783006668091, |
| "learning_rate": 0.00040904212908091797, |
| "loss": 3.4731, |
| "step": 29650 |
| }, |
| { |
| "epoch": 3.1966419115272844, |
| "grad_norm": 0.6810404062271118, |
| "learning_rate": 0.00040871888805085656, |
| "loss": 3.4522, |
| "step": 29700 |
| }, |
| { |
| "epoch": 3.2020234635668925, |
| "grad_norm": 0.6131249666213989, |
| "learning_rate": 0.0004083956470207951, |
| "loss": 3.4594, |
| "step": 29750 |
| }, |
| { |
| "epoch": 3.207405015606501, |
| "grad_norm": 0.6485390663146973, |
| "learning_rate": 0.0004080724059907337, |
| "loss": 3.4409, |
| "step": 29800 |
| }, |
| { |
| "epoch": 3.212786567646109, |
| "grad_norm": 0.6126716136932373, |
| "learning_rate": 0.00040774916496067235, |
| "loss": 3.4303, |
| "step": 29850 |
| }, |
| { |
| "epoch": 3.2181681196857173, |
| "grad_norm": 0.626392662525177, |
| "learning_rate": 0.0004074259239306109, |
| "loss": 3.4539, |
| "step": 29900 |
| }, |
| { |
| "epoch": 3.2235496717253254, |
| "grad_norm": 0.5780983567237854, |
| "learning_rate": 0.0004071026829005495, |
| "loss": 3.4479, |
| "step": 29950 |
| }, |
| { |
| "epoch": 3.228931223764934, |
| "grad_norm": 0.5982878804206848, |
| "learning_rate": 0.0004067794418704881, |
| "loss": 3.445, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.228931223764934, |
| "eval_accuracy": 0.37137198500459767, |
| "eval_loss": 3.49147367477417, |
| "eval_runtime": 179.7703, |
| "eval_samples_per_second": 100.189, |
| "eval_steps_per_second": 6.264, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.234312775804542, |
| "grad_norm": 0.5977009534835815, |
| "learning_rate": 0.0004064562008404266, |
| "loss": 3.4542, |
| "step": 30050 |
| }, |
| { |
| "epoch": 3.23969432784415, |
| "grad_norm": 0.5981681942939758, |
| "learning_rate": 0.0004061329598103652, |
| "loss": 3.4456, |
| "step": 30100 |
| }, |
| { |
| "epoch": 3.2450758798837587, |
| "grad_norm": 0.5974931120872498, |
| "learning_rate": 0.00040580971878030386, |
| "loss": 3.4531, |
| "step": 30150 |
| }, |
| { |
| "epoch": 3.250457431923367, |
| "grad_norm": 0.6247798800468445, |
| "learning_rate": 0.0004054864777502424, |
| "loss": 3.4451, |
| "step": 30200 |
| }, |
| { |
| "epoch": 3.255838983962975, |
| "grad_norm": 0.6065713167190552, |
| "learning_rate": 0.000405163236720181, |
| "loss": 3.4398, |
| "step": 30250 |
| }, |
| { |
| "epoch": 3.261220536002583, |
| "grad_norm": 0.5848674774169922, |
| "learning_rate": 0.00040483999569011954, |
| "loss": 3.4452, |
| "step": 30300 |
| }, |
| { |
| "epoch": 3.2666020880421915, |
| "grad_norm": 0.6002523899078369, |
| "learning_rate": 0.00040451675466005813, |
| "loss": 3.4577, |
| "step": 30350 |
| }, |
| { |
| "epoch": 3.2719836400817996, |
| "grad_norm": 0.6616047024726868, |
| "learning_rate": 0.0004041935136299967, |
| "loss": 3.4653, |
| "step": 30400 |
| }, |
| { |
| "epoch": 3.2773651921214078, |
| "grad_norm": 0.610581636428833, |
| "learning_rate": 0.0004038702725999353, |
| "loss": 3.4604, |
| "step": 30450 |
| }, |
| { |
| "epoch": 3.282746744161016, |
| "grad_norm": 0.6293004155158997, |
| "learning_rate": 0.0004035470315698739, |
| "loss": 3.4615, |
| "step": 30500 |
| }, |
| { |
| "epoch": 3.2881282962006244, |
| "grad_norm": 0.5750092267990112, |
| "learning_rate": 0.0004032237905398125, |
| "loss": 3.4553, |
| "step": 30550 |
| }, |
| { |
| "epoch": 3.2935098482402325, |
| "grad_norm": 0.5964309573173523, |
| "learning_rate": 0.00040290054950975105, |
| "loss": 3.4681, |
| "step": 30600 |
| }, |
| { |
| "epoch": 3.2988914002798406, |
| "grad_norm": 0.5653200149536133, |
| "learning_rate": 0.00040257730847968965, |
| "loss": 3.455, |
| "step": 30650 |
| }, |
| { |
| "epoch": 3.304272952319449, |
| "grad_norm": 0.5926148891448975, |
| "learning_rate": 0.0004022540674496283, |
| "loss": 3.4585, |
| "step": 30700 |
| }, |
| { |
| "epoch": 3.3096545043590573, |
| "grad_norm": 0.6698280572891235, |
| "learning_rate": 0.00040193082641956684, |
| "loss": 3.4393, |
| "step": 30750 |
| }, |
| { |
| "epoch": 3.3150360563986654, |
| "grad_norm": 0.6018702983856201, |
| "learning_rate": 0.00040160758538950543, |
| "loss": 3.4479, |
| "step": 30800 |
| }, |
| { |
| "epoch": 3.3204176084382735, |
| "grad_norm": 0.6161127090454102, |
| "learning_rate": 0.00040128434435944397, |
| "loss": 3.4725, |
| "step": 30850 |
| }, |
| { |
| "epoch": 3.3257991604778816, |
| "grad_norm": 0.6327736973762512, |
| "learning_rate": 0.00040096110332938257, |
| "loss": 3.4554, |
| "step": 30900 |
| }, |
| { |
| "epoch": 3.33118071251749, |
| "grad_norm": 0.55366051197052, |
| "learning_rate": 0.00040063786229932116, |
| "loss": 3.4484, |
| "step": 30950 |
| }, |
| { |
| "epoch": 3.3365622645570983, |
| "grad_norm": 0.6205126643180847, |
| "learning_rate": 0.00040031462126925975, |
| "loss": 3.4523, |
| "step": 31000 |
| }, |
| { |
| "epoch": 3.3365622645570983, |
| "eval_accuracy": 0.3720559549333898, |
| "eval_loss": 3.4844534397125244, |
| "eval_runtime": 180.3473, |
| "eval_samples_per_second": 99.868, |
| "eval_steps_per_second": 6.244, |
| "step": 31000 |
| }, |
| { |
| "epoch": 3.3419438165967064, |
| "grad_norm": 0.6246373057365417, |
| "learning_rate": 0.00039999138023919835, |
| "loss": 3.455, |
| "step": 31050 |
| }, |
| { |
| "epoch": 3.347325368636315, |
| "grad_norm": 0.5978841781616211, |
| "learning_rate": 0.00039966813920913694, |
| "loss": 3.4602, |
| "step": 31100 |
| }, |
| { |
| "epoch": 3.352706920675923, |
| "grad_norm": 0.6038092374801636, |
| "learning_rate": 0.0003993448981790755, |
| "loss": 3.4566, |
| "step": 31150 |
| }, |
| { |
| "epoch": 3.358088472715531, |
| "grad_norm": 0.6315546631813049, |
| "learning_rate": 0.0003990216571490141, |
| "loss": 3.4436, |
| "step": 31200 |
| }, |
| { |
| "epoch": 3.3634700247551392, |
| "grad_norm": 0.6053710579872131, |
| "learning_rate": 0.0003986984161189526, |
| "loss": 3.4482, |
| "step": 31250 |
| }, |
| { |
| "epoch": 3.368851576794748, |
| "grad_norm": 0.6105901598930359, |
| "learning_rate": 0.00039837517508889127, |
| "loss": 3.4674, |
| "step": 31300 |
| }, |
| { |
| "epoch": 3.374233128834356, |
| "grad_norm": 0.5860775113105774, |
| "learning_rate": 0.00039805193405882986, |
| "loss": 3.4504, |
| "step": 31350 |
| }, |
| { |
| "epoch": 3.379614680873964, |
| "grad_norm": 0.6133190989494324, |
| "learning_rate": 0.0003977286930287684, |
| "loss": 3.455, |
| "step": 31400 |
| }, |
| { |
| "epoch": 3.384996232913572, |
| "grad_norm": 0.6758852005004883, |
| "learning_rate": 0.000397405451998707, |
| "loss": 3.4575, |
| "step": 31450 |
| }, |
| { |
| "epoch": 3.3903777849531807, |
| "grad_norm": 0.5996787548065186, |
| "learning_rate": 0.00039708221096864554, |
| "loss": 3.4638, |
| "step": 31500 |
| }, |
| { |
| "epoch": 3.3957593369927888, |
| "grad_norm": 0.6079868674278259, |
| "learning_rate": 0.00039675896993858413, |
| "loss": 3.4638, |
| "step": 31550 |
| }, |
| { |
| "epoch": 3.401140889032397, |
| "grad_norm": 0.5919036865234375, |
| "learning_rate": 0.0003964357289085228, |
| "loss": 3.4436, |
| "step": 31600 |
| }, |
| { |
| "epoch": 3.4065224410720054, |
| "grad_norm": 0.6195621490478516, |
| "learning_rate": 0.0003961124878784613, |
| "loss": 3.4653, |
| "step": 31650 |
| }, |
| { |
| "epoch": 3.4119039931116135, |
| "grad_norm": 0.6306630969047546, |
| "learning_rate": 0.0003957892468483999, |
| "loss": 3.4508, |
| "step": 31700 |
| }, |
| { |
| "epoch": 3.4172855451512216, |
| "grad_norm": 0.6084063053131104, |
| "learning_rate": 0.0003954660058183385, |
| "loss": 3.4406, |
| "step": 31750 |
| }, |
| { |
| "epoch": 3.4226670971908297, |
| "grad_norm": 0.6287810206413269, |
| "learning_rate": 0.00039514276478827705, |
| "loss": 3.4555, |
| "step": 31800 |
| }, |
| { |
| "epoch": 3.428048649230438, |
| "grad_norm": 0.598998486995697, |
| "learning_rate": 0.0003948195237582157, |
| "loss": 3.4591, |
| "step": 31850 |
| }, |
| { |
| "epoch": 3.4334302012700464, |
| "grad_norm": 0.6260305047035217, |
| "learning_rate": 0.0003944962827281543, |
| "loss": 3.4522, |
| "step": 31900 |
| }, |
| { |
| "epoch": 3.4388117533096545, |
| "grad_norm": 0.6597923040390015, |
| "learning_rate": 0.00039417304169809284, |
| "loss": 3.4599, |
| "step": 31950 |
| }, |
| { |
| "epoch": 3.4441933053492626, |
| "grad_norm": 0.6098554134368896, |
| "learning_rate": 0.00039384980066803143, |
| "loss": 3.4609, |
| "step": 32000 |
| }, |
| { |
| "epoch": 3.4441933053492626, |
| "eval_accuracy": 0.37268233883164026, |
| "eval_loss": 3.4774160385131836, |
| "eval_runtime": 179.8913, |
| "eval_samples_per_second": 100.122, |
| "eval_steps_per_second": 6.259, |
| "step": 32000 |
| }, |
| { |
| "epoch": 3.449574857388871, |
| "grad_norm": 0.574020504951477, |
| "learning_rate": 0.00039352655963796997, |
| "loss": 3.4806, |
| "step": 32050 |
| }, |
| { |
| "epoch": 3.4549564094284793, |
| "grad_norm": 0.6245083808898926, |
| "learning_rate": 0.00039320331860790857, |
| "loss": 3.4747, |
| "step": 32100 |
| }, |
| { |
| "epoch": 3.4603379614680874, |
| "grad_norm": 0.5653278827667236, |
| "learning_rate": 0.0003928800775778472, |
| "loss": 3.4557, |
| "step": 32150 |
| }, |
| { |
| "epoch": 3.4657195135076955, |
| "grad_norm": 0.6541618704795837, |
| "learning_rate": 0.00039255683654778576, |
| "loss": 3.4788, |
| "step": 32200 |
| }, |
| { |
| "epoch": 3.471101065547304, |
| "grad_norm": 0.5819790363311768, |
| "learning_rate": 0.00039223359551772435, |
| "loss": 3.4577, |
| "step": 32250 |
| }, |
| { |
| "epoch": 3.476482617586912, |
| "grad_norm": 0.612333357334137, |
| "learning_rate": 0.00039191035448766294, |
| "loss": 3.4452, |
| "step": 32300 |
| }, |
| { |
| "epoch": 3.4818641696265202, |
| "grad_norm": 0.616150975227356, |
| "learning_rate": 0.0003915871134576015, |
| "loss": 3.445, |
| "step": 32350 |
| }, |
| { |
| "epoch": 3.4872457216661283, |
| "grad_norm": 0.6523071527481079, |
| "learning_rate": 0.0003912638724275401, |
| "loss": 3.4603, |
| "step": 32400 |
| }, |
| { |
| "epoch": 3.492627273705737, |
| "grad_norm": 0.6022084951400757, |
| "learning_rate": 0.00039094063139747873, |
| "loss": 3.4648, |
| "step": 32450 |
| }, |
| { |
| "epoch": 3.498008825745345, |
| "grad_norm": 0.6120687127113342, |
| "learning_rate": 0.00039061739036741727, |
| "loss": 3.4401, |
| "step": 32500 |
| }, |
| { |
| "epoch": 3.503390377784953, |
| "grad_norm": 0.6045283675193787, |
| "learning_rate": 0.00039029414933735586, |
| "loss": 3.4597, |
| "step": 32550 |
| }, |
| { |
| "epoch": 3.5087719298245617, |
| "grad_norm": 0.5807498097419739, |
| "learning_rate": 0.0003899709083072944, |
| "loss": 3.4839, |
| "step": 32600 |
| }, |
| { |
| "epoch": 3.5141534818641698, |
| "grad_norm": 0.6208375692367554, |
| "learning_rate": 0.000389647667277233, |
| "loss": 3.4497, |
| "step": 32650 |
| }, |
| { |
| "epoch": 3.519535033903778, |
| "grad_norm": 0.6369626522064209, |
| "learning_rate": 0.00038932442624717165, |
| "loss": 3.4686, |
| "step": 32700 |
| }, |
| { |
| "epoch": 3.524916585943386, |
| "grad_norm": 0.5826206207275391, |
| "learning_rate": 0.0003890011852171102, |
| "loss": 3.4625, |
| "step": 32750 |
| }, |
| { |
| "epoch": 3.530298137982994, |
| "grad_norm": 0.6600220203399658, |
| "learning_rate": 0.0003886779441870488, |
| "loss": 3.4547, |
| "step": 32800 |
| }, |
| { |
| "epoch": 3.5356796900226026, |
| "grad_norm": 0.6286580562591553, |
| "learning_rate": 0.0003883547031569874, |
| "loss": 3.4523, |
| "step": 32850 |
| }, |
| { |
| "epoch": 3.5410612420622107, |
| "grad_norm": 0.5954705476760864, |
| "learning_rate": 0.0003880314621269259, |
| "loss": 3.4432, |
| "step": 32900 |
| }, |
| { |
| "epoch": 3.546442794101819, |
| "grad_norm": 0.5685660243034363, |
| "learning_rate": 0.0003877082210968645, |
| "loss": 3.4673, |
| "step": 32950 |
| }, |
| { |
| "epoch": 3.5518243461414274, |
| "grad_norm": 0.5767381191253662, |
| "learning_rate": 0.00038738498006680316, |
| "loss": 3.4729, |
| "step": 33000 |
| }, |
| { |
| "epoch": 3.5518243461414274, |
| "eval_accuracy": 0.3732153898992581, |
| "eval_loss": 3.4731204509735107, |
| "eval_runtime": 180.0807, |
| "eval_samples_per_second": 100.016, |
| "eval_steps_per_second": 6.253, |
| "step": 33000 |
| }, |
| { |
| "epoch": 3.5572058981810355, |
| "grad_norm": 0.5783354043960571, |
| "learning_rate": 0.0003870617390367417, |
| "loss": 3.4849, |
| "step": 33050 |
| }, |
| { |
| "epoch": 3.5625874502206436, |
| "grad_norm": 0.6520091891288757, |
| "learning_rate": 0.0003867384980066803, |
| "loss": 3.4411, |
| "step": 33100 |
| }, |
| { |
| "epoch": 3.5679690022602517, |
| "grad_norm": 0.6067547798156738, |
| "learning_rate": 0.00038641525697661884, |
| "loss": 3.4514, |
| "step": 33150 |
| }, |
| { |
| "epoch": 3.57335055429986, |
| "grad_norm": 0.6351250410079956, |
| "learning_rate": 0.00038609201594655743, |
| "loss": 3.4697, |
| "step": 33200 |
| }, |
| { |
| "epoch": 3.5787321063394684, |
| "grad_norm": 0.5866857171058655, |
| "learning_rate": 0.000385768774916496, |
| "loss": 3.4711, |
| "step": 33250 |
| }, |
| { |
| "epoch": 3.5841136583790765, |
| "grad_norm": 0.6151683330535889, |
| "learning_rate": 0.0003854455338864346, |
| "loss": 3.4429, |
| "step": 33300 |
| }, |
| { |
| "epoch": 3.5894952104186846, |
| "grad_norm": 0.580003023147583, |
| "learning_rate": 0.0003851222928563732, |
| "loss": 3.4521, |
| "step": 33350 |
| }, |
| { |
| "epoch": 3.594876762458293, |
| "grad_norm": 0.6485602259635925, |
| "learning_rate": 0.0003847990518263118, |
| "loss": 3.4508, |
| "step": 33400 |
| }, |
| { |
| "epoch": 3.6002583144979012, |
| "grad_norm": 0.6691255569458008, |
| "learning_rate": 0.00038447581079625035, |
| "loss": 3.4681, |
| "step": 33450 |
| }, |
| { |
| "epoch": 3.6056398665375093, |
| "grad_norm": 0.5927985310554504, |
| "learning_rate": 0.00038415903458679016, |
| "loss": 3.4552, |
| "step": 33500 |
| }, |
| { |
| "epoch": 3.611021418577118, |
| "grad_norm": 0.6414245367050171, |
| "learning_rate": 0.00038383579355672875, |
| "loss": 3.4548, |
| "step": 33550 |
| }, |
| { |
| "epoch": 3.616402970616726, |
| "grad_norm": 0.6126065850257874, |
| "learning_rate": 0.00038351255252666735, |
| "loss": 3.4498, |
| "step": 33600 |
| }, |
| { |
| "epoch": 3.621784522656334, |
| "grad_norm": 0.5922813415527344, |
| "learning_rate": 0.00038318931149660594, |
| "loss": 3.4548, |
| "step": 33650 |
| }, |
| { |
| "epoch": 3.627166074695942, |
| "grad_norm": 0.6042068004608154, |
| "learning_rate": 0.00038286607046654454, |
| "loss": 3.4562, |
| "step": 33700 |
| }, |
| { |
| "epoch": 3.6325476267355503, |
| "grad_norm": 0.6460469961166382, |
| "learning_rate": 0.00038254929425708434, |
| "loss": 3.4569, |
| "step": 33750 |
| }, |
| { |
| "epoch": 3.637929178775159, |
| "grad_norm": 0.6714599132537842, |
| "learning_rate": 0.00038222605322702294, |
| "loss": 3.4775, |
| "step": 33800 |
| }, |
| { |
| "epoch": 3.643310730814767, |
| "grad_norm": 0.6167879104614258, |
| "learning_rate": 0.00038190281219696153, |
| "loss": 3.4382, |
| "step": 33850 |
| }, |
| { |
| "epoch": 3.648692282854375, |
| "grad_norm": 0.5907341241836548, |
| "learning_rate": 0.00038157957116690007, |
| "loss": 3.4661, |
| "step": 33900 |
| }, |
| { |
| "epoch": 3.6540738348939836, |
| "grad_norm": 0.6232337355613708, |
| "learning_rate": 0.00038125633013683867, |
| "loss": 3.4681, |
| "step": 33950 |
| }, |
| { |
| "epoch": 3.6594553869335917, |
| "grad_norm": 0.6158431768417358, |
| "learning_rate": 0.0003809330891067772, |
| "loss": 3.4414, |
| "step": 34000 |
| }, |
| { |
| "epoch": 3.6594553869335917, |
| "eval_accuracy": 0.3739183740834178, |
| "eval_loss": 3.4675493240356445, |
| "eval_runtime": 179.9271, |
| "eval_samples_per_second": 100.102, |
| "eval_steps_per_second": 6.258, |
| "step": 34000 |
| }, |
| { |
| "epoch": 3.6648369389732, |
| "grad_norm": 0.6306874752044678, |
| "learning_rate": 0.00038060984807671586, |
| "loss": 3.4467, |
| "step": 34050 |
| }, |
| { |
| "epoch": 3.670218491012808, |
| "grad_norm": 0.6371271014213562, |
| "learning_rate": 0.00038028660704665445, |
| "loss": 3.4585, |
| "step": 34100 |
| }, |
| { |
| "epoch": 3.675600043052416, |
| "grad_norm": 0.5911433696746826, |
| "learning_rate": 0.000379963366016593, |
| "loss": 3.4271, |
| "step": 34150 |
| }, |
| { |
| "epoch": 3.6809815950920246, |
| "grad_norm": 0.610369086265564, |
| "learning_rate": 0.0003796401249865316, |
| "loss": 3.4656, |
| "step": 34200 |
| }, |
| { |
| "epoch": 3.6863631471316327, |
| "grad_norm": 0.6166794300079346, |
| "learning_rate": 0.0003793168839564701, |
| "loss": 3.4541, |
| "step": 34250 |
| }, |
| { |
| "epoch": 3.691744699171241, |
| "grad_norm": 0.6205360889434814, |
| "learning_rate": 0.0003789936429264088, |
| "loss": 3.4503, |
| "step": 34300 |
| }, |
| { |
| "epoch": 3.6971262512108494, |
| "grad_norm": 0.6370207071304321, |
| "learning_rate": 0.00037867040189634737, |
| "loss": 3.4441, |
| "step": 34350 |
| }, |
| { |
| "epoch": 3.7025078032504575, |
| "grad_norm": 0.6453512907028198, |
| "learning_rate": 0.0003783471608662859, |
| "loss": 3.4555, |
| "step": 34400 |
| }, |
| { |
| "epoch": 3.7078893552900656, |
| "grad_norm": 0.5642420649528503, |
| "learning_rate": 0.0003780239198362245, |
| "loss": 3.45, |
| "step": 34450 |
| }, |
| { |
| "epoch": 3.713270907329674, |
| "grad_norm": 0.5417167544364929, |
| "learning_rate": 0.0003777006788061631, |
| "loss": 3.4647, |
| "step": 34500 |
| }, |
| { |
| "epoch": 3.7186524593692822, |
| "grad_norm": 0.5547471642494202, |
| "learning_rate": 0.00037737743777610164, |
| "loss": 3.4465, |
| "step": 34550 |
| }, |
| { |
| "epoch": 3.7240340114088903, |
| "grad_norm": 0.5820665955543518, |
| "learning_rate": 0.0003770541967460403, |
| "loss": 3.4561, |
| "step": 34600 |
| }, |
| { |
| "epoch": 3.7294155634484984, |
| "grad_norm": 0.5901154279708862, |
| "learning_rate": 0.0003767309557159789, |
| "loss": 3.4465, |
| "step": 34650 |
| }, |
| { |
| "epoch": 3.7347971154881066, |
| "grad_norm": 0.5940930247306824, |
| "learning_rate": 0.0003764077146859174, |
| "loss": 3.4482, |
| "step": 34700 |
| }, |
| { |
| "epoch": 3.740178667527715, |
| "grad_norm": 0.6073951721191406, |
| "learning_rate": 0.000376084473655856, |
| "loss": 3.4522, |
| "step": 34750 |
| }, |
| { |
| "epoch": 3.745560219567323, |
| "grad_norm": 0.573229968547821, |
| "learning_rate": 0.00037576123262579456, |
| "loss": 3.4248, |
| "step": 34800 |
| }, |
| { |
| "epoch": 3.7509417716069313, |
| "grad_norm": 0.6249234676361084, |
| "learning_rate": 0.00037543799159573315, |
| "loss": 3.4414, |
| "step": 34850 |
| }, |
| { |
| "epoch": 3.75632332364654, |
| "grad_norm": 0.5785616040229797, |
| "learning_rate": 0.0003751147505656718, |
| "loss": 3.4517, |
| "step": 34900 |
| }, |
| { |
| "epoch": 3.761704875686148, |
| "grad_norm": 0.5984014272689819, |
| "learning_rate": 0.00037479150953561034, |
| "loss": 3.4535, |
| "step": 34950 |
| }, |
| { |
| "epoch": 3.767086427725756, |
| "grad_norm": 0.575398862361908, |
| "learning_rate": 0.00037446826850554894, |
| "loss": 3.4656, |
| "step": 35000 |
| }, |
| { |
| "epoch": 3.767086427725756, |
| "eval_accuracy": 0.3745656193361287, |
| "eval_loss": 3.462155818939209, |
| "eval_runtime": 180.1892, |
| "eval_samples_per_second": 99.956, |
| "eval_steps_per_second": 6.249, |
| "step": 35000 |
| }, |
| { |
| "epoch": 3.772467979765364, |
| "grad_norm": 0.5907803177833557, |
| "learning_rate": 0.00037414502747548753, |
| "loss": 3.4421, |
| "step": 35050 |
| }, |
| { |
| "epoch": 3.7778495318049723, |
| "grad_norm": 0.5853913426399231, |
| "learning_rate": 0.00037382178644542607, |
| "loss": 3.4405, |
| "step": 35100 |
| }, |
| { |
| "epoch": 3.783231083844581, |
| "grad_norm": 0.6257376074790955, |
| "learning_rate": 0.00037349854541536467, |
| "loss": 3.4686, |
| "step": 35150 |
| }, |
| { |
| "epoch": 3.788612635884189, |
| "grad_norm": 0.6458983421325684, |
| "learning_rate": 0.0003731753043853033, |
| "loss": 3.4546, |
| "step": 35200 |
| }, |
| { |
| "epoch": 3.793994187923797, |
| "grad_norm": 0.612881064414978, |
| "learning_rate": 0.00037285206335524186, |
| "loss": 3.4791, |
| "step": 35250 |
| }, |
| { |
| "epoch": 3.7993757399634056, |
| "grad_norm": 0.5735159516334534, |
| "learning_rate": 0.00037252882232518045, |
| "loss": 3.4668, |
| "step": 35300 |
| }, |
| { |
| "epoch": 3.8047572920030137, |
| "grad_norm": 0.6126082539558411, |
| "learning_rate": 0.000372205581295119, |
| "loss": 3.456, |
| "step": 35350 |
| }, |
| { |
| "epoch": 3.810138844042622, |
| "grad_norm": 0.5752865672111511, |
| "learning_rate": 0.0003718823402650576, |
| "loss": 3.4603, |
| "step": 35400 |
| }, |
| { |
| "epoch": 3.8155203960822304, |
| "grad_norm": 0.6268784999847412, |
| "learning_rate": 0.00037155909923499624, |
| "loss": 3.4538, |
| "step": 35450 |
| }, |
| { |
| "epoch": 3.8209019481218385, |
| "grad_norm": 0.613398015499115, |
| "learning_rate": 0.0003712358582049348, |
| "loss": 3.4408, |
| "step": 35500 |
| }, |
| { |
| "epoch": 3.8262835001614466, |
| "grad_norm": 0.6554962992668152, |
| "learning_rate": 0.00037091261717487337, |
| "loss": 3.4481, |
| "step": 35550 |
| }, |
| { |
| "epoch": 3.8316650522010547, |
| "grad_norm": 0.6721749901771545, |
| "learning_rate": 0.00037058937614481197, |
| "loss": 3.4502, |
| "step": 35600 |
| }, |
| { |
| "epoch": 3.837046604240663, |
| "grad_norm": 0.630124568939209, |
| "learning_rate": 0.0003702661351147505, |
| "loss": 3.4629, |
| "step": 35650 |
| }, |
| { |
| "epoch": 3.8424281562802713, |
| "grad_norm": 0.5516995787620544, |
| "learning_rate": 0.0003699428940846891, |
| "loss": 3.4512, |
| "step": 35700 |
| }, |
| { |
| "epoch": 3.8478097083198795, |
| "grad_norm": 0.7088376879692078, |
| "learning_rate": 0.00036961965305462775, |
| "loss": 3.4578, |
| "step": 35750 |
| }, |
| { |
| "epoch": 3.8531912603594876, |
| "grad_norm": 0.6366024613380432, |
| "learning_rate": 0.0003692964120245663, |
| "loss": 3.4455, |
| "step": 35800 |
| }, |
| { |
| "epoch": 3.858572812399096, |
| "grad_norm": 0.6255238056182861, |
| "learning_rate": 0.0003689731709945049, |
| "loss": 3.4399, |
| "step": 35850 |
| }, |
| { |
| "epoch": 3.863954364438704, |
| "grad_norm": 0.5794286131858826, |
| "learning_rate": 0.0003686499299644434, |
| "loss": 3.4501, |
| "step": 35900 |
| }, |
| { |
| "epoch": 3.8693359164783123, |
| "grad_norm": 0.6042854189872742, |
| "learning_rate": 0.000368326688934382, |
| "loss": 3.4426, |
| "step": 35950 |
| }, |
| { |
| "epoch": 3.8747174685179204, |
| "grad_norm": 0.6024532914161682, |
| "learning_rate": 0.0003680034479043206, |
| "loss": 3.4696, |
| "step": 36000 |
| }, |
| { |
| "epoch": 3.8747174685179204, |
| "eval_accuracy": 0.3750145730685781, |
| "eval_loss": 3.4540035724639893, |
| "eval_runtime": 179.8843, |
| "eval_samples_per_second": 100.125, |
| "eval_steps_per_second": 6.26, |
| "step": 36000 |
| }, |
| { |
| "epoch": 3.8800990205575285, |
| "grad_norm": 0.5712722539901733, |
| "learning_rate": 0.0003676802068742592, |
| "loss": 3.4696, |
| "step": 36050 |
| }, |
| { |
| "epoch": 3.885480572597137, |
| "grad_norm": 0.6209331750869751, |
| "learning_rate": 0.0003673569658441978, |
| "loss": 3.4485, |
| "step": 36100 |
| }, |
| { |
| "epoch": 3.890862124636745, |
| "grad_norm": 0.573322594165802, |
| "learning_rate": 0.0003670337248141364, |
| "loss": 3.4665, |
| "step": 36150 |
| }, |
| { |
| "epoch": 3.8962436766763533, |
| "grad_norm": 0.6659927368164062, |
| "learning_rate": 0.00036671048378407494, |
| "loss": 3.437, |
| "step": 36200 |
| }, |
| { |
| "epoch": 3.901625228715962, |
| "grad_norm": 0.6092972159385681, |
| "learning_rate": 0.00036638724275401353, |
| "loss": 3.4379, |
| "step": 36250 |
| }, |
| { |
| "epoch": 3.90700678075557, |
| "grad_norm": 0.5966007113456726, |
| "learning_rate": 0.0003660640017239522, |
| "loss": 3.4472, |
| "step": 36300 |
| }, |
| { |
| "epoch": 3.912388332795178, |
| "grad_norm": 0.6116428971290588, |
| "learning_rate": 0.0003657407606938907, |
| "loss": 3.4242, |
| "step": 36350 |
| }, |
| { |
| "epoch": 3.9177698848347866, |
| "grad_norm": 0.6971980333328247, |
| "learning_rate": 0.0003654175196638293, |
| "loss": 3.4542, |
| "step": 36400 |
| }, |
| { |
| "epoch": 3.9231514368743947, |
| "grad_norm": 0.5674816966056824, |
| "learning_rate": 0.00036509427863376786, |
| "loss": 3.4674, |
| "step": 36450 |
| }, |
| { |
| "epoch": 3.928532988914003, |
| "grad_norm": 0.596181333065033, |
| "learning_rate": 0.00036477103760370645, |
| "loss": 3.4494, |
| "step": 36500 |
| }, |
| { |
| "epoch": 3.933914540953611, |
| "grad_norm": 0.6476390957832336, |
| "learning_rate": 0.00036444779657364505, |
| "loss": 3.4458, |
| "step": 36550 |
| }, |
| { |
| "epoch": 3.939296092993219, |
| "grad_norm": 0.6111776828765869, |
| "learning_rate": 0.00036412455554358364, |
| "loss": 3.4511, |
| "step": 36600 |
| }, |
| { |
| "epoch": 3.9446776450328276, |
| "grad_norm": 0.6062666773796082, |
| "learning_rate": 0.00036380131451352224, |
| "loss": 3.4604, |
| "step": 36650 |
| }, |
| { |
| "epoch": 3.9500591970724357, |
| "grad_norm": 0.6287407279014587, |
| "learning_rate": 0.00036347807348346083, |
| "loss": 3.4553, |
| "step": 36700 |
| }, |
| { |
| "epoch": 3.955440749112044, |
| "grad_norm": 0.6396850347518921, |
| "learning_rate": 0.00036315483245339937, |
| "loss": 3.4422, |
| "step": 36750 |
| }, |
| { |
| "epoch": 3.9608223011516523, |
| "grad_norm": 0.6441523432731628, |
| "learning_rate": 0.00036283159142333797, |
| "loss": 3.4455, |
| "step": 36800 |
| }, |
| { |
| "epoch": 3.9662038531912605, |
| "grad_norm": 0.5901207327842712, |
| "learning_rate": 0.0003625083503932765, |
| "loss": 3.4434, |
| "step": 36850 |
| }, |
| { |
| "epoch": 3.9715854052308686, |
| "grad_norm": 0.6204649209976196, |
| "learning_rate": 0.00036218510936321516, |
| "loss": 3.4559, |
| "step": 36900 |
| }, |
| { |
| "epoch": 3.9769669572704767, |
| "grad_norm": 0.6309411525726318, |
| "learning_rate": 0.00036186186833315375, |
| "loss": 3.4507, |
| "step": 36950 |
| }, |
| { |
| "epoch": 3.9823485093100848, |
| "grad_norm": 0.6167184114456177, |
| "learning_rate": 0.0003615386273030923, |
| "loss": 3.4349, |
| "step": 37000 |
| }, |
| { |
| "epoch": 3.9823485093100848, |
| "eval_accuracy": 0.3759814751172392, |
| "eval_loss": 3.4468631744384766, |
| "eval_runtime": 179.8588, |
| "eval_samples_per_second": 100.14, |
| "eval_steps_per_second": 6.26, |
| "step": 37000 |
| }, |
| { |
| "epoch": 3.9877300613496933, |
| "grad_norm": 0.6533148884773254, |
| "learning_rate": 0.0003612153862730309, |
| "loss": 3.4514, |
| "step": 37050 |
| }, |
| { |
| "epoch": 3.9931116133893014, |
| "grad_norm": 0.5957464575767517, |
| "learning_rate": 0.0003608921452429695, |
| "loss": 3.4542, |
| "step": 37100 |
| }, |
| { |
| "epoch": 3.9984931654289095, |
| "grad_norm": 0.6055651903152466, |
| "learning_rate": 0.000360568904212908, |
| "loss": 3.4395, |
| "step": 37150 |
| }, |
| { |
| "epoch": 4.003874717468518, |
| "grad_norm": 0.579195499420166, |
| "learning_rate": 0.00036024566318284667, |
| "loss": 3.3731, |
| "step": 37200 |
| }, |
| { |
| "epoch": 4.009256269508126, |
| "grad_norm": 0.6339341998100281, |
| "learning_rate": 0.00035992242215278526, |
| "loss": 3.3399, |
| "step": 37250 |
| }, |
| { |
| "epoch": 4.014637821547734, |
| "grad_norm": 0.6784026622772217, |
| "learning_rate": 0.0003595991811227238, |
| "loss": 3.3643, |
| "step": 37300 |
| }, |
| { |
| "epoch": 4.020019373587343, |
| "grad_norm": 0.5875265598297119, |
| "learning_rate": 0.0003592759400926624, |
| "loss": 3.356, |
| "step": 37350 |
| }, |
| { |
| "epoch": 4.0254009256269505, |
| "grad_norm": 0.6232568621635437, |
| "learning_rate": 0.00035895269906260094, |
| "loss": 3.3561, |
| "step": 37400 |
| }, |
| { |
| "epoch": 4.030782477666559, |
| "grad_norm": 0.5812575221061707, |
| "learning_rate": 0.0003586294580325396, |
| "loss": 3.3639, |
| "step": 37450 |
| }, |
| { |
| "epoch": 4.036164029706168, |
| "grad_norm": 0.622757613658905, |
| "learning_rate": 0.0003583062170024782, |
| "loss": 3.34, |
| "step": 37500 |
| }, |
| { |
| "epoch": 4.041545581745775, |
| "grad_norm": 0.6129567623138428, |
| "learning_rate": 0.0003579829759724167, |
| "loss": 3.3655, |
| "step": 37550 |
| }, |
| { |
| "epoch": 4.046927133785384, |
| "grad_norm": 0.6567339301109314, |
| "learning_rate": 0.0003576597349423553, |
| "loss": 3.3565, |
| "step": 37600 |
| }, |
| { |
| "epoch": 4.0523086858249915, |
| "grad_norm": 0.6535717248916626, |
| "learning_rate": 0.0003573364939122939, |
| "loss": 3.3767, |
| "step": 37650 |
| }, |
| { |
| "epoch": 4.0576902378646, |
| "grad_norm": 0.6005692481994629, |
| "learning_rate": 0.00035701325288223245, |
| "loss": 3.3669, |
| "step": 37700 |
| }, |
| { |
| "epoch": 4.063071789904209, |
| "grad_norm": 0.5821049213409424, |
| "learning_rate": 0.0003566900118521711, |
| "loss": 3.3626, |
| "step": 37750 |
| }, |
| { |
| "epoch": 4.068453341943816, |
| "grad_norm": 0.6144912838935852, |
| "learning_rate": 0.0003563667708221097, |
| "loss": 3.3676, |
| "step": 37800 |
| }, |
| { |
| "epoch": 4.073834893983425, |
| "grad_norm": 0.5643580555915833, |
| "learning_rate": 0.00035604352979204824, |
| "loss": 3.3654, |
| "step": 37850 |
| }, |
| { |
| "epoch": 4.079216446023033, |
| "grad_norm": 0.5872395634651184, |
| "learning_rate": 0.00035572675358258804, |
| "loss": 3.3534, |
| "step": 37900 |
| }, |
| { |
| "epoch": 4.084597998062641, |
| "grad_norm": 0.6200112700462341, |
| "learning_rate": 0.00035540351255252664, |
| "loss": 3.368, |
| "step": 37950 |
| }, |
| { |
| "epoch": 4.08997955010225, |
| "grad_norm": 0.6187693476676941, |
| "learning_rate": 0.00035508027152246523, |
| "loss": 3.3756, |
| "step": 38000 |
| }, |
| { |
| "epoch": 4.08997955010225, |
| "eval_accuracy": 0.3760597051964656, |
| "eval_loss": 3.4495081901550293, |
| "eval_runtime": 179.77, |
| "eval_samples_per_second": 100.189, |
| "eval_steps_per_second": 6.264, |
| "step": 38000 |
| }, |
| { |
| "epoch": 4.095361102141858, |
| "grad_norm": 0.6478372812271118, |
| "learning_rate": 0.0003547570304924038, |
| "loss": 3.3678, |
| "step": 38050 |
| }, |
| { |
| "epoch": 4.100742654181466, |
| "grad_norm": 0.6313759088516235, |
| "learning_rate": 0.0003544337894623424, |
| "loss": 3.3722, |
| "step": 38100 |
| }, |
| { |
| "epoch": 4.106124206221074, |
| "grad_norm": 0.6767866611480713, |
| "learning_rate": 0.000354110548432281, |
| "loss": 3.3723, |
| "step": 38150 |
| }, |
| { |
| "epoch": 4.111505758260682, |
| "grad_norm": 0.5833486318588257, |
| "learning_rate": 0.00035378730740221956, |
| "loss": 3.3513, |
| "step": 38200 |
| }, |
| { |
| "epoch": 4.1168873103002905, |
| "grad_norm": 0.5952380299568176, |
| "learning_rate": 0.00035346406637215815, |
| "loss": 3.3763, |
| "step": 38250 |
| }, |
| { |
| "epoch": 4.122268862339899, |
| "grad_norm": 0.634947657585144, |
| "learning_rate": 0.0003531408253420967, |
| "loss": 3.3668, |
| "step": 38300 |
| }, |
| { |
| "epoch": 4.127650414379507, |
| "grad_norm": 0.6541725397109985, |
| "learning_rate": 0.0003528175843120353, |
| "loss": 3.3724, |
| "step": 38350 |
| }, |
| { |
| "epoch": 4.133031966419115, |
| "grad_norm": 0.6105016469955444, |
| "learning_rate": 0.00035249434328197394, |
| "loss": 3.367, |
| "step": 38400 |
| }, |
| { |
| "epoch": 4.138413518458724, |
| "grad_norm": 0.6355590224266052, |
| "learning_rate": 0.0003521711022519125, |
| "loss": 3.3708, |
| "step": 38450 |
| }, |
| { |
| "epoch": 4.1437950704983315, |
| "grad_norm": 0.5767462253570557, |
| "learning_rate": 0.00035185432604245234, |
| "loss": 3.39, |
| "step": 38500 |
| }, |
| { |
| "epoch": 4.14917662253794, |
| "grad_norm": 0.6215607523918152, |
| "learning_rate": 0.0003515310850123909, |
| "loss": 3.3777, |
| "step": 38550 |
| }, |
| { |
| "epoch": 4.154558174577549, |
| "grad_norm": 0.6338831186294556, |
| "learning_rate": 0.00035120784398232947, |
| "loss": 3.3888, |
| "step": 38600 |
| }, |
| { |
| "epoch": 4.159939726617156, |
| "grad_norm": 0.6333625912666321, |
| "learning_rate": 0.000350884602952268, |
| "loss": 3.3751, |
| "step": 38650 |
| }, |
| { |
| "epoch": 4.165321278656765, |
| "grad_norm": 0.6430308222770691, |
| "learning_rate": 0.0003505613619222066, |
| "loss": 3.385, |
| "step": 38700 |
| }, |
| { |
| "epoch": 4.1707028306963725, |
| "grad_norm": 0.5828115344047546, |
| "learning_rate": 0.0003502381208921452, |
| "loss": 3.4078, |
| "step": 38750 |
| }, |
| { |
| "epoch": 4.176084382735981, |
| "grad_norm": 0.652651846408844, |
| "learning_rate": 0.0003499148798620838, |
| "loss": 3.3657, |
| "step": 38800 |
| }, |
| { |
| "epoch": 4.18146593477559, |
| "grad_norm": 0.6599338054656982, |
| "learning_rate": 0.0003495916388320224, |
| "loss": 3.3897, |
| "step": 38850 |
| }, |
| { |
| "epoch": 4.186847486815197, |
| "grad_norm": 0.5994952917098999, |
| "learning_rate": 0.000349268397801961, |
| "loss": 3.3771, |
| "step": 38900 |
| }, |
| { |
| "epoch": 4.192229038854806, |
| "grad_norm": 0.6271725296974182, |
| "learning_rate": 0.0003489451567718995, |
| "loss": 3.3818, |
| "step": 38950 |
| }, |
| { |
| "epoch": 4.197610590894414, |
| "grad_norm": 0.617135226726532, |
| "learning_rate": 0.0003486219157418381, |
| "loss": 3.3611, |
| "step": 39000 |
| }, |
| { |
| "epoch": 4.197610590894414, |
| "eval_accuracy": 0.37661372627143164, |
| "eval_loss": 3.4469828605651855, |
| "eval_runtime": 179.8408, |
| "eval_samples_per_second": 100.15, |
| "eval_steps_per_second": 6.261, |
| "step": 39000 |
| }, |
| { |
| "epoch": 4.202992142934022, |
| "grad_norm": 0.61525958776474, |
| "learning_rate": 0.00034829867471177677, |
| "loss": 3.3823, |
| "step": 39050 |
| }, |
| { |
| "epoch": 4.208373694973631, |
| "grad_norm": 0.638131320476532, |
| "learning_rate": 0.0003479754336817153, |
| "loss": 3.384, |
| "step": 39100 |
| }, |
| { |
| "epoch": 4.213755247013238, |
| "grad_norm": 0.6497801542282104, |
| "learning_rate": 0.0003476521926516539, |
| "loss": 3.3798, |
| "step": 39150 |
| }, |
| { |
| "epoch": 4.219136799052847, |
| "grad_norm": 0.6336971521377563, |
| "learning_rate": 0.00034732895162159245, |
| "loss": 3.3859, |
| "step": 39200 |
| }, |
| { |
| "epoch": 4.224518351092455, |
| "grad_norm": 0.7161690592765808, |
| "learning_rate": 0.00034700571059153104, |
| "loss": 3.3821, |
| "step": 39250 |
| }, |
| { |
| "epoch": 4.229899903132063, |
| "grad_norm": 0.5993850231170654, |
| "learning_rate": 0.00034668246956146963, |
| "loss": 3.3726, |
| "step": 39300 |
| }, |
| { |
| "epoch": 4.2352814551716715, |
| "grad_norm": 0.6245800256729126, |
| "learning_rate": 0.00034635922853140823, |
| "loss": 3.398, |
| "step": 39350 |
| }, |
| { |
| "epoch": 4.24066300721128, |
| "grad_norm": 0.599709689617157, |
| "learning_rate": 0.0003460359875013468, |
| "loss": 3.3838, |
| "step": 39400 |
| }, |
| { |
| "epoch": 4.246044559250888, |
| "grad_norm": 0.6475380063056946, |
| "learning_rate": 0.0003457127464712854, |
| "loss": 3.3766, |
| "step": 39450 |
| }, |
| { |
| "epoch": 4.251426111290496, |
| "grad_norm": 0.6066352128982544, |
| "learning_rate": 0.00034538950544122396, |
| "loss": 3.3921, |
| "step": 39500 |
| }, |
| { |
| "epoch": 4.256807663330104, |
| "grad_norm": 0.5933494567871094, |
| "learning_rate": 0.00034506626441116255, |
| "loss": 3.3822, |
| "step": 39550 |
| }, |
| { |
| "epoch": 4.2621892153697125, |
| "grad_norm": 0.6642129421234131, |
| "learning_rate": 0.0003447430233811011, |
| "loss": 3.3944, |
| "step": 39600 |
| }, |
| { |
| "epoch": 4.267570767409321, |
| "grad_norm": 0.6359052062034607, |
| "learning_rate": 0.00034441978235103974, |
| "loss": 3.3894, |
| "step": 39650 |
| }, |
| { |
| "epoch": 4.272952319448929, |
| "grad_norm": 0.6086246967315674, |
| "learning_rate": 0.00034409654132097834, |
| "loss": 3.3938, |
| "step": 39700 |
| }, |
| { |
| "epoch": 4.278333871488537, |
| "grad_norm": 0.6143749356269836, |
| "learning_rate": 0.0003437733002909169, |
| "loss": 3.3853, |
| "step": 39750 |
| }, |
| { |
| "epoch": 4.283715423528146, |
| "grad_norm": 0.6609334945678711, |
| "learning_rate": 0.00034345005926085547, |
| "loss": 3.3985, |
| "step": 39800 |
| }, |
| { |
| "epoch": 4.2890969755677535, |
| "grad_norm": 0.6083665490150452, |
| "learning_rate": 0.00034312681823079407, |
| "loss": 3.361, |
| "step": 39850 |
| }, |
| { |
| "epoch": 4.294478527607362, |
| "grad_norm": 0.6407836675643921, |
| "learning_rate": 0.00034280357720073266, |
| "loss": 3.3881, |
| "step": 39900 |
| }, |
| { |
| "epoch": 4.299860079646971, |
| "grad_norm": 0.6500919461250305, |
| "learning_rate": 0.00034248033617067126, |
| "loss": 3.4023, |
| "step": 39950 |
| }, |
| { |
| "epoch": 4.305241631686578, |
| "grad_norm": 0.6269468665122986, |
| "learning_rate": 0.00034215709514060985, |
| "loss": 3.3894, |
| "step": 40000 |
| }, |
| { |
| "epoch": 4.305241631686578, |
| "eval_accuracy": 0.37716155413179225, |
| "eval_loss": 3.440868377685547, |
| "eval_runtime": 179.5856, |
| "eval_samples_per_second": 100.292, |
| "eval_steps_per_second": 6.27, |
| "step": 40000 |
| }, |
| { |
| "epoch": 4.310623183726187, |
| "grad_norm": 0.6156846284866333, |
| "learning_rate": 0.0003418338541105484, |
| "loss": 3.3992, |
| "step": 40050 |
| }, |
| { |
| "epoch": 4.3160047357657945, |
| "grad_norm": 0.6544471979141235, |
| "learning_rate": 0.000341510613080487, |
| "loss": 3.3941, |
| "step": 40100 |
| }, |
| { |
| "epoch": 4.321386287805403, |
| "grad_norm": 0.6093199849128723, |
| "learning_rate": 0.0003411873720504255, |
| "loss": 3.3693, |
| "step": 40150 |
| }, |
| { |
| "epoch": 4.326767839845012, |
| "grad_norm": 0.5988791584968567, |
| "learning_rate": 0.0003408641310203642, |
| "loss": 3.3889, |
| "step": 40200 |
| }, |
| { |
| "epoch": 4.332149391884619, |
| "grad_norm": 0.640820324420929, |
| "learning_rate": 0.00034054088999030277, |
| "loss": 3.3808, |
| "step": 40250 |
| }, |
| { |
| "epoch": 4.337530943924228, |
| "grad_norm": 0.607883632183075, |
| "learning_rate": 0.0003402176489602413, |
| "loss": 3.3864, |
| "step": 40300 |
| }, |
| { |
| "epoch": 4.342912495963836, |
| "grad_norm": 0.6579638123512268, |
| "learning_rate": 0.0003398944079301799, |
| "loss": 3.394, |
| "step": 40350 |
| }, |
| { |
| "epoch": 4.348294048003444, |
| "grad_norm": 0.587094247341156, |
| "learning_rate": 0.0003395711669001185, |
| "loss": 3.3886, |
| "step": 40400 |
| }, |
| { |
| "epoch": 4.3536756000430525, |
| "grad_norm": 0.6616682410240173, |
| "learning_rate": 0.00033924792587005704, |
| "loss": 3.3742, |
| "step": 40450 |
| }, |
| { |
| "epoch": 4.359057152082661, |
| "grad_norm": 0.6799524426460266, |
| "learning_rate": 0.0003389246848399957, |
| "loss": 3.3839, |
| "step": 40500 |
| }, |
| { |
| "epoch": 4.364438704122269, |
| "grad_norm": 0.6899614930152893, |
| "learning_rate": 0.0003386014438099343, |
| "loss": 3.3992, |
| "step": 40550 |
| }, |
| { |
| "epoch": 4.369820256161877, |
| "grad_norm": 0.6630707383155823, |
| "learning_rate": 0.0003382782027798728, |
| "loss": 3.3764, |
| "step": 40600 |
| }, |
| { |
| "epoch": 4.375201808201485, |
| "grad_norm": 0.598388671875, |
| "learning_rate": 0.0003379549617498114, |
| "loss": 3.3942, |
| "step": 40650 |
| }, |
| { |
| "epoch": 4.3805833602410935, |
| "grad_norm": 0.6070818305015564, |
| "learning_rate": 0.00033763172071974996, |
| "loss": 3.3867, |
| "step": 40700 |
| }, |
| { |
| "epoch": 4.385964912280702, |
| "grad_norm": 0.5876506567001343, |
| "learning_rate": 0.0003373149445102898, |
| "loss": 3.3963, |
| "step": 40750 |
| }, |
| { |
| "epoch": 4.39134646432031, |
| "grad_norm": 0.6956170797348022, |
| "learning_rate": 0.00033699170348022836, |
| "loss": 3.3942, |
| "step": 40800 |
| }, |
| { |
| "epoch": 4.396728016359918, |
| "grad_norm": 0.6381921172142029, |
| "learning_rate": 0.000336668462450167, |
| "loss": 3.3903, |
| "step": 40850 |
| }, |
| { |
| "epoch": 4.402109568399527, |
| "grad_norm": 0.5954846143722534, |
| "learning_rate": 0.0003363452214201056, |
| "loss": 3.3904, |
| "step": 40900 |
| }, |
| { |
| "epoch": 4.4074911204391345, |
| "grad_norm": 0.6178380250930786, |
| "learning_rate": 0.00033602198039004415, |
| "loss": 3.3963, |
| "step": 40950 |
| }, |
| { |
| "epoch": 4.412872672478743, |
| "grad_norm": 0.6097917556762695, |
| "learning_rate": 0.00033569873935998274, |
| "loss": 3.3884, |
| "step": 41000 |
| }, |
| { |
| "epoch": 4.412872672478743, |
| "eval_accuracy": 0.3774454641276515, |
| "eval_loss": 3.4369654655456543, |
| "eval_runtime": 180.2414, |
| "eval_samples_per_second": 99.927, |
| "eval_steps_per_second": 6.247, |
| "step": 41000 |
| }, |
| { |
| "epoch": 4.418254224518351, |
| "grad_norm": 0.6266130208969116, |
| "learning_rate": 0.0003353754983299213, |
| "loss": 3.3913, |
| "step": 41050 |
| }, |
| { |
| "epoch": 4.423635776557959, |
| "grad_norm": 0.703901469707489, |
| "learning_rate": 0.0003350522572998599, |
| "loss": 3.3757, |
| "step": 41100 |
| }, |
| { |
| "epoch": 4.429017328597568, |
| "grad_norm": 0.6588935256004333, |
| "learning_rate": 0.0003347290162697985, |
| "loss": 3.4056, |
| "step": 41150 |
| }, |
| { |
| "epoch": 4.4343988806371755, |
| "grad_norm": 0.6265673041343689, |
| "learning_rate": 0.00033440577523973706, |
| "loss": 3.3899, |
| "step": 41200 |
| }, |
| { |
| "epoch": 4.439780432676784, |
| "grad_norm": 0.6723883152008057, |
| "learning_rate": 0.00033408253420967566, |
| "loss": 3.3989, |
| "step": 41250 |
| }, |
| { |
| "epoch": 4.445161984716393, |
| "grad_norm": 0.5973272323608398, |
| "learning_rate": 0.00033375929317961425, |
| "loss": 3.3731, |
| "step": 41300 |
| }, |
| { |
| "epoch": 4.450543536756, |
| "grad_norm": 0.6177666783332825, |
| "learning_rate": 0.0003334360521495528, |
| "loss": 3.4056, |
| "step": 41350 |
| }, |
| { |
| "epoch": 4.455925088795609, |
| "grad_norm": 0.6231200695037842, |
| "learning_rate": 0.0003331128111194914, |
| "loss": 3.3935, |
| "step": 41400 |
| }, |
| { |
| "epoch": 4.461306640835216, |
| "grad_norm": 0.6817479729652405, |
| "learning_rate": 0.00033278957008943004, |
| "loss": 3.3891, |
| "step": 41450 |
| }, |
| { |
| "epoch": 4.466688192874825, |
| "grad_norm": 0.6218167543411255, |
| "learning_rate": 0.0003324663290593686, |
| "loss": 3.3724, |
| "step": 41500 |
| }, |
| { |
| "epoch": 4.4720697449144335, |
| "grad_norm": 0.6025295257568359, |
| "learning_rate": 0.00033214308802930717, |
| "loss": 3.3758, |
| "step": 41550 |
| }, |
| { |
| "epoch": 4.477451296954041, |
| "grad_norm": 0.6618101000785828, |
| "learning_rate": 0.0003318198469992457, |
| "loss": 3.3954, |
| "step": 41600 |
| }, |
| { |
| "epoch": 4.48283284899365, |
| "grad_norm": 0.6926753520965576, |
| "learning_rate": 0.0003314966059691843, |
| "loss": 3.3904, |
| "step": 41650 |
| }, |
| { |
| "epoch": 4.488214401033258, |
| "grad_norm": 0.6272509098052979, |
| "learning_rate": 0.00033117336493912296, |
| "loss": 3.4091, |
| "step": 41700 |
| }, |
| { |
| "epoch": 4.493595953072866, |
| "grad_norm": 0.6476884484291077, |
| "learning_rate": 0.0003308501239090615, |
| "loss": 3.3659, |
| "step": 41750 |
| }, |
| { |
| "epoch": 4.4989775051124745, |
| "grad_norm": 0.6307544708251953, |
| "learning_rate": 0.0003305268828790001, |
| "loss": 3.4048, |
| "step": 41800 |
| }, |
| { |
| "epoch": 4.504359057152083, |
| "grad_norm": 0.6108989715576172, |
| "learning_rate": 0.0003302036418489387, |
| "loss": 3.383, |
| "step": 41850 |
| }, |
| { |
| "epoch": 4.509740609191691, |
| "grad_norm": 0.6389515995979309, |
| "learning_rate": 0.0003298804008188772, |
| "loss": 3.3898, |
| "step": 41900 |
| }, |
| { |
| "epoch": 4.515122161231299, |
| "grad_norm": 0.6314558386802673, |
| "learning_rate": 0.0003295571597888158, |
| "loss": 3.3848, |
| "step": 41950 |
| }, |
| { |
| "epoch": 4.520503713270907, |
| "grad_norm": 0.6694413423538208, |
| "learning_rate": 0.00032923391875875447, |
| "loss": 3.3698, |
| "step": 42000 |
| }, |
| { |
| "epoch": 4.520503713270907, |
| "eval_accuracy": 0.3781866941283219, |
| "eval_loss": 3.4318511486053467, |
| "eval_runtime": 179.6259, |
| "eval_samples_per_second": 100.27, |
| "eval_steps_per_second": 6.269, |
| "step": 42000 |
| }, |
| { |
| "epoch": 4.5258852653105155, |
| "grad_norm": 0.6342461705207825, |
| "learning_rate": 0.000328910677728693, |
| "loss": 3.3916, |
| "step": 42050 |
| }, |
| { |
| "epoch": 4.531266817350124, |
| "grad_norm": 0.6582828760147095, |
| "learning_rate": 0.0003285874366986316, |
| "loss": 3.4096, |
| "step": 42100 |
| }, |
| { |
| "epoch": 4.536648369389732, |
| "grad_norm": 0.6246548295021057, |
| "learning_rate": 0.00032826419566857015, |
| "loss": 3.3845, |
| "step": 42150 |
| }, |
| { |
| "epoch": 4.54202992142934, |
| "grad_norm": 0.6190484762191772, |
| "learning_rate": 0.00032794095463850874, |
| "loss": 3.396, |
| "step": 42200 |
| }, |
| { |
| "epoch": 4.547411473468949, |
| "grad_norm": 0.6867701411247253, |
| "learning_rate": 0.0003276177136084473, |
| "loss": 3.4047, |
| "step": 42250 |
| }, |
| { |
| "epoch": 4.5527930255085565, |
| "grad_norm": 0.6821978688240051, |
| "learning_rate": 0.00032729447257838593, |
| "loss": 3.3873, |
| "step": 42300 |
| }, |
| { |
| "epoch": 4.558174577548165, |
| "grad_norm": 0.6484827995300293, |
| "learning_rate": 0.0003269712315483245, |
| "loss": 3.382, |
| "step": 42350 |
| }, |
| { |
| "epoch": 4.563556129587774, |
| "grad_norm": 0.611552357673645, |
| "learning_rate": 0.00032664799051826306, |
| "loss": 3.3833, |
| "step": 42400 |
| }, |
| { |
| "epoch": 4.568937681627381, |
| "grad_norm": 0.6681063771247864, |
| "learning_rate": 0.00032632474948820166, |
| "loss": 3.3999, |
| "step": 42450 |
| }, |
| { |
| "epoch": 4.57431923366699, |
| "grad_norm": 0.6429010033607483, |
| "learning_rate": 0.00032600150845814025, |
| "loss": 3.3881, |
| "step": 42500 |
| }, |
| { |
| "epoch": 4.579700785706597, |
| "grad_norm": 0.6277284026145935, |
| "learning_rate": 0.0003256782674280788, |
| "loss": 3.394, |
| "step": 42550 |
| }, |
| { |
| "epoch": 4.585082337746206, |
| "grad_norm": 0.6517542600631714, |
| "learning_rate": 0.00032535502639801744, |
| "loss": 3.3899, |
| "step": 42600 |
| }, |
| { |
| "epoch": 4.5904638897858145, |
| "grad_norm": 0.5616545677185059, |
| "learning_rate": 0.00032503178536795604, |
| "loss": 3.3953, |
| "step": 42650 |
| }, |
| { |
| "epoch": 4.595845441825422, |
| "grad_norm": 0.6394187211990356, |
| "learning_rate": 0.0003247085443378946, |
| "loss": 3.3838, |
| "step": 42700 |
| }, |
| { |
| "epoch": 4.601226993865031, |
| "grad_norm": 0.604969322681427, |
| "learning_rate": 0.0003243853033078332, |
| "loss": 3.4029, |
| "step": 42750 |
| }, |
| { |
| "epoch": 4.606608545904638, |
| "grad_norm": 0.6396239995956421, |
| "learning_rate": 0.0003240620622777717, |
| "loss": 3.3879, |
| "step": 42800 |
| }, |
| { |
| "epoch": 4.611990097944247, |
| "grad_norm": 0.625914454460144, |
| "learning_rate": 0.00032373882124771036, |
| "loss": 3.39, |
| "step": 42850 |
| }, |
| { |
| "epoch": 4.6173716499838555, |
| "grad_norm": 0.6272422671318054, |
| "learning_rate": 0.00032341558021764896, |
| "loss": 3.3859, |
| "step": 42900 |
| }, |
| { |
| "epoch": 4.622753202023463, |
| "grad_norm": 0.628781795501709, |
| "learning_rate": 0.0003230923391875875, |
| "loss": 3.3976, |
| "step": 42950 |
| }, |
| { |
| "epoch": 4.628134754063072, |
| "grad_norm": 0.690066933631897, |
| "learning_rate": 0.0003227690981575261, |
| "loss": 3.3775, |
| "step": 43000 |
| }, |
| { |
| "epoch": 4.628134754063072, |
| "eval_accuracy": 0.3786348872905566, |
| "eval_loss": 3.4256374835968018, |
| "eval_runtime": 179.7124, |
| "eval_samples_per_second": 100.221, |
| "eval_steps_per_second": 6.266, |
| "step": 43000 |
| }, |
| { |
| "epoch": 4.63351630610268, |
| "grad_norm": 0.6105633974075317, |
| "learning_rate": 0.0003224458571274647, |
| "loss": 3.4001, |
| "step": 43050 |
| }, |
| { |
| "epoch": 4.638897858142288, |
| "grad_norm": 0.6365822553634644, |
| "learning_rate": 0.00032212261609740323, |
| "loss": 3.3949, |
| "step": 43100 |
| }, |
| { |
| "epoch": 4.6442794101818965, |
| "grad_norm": 0.6391448378562927, |
| "learning_rate": 0.0003217993750673419, |
| "loss": 3.3986, |
| "step": 43150 |
| }, |
| { |
| "epoch": 4.649660962221505, |
| "grad_norm": 0.645176351070404, |
| "learning_rate": 0.00032147613403728047, |
| "loss": 3.3902, |
| "step": 43200 |
| }, |
| { |
| "epoch": 4.655042514261113, |
| "grad_norm": 0.6706348657608032, |
| "learning_rate": 0.000321152893007219, |
| "loss": 3.3902, |
| "step": 43250 |
| }, |
| { |
| "epoch": 4.660424066300721, |
| "grad_norm": 0.6370499730110168, |
| "learning_rate": 0.0003208296519771576, |
| "loss": 3.4003, |
| "step": 43300 |
| }, |
| { |
| "epoch": 4.665805618340329, |
| "grad_norm": 0.6663179397583008, |
| "learning_rate": 0.00032050641094709615, |
| "loss": 3.4038, |
| "step": 43350 |
| }, |
| { |
| "epoch": 4.6711871703799375, |
| "grad_norm": 0.6630804538726807, |
| "learning_rate": 0.00032018316991703474, |
| "loss": 3.3995, |
| "step": 43400 |
| }, |
| { |
| "epoch": 4.676568722419546, |
| "grad_norm": 0.6663694977760315, |
| "learning_rate": 0.0003198599288869734, |
| "loss": 3.3829, |
| "step": 43450 |
| }, |
| { |
| "epoch": 4.681950274459154, |
| "grad_norm": 0.6295120120048523, |
| "learning_rate": 0.00031953668785691193, |
| "loss": 3.4082, |
| "step": 43500 |
| }, |
| { |
| "epoch": 4.687331826498762, |
| "grad_norm": 0.6530965566635132, |
| "learning_rate": 0.0003192134468268505, |
| "loss": 3.4099, |
| "step": 43550 |
| }, |
| { |
| "epoch": 4.692713378538371, |
| "grad_norm": 0.6678366661071777, |
| "learning_rate": 0.0003188902057967891, |
| "loss": 3.3738, |
| "step": 43600 |
| }, |
| { |
| "epoch": 4.6980949305779784, |
| "grad_norm": 0.650449275970459, |
| "learning_rate": 0.00031856696476672766, |
| "loss": 3.4114, |
| "step": 43650 |
| }, |
| { |
| "epoch": 4.703476482617587, |
| "grad_norm": 0.6517276763916016, |
| "learning_rate": 0.0003182437237366663, |
| "loss": 3.3892, |
| "step": 43700 |
| }, |
| { |
| "epoch": 4.7088580346571955, |
| "grad_norm": 0.651042103767395, |
| "learning_rate": 0.0003179204827066049, |
| "loss": 3.4098, |
| "step": 43750 |
| }, |
| { |
| "epoch": 4.714239586696803, |
| "grad_norm": 0.6547741293907166, |
| "learning_rate": 0.00031759724167654344, |
| "loss": 3.3836, |
| "step": 43800 |
| }, |
| { |
| "epoch": 4.719621138736412, |
| "grad_norm": 0.6325539946556091, |
| "learning_rate": 0.00031727400064648204, |
| "loss": 3.387, |
| "step": 43850 |
| }, |
| { |
| "epoch": 4.725002690776019, |
| "grad_norm": 0.6404955387115479, |
| "learning_rate": 0.0003169507596164206, |
| "loss": 3.3987, |
| "step": 43900 |
| }, |
| { |
| "epoch": 4.730384242815628, |
| "grad_norm": 0.6595001220703125, |
| "learning_rate": 0.0003166275185863592, |
| "loss": 3.3737, |
| "step": 43950 |
| }, |
| { |
| "epoch": 4.7357657948552365, |
| "grad_norm": 0.6610948443412781, |
| "learning_rate": 0.0003163042775562978, |
| "loss": 3.3787, |
| "step": 44000 |
| }, |
| { |
| "epoch": 4.7357657948552365, |
| "eval_accuracy": 0.37884719503334613, |
| "eval_loss": 3.4220328330993652, |
| "eval_runtime": 180.1789, |
| "eval_samples_per_second": 99.962, |
| "eval_steps_per_second": 6.249, |
| "step": 44000 |
| }, |
| { |
| "epoch": 4.741147346894844, |
| "grad_norm": 0.7104712724685669, |
| "learning_rate": 0.00031598103652623636, |
| "loss": 3.3982, |
| "step": 44050 |
| }, |
| { |
| "epoch": 4.746528898934453, |
| "grad_norm": 0.6293849945068359, |
| "learning_rate": 0.00031565779549617496, |
| "loss": 3.3637, |
| "step": 44100 |
| }, |
| { |
| "epoch": 4.751910450974061, |
| "grad_norm": 0.6018164753913879, |
| "learning_rate": 0.00031533455446611355, |
| "loss": 3.3908, |
| "step": 44150 |
| }, |
| { |
| "epoch": 4.757292003013669, |
| "grad_norm": 0.646990180015564, |
| "learning_rate": 0.0003150113134360521, |
| "loss": 3.3998, |
| "step": 44200 |
| }, |
| { |
| "epoch": 4.7626735550532775, |
| "grad_norm": 0.6214666366577148, |
| "learning_rate": 0.0003146880724059907, |
| "loss": 3.3958, |
| "step": 44250 |
| }, |
| { |
| "epoch": 4.768055107092886, |
| "grad_norm": 0.652014434337616, |
| "learning_rate": 0.00031436483137592934, |
| "loss": 3.3836, |
| "step": 44300 |
| }, |
| { |
| "epoch": 4.773436659132494, |
| "grad_norm": 0.6798874139785767, |
| "learning_rate": 0.0003140415903458679, |
| "loss": 3.3979, |
| "step": 44350 |
| }, |
| { |
| "epoch": 4.778818211172102, |
| "grad_norm": 0.6357735991477966, |
| "learning_rate": 0.00031371834931580647, |
| "loss": 3.3785, |
| "step": 44400 |
| }, |
| { |
| "epoch": 4.78419976321171, |
| "grad_norm": 0.661415696144104, |
| "learning_rate": 0.000313395108285745, |
| "loss": 3.3966, |
| "step": 44450 |
| }, |
| { |
| "epoch": 4.7895813152513185, |
| "grad_norm": 0.6242609024047852, |
| "learning_rate": 0.0003130718672556836, |
| "loss": 3.3797, |
| "step": 44500 |
| }, |
| { |
| "epoch": 4.794962867290927, |
| "grad_norm": 0.6368847489356995, |
| "learning_rate": 0.0003127486262256222, |
| "loss": 3.3922, |
| "step": 44550 |
| }, |
| { |
| "epoch": 4.800344419330535, |
| "grad_norm": 0.626862108707428, |
| "learning_rate": 0.0003124253851955608, |
| "loss": 3.3845, |
| "step": 44600 |
| }, |
| { |
| "epoch": 4.805725971370143, |
| "grad_norm": 0.6585181355476379, |
| "learning_rate": 0.0003121021441654994, |
| "loss": 3.411, |
| "step": 44650 |
| }, |
| { |
| "epoch": 4.811107523409751, |
| "grad_norm": 0.6374171376228333, |
| "learning_rate": 0.000311778903135438, |
| "loss": 3.3719, |
| "step": 44700 |
| }, |
| { |
| "epoch": 4.8164890754493594, |
| "grad_norm": 0.6377184987068176, |
| "learning_rate": 0.0003114556621053765, |
| "loss": 3.4087, |
| "step": 44750 |
| }, |
| { |
| "epoch": 4.821870627488968, |
| "grad_norm": 0.6534616351127625, |
| "learning_rate": 0.00031113888589591633, |
| "loss": 3.3769, |
| "step": 44800 |
| }, |
| { |
| "epoch": 4.827252179528576, |
| "grad_norm": 0.6503520607948303, |
| "learning_rate": 0.00031081564486585493, |
| "loss": 3.3823, |
| "step": 44850 |
| }, |
| { |
| "epoch": 4.832633731568184, |
| "grad_norm": 0.6115184426307678, |
| "learning_rate": 0.0003104924038357935, |
| "loss": 3.3727, |
| "step": 44900 |
| }, |
| { |
| "epoch": 4.838015283607793, |
| "grad_norm": 0.6773943901062012, |
| "learning_rate": 0.0003101691628057321, |
| "loss": 3.382, |
| "step": 44950 |
| }, |
| { |
| "epoch": 4.8433968356474, |
| "grad_norm": 0.64551842212677, |
| "learning_rate": 0.0003098459217756707, |
| "loss": 3.3814, |
| "step": 45000 |
| }, |
| { |
| "epoch": 4.8433968356474, |
| "eval_accuracy": 0.37974640633289874, |
| "eval_loss": 3.416212558746338, |
| "eval_runtime": 179.8221, |
| "eval_samples_per_second": 100.16, |
| "eval_steps_per_second": 6.262, |
| "step": 45000 |
| }, |
| { |
| "epoch": 4.848778387687009, |
| "grad_norm": 0.620168924331665, |
| "learning_rate": 0.0003095226807456093, |
| "loss": 3.3979, |
| "step": 45050 |
| }, |
| { |
| "epoch": 4.8541599397266175, |
| "grad_norm": 0.6280816197395325, |
| "learning_rate": 0.00030919943971554785, |
| "loss": 3.3942, |
| "step": 45100 |
| }, |
| { |
| "epoch": 4.859541491766225, |
| "grad_norm": 0.6292531490325928, |
| "learning_rate": 0.00030888266350608765, |
| "loss": 3.3716, |
| "step": 45150 |
| }, |
| { |
| "epoch": 4.864923043805834, |
| "grad_norm": 0.64167320728302, |
| "learning_rate": 0.00030855942247602625, |
| "loss": 3.3863, |
| "step": 45200 |
| }, |
| { |
| "epoch": 4.870304595845441, |
| "grad_norm": 0.6454716920852661, |
| "learning_rate": 0.00030823618144596484, |
| "loss": 3.3806, |
| "step": 45250 |
| }, |
| { |
| "epoch": 4.87568614788505, |
| "grad_norm": 0.6556320190429688, |
| "learning_rate": 0.0003079129404159035, |
| "loss": 3.4006, |
| "step": 45300 |
| }, |
| { |
| "epoch": 4.8810676999246585, |
| "grad_norm": 0.7040285468101501, |
| "learning_rate": 0.00030758969938584203, |
| "loss": 3.3745, |
| "step": 45350 |
| }, |
| { |
| "epoch": 4.886449251964266, |
| "grad_norm": 0.6378863453865051, |
| "learning_rate": 0.0003072664583557806, |
| "loss": 3.3978, |
| "step": 45400 |
| }, |
| { |
| "epoch": 4.891830804003875, |
| "grad_norm": 0.654747486114502, |
| "learning_rate": 0.00030694321732571917, |
| "loss": 3.4139, |
| "step": 45450 |
| }, |
| { |
| "epoch": 4.897212356043483, |
| "grad_norm": 0.7145013213157654, |
| "learning_rate": 0.00030661997629565776, |
| "loss": 3.3888, |
| "step": 45500 |
| }, |
| { |
| "epoch": 4.902593908083091, |
| "grad_norm": 0.7178427577018738, |
| "learning_rate": 0.0003062967352655963, |
| "loss": 3.3789, |
| "step": 45550 |
| }, |
| { |
| "epoch": 4.9079754601226995, |
| "grad_norm": 0.6693344712257385, |
| "learning_rate": 0.00030597349423553495, |
| "loss": 3.3774, |
| "step": 45600 |
| }, |
| { |
| "epoch": 4.913357012162308, |
| "grad_norm": 0.6698407530784607, |
| "learning_rate": 0.00030565025320547355, |
| "loss": 3.3774, |
| "step": 45650 |
| }, |
| { |
| "epoch": 4.918738564201916, |
| "grad_norm": 0.625595211982727, |
| "learning_rate": 0.0003053270121754121, |
| "loss": 3.3888, |
| "step": 45700 |
| }, |
| { |
| "epoch": 4.924120116241524, |
| "grad_norm": 0.6924371719360352, |
| "learning_rate": 0.0003050037711453507, |
| "loss": 3.3928, |
| "step": 45750 |
| }, |
| { |
| "epoch": 4.929501668281132, |
| "grad_norm": 0.6393285989761353, |
| "learning_rate": 0.0003046805301152893, |
| "loss": 3.3896, |
| "step": 45800 |
| }, |
| { |
| "epoch": 4.9348832203207404, |
| "grad_norm": 0.6794965267181396, |
| "learning_rate": 0.0003043572890852278, |
| "loss": 3.3805, |
| "step": 45850 |
| }, |
| { |
| "epoch": 4.940264772360349, |
| "grad_norm": 0.6193264722824097, |
| "learning_rate": 0.00030403404805516646, |
| "loss": 3.3845, |
| "step": 45900 |
| }, |
| { |
| "epoch": 4.945646324399957, |
| "grad_norm": 0.8187567591667175, |
| "learning_rate": 0.00030371080702510506, |
| "loss": 3.3807, |
| "step": 45950 |
| }, |
| { |
| "epoch": 4.951027876439565, |
| "grad_norm": 0.7009637951850891, |
| "learning_rate": 0.0003033875659950436, |
| "loss": 3.3892, |
| "step": 46000 |
| }, |
| { |
| "epoch": 4.951027876439565, |
| "eval_accuracy": 0.38006943136837124, |
| "eval_loss": 3.4113309383392334, |
| "eval_runtime": 179.7172, |
| "eval_samples_per_second": 100.219, |
| "eval_steps_per_second": 6.265, |
| "step": 46000 |
| }, |
| { |
| "epoch": 4.956409428479174, |
| "grad_norm": 0.7120588421821594, |
| "learning_rate": 0.0003030643249649822, |
| "loss": 3.3983, |
| "step": 46050 |
| }, |
| { |
| "epoch": 4.961790980518781, |
| "grad_norm": 0.66907799243927, |
| "learning_rate": 0.00030274108393492073, |
| "loss": 3.3768, |
| "step": 46100 |
| }, |
| { |
| "epoch": 4.96717253255839, |
| "grad_norm": 0.6506483554840088, |
| "learning_rate": 0.00030241784290485933, |
| "loss": 3.3993, |
| "step": 46150 |
| }, |
| { |
| "epoch": 4.9725540845979985, |
| "grad_norm": 0.6320881843566895, |
| "learning_rate": 0.000302094601874798, |
| "loss": 3.4016, |
| "step": 46200 |
| }, |
| { |
| "epoch": 4.977935636637606, |
| "grad_norm": 0.6613715291023254, |
| "learning_rate": 0.0003017713608447365, |
| "loss": 3.3842, |
| "step": 46250 |
| }, |
| { |
| "epoch": 4.983317188677215, |
| "grad_norm": 0.6871373653411865, |
| "learning_rate": 0.0003014481198146751, |
| "loss": 3.39, |
| "step": 46300 |
| }, |
| { |
| "epoch": 4.988698740716822, |
| "grad_norm": 0.6517271399497986, |
| "learning_rate": 0.0003011248787846137, |
| "loss": 3.3789, |
| "step": 46350 |
| }, |
| { |
| "epoch": 4.994080292756431, |
| "grad_norm": 0.6434972882270813, |
| "learning_rate": 0.00030080163775455225, |
| "loss": 3.4122, |
| "step": 46400 |
| }, |
| { |
| "epoch": 4.9994618447960395, |
| "grad_norm": 0.6256486773490906, |
| "learning_rate": 0.0003004783967244909, |
| "loss": 3.4034, |
| "step": 46450 |
| }, |
| { |
| "epoch": 5.004843396835647, |
| "grad_norm": 0.6456863284111023, |
| "learning_rate": 0.0003001551556944295, |
| "loss": 3.3024, |
| "step": 46500 |
| }, |
| { |
| "epoch": 5.010224948875256, |
| "grad_norm": 0.7082270383834839, |
| "learning_rate": 0.00029983191466436803, |
| "loss": 3.3113, |
| "step": 46550 |
| }, |
| { |
| "epoch": 5.015606500914864, |
| "grad_norm": 0.6580175161361694, |
| "learning_rate": 0.0002995086736343066, |
| "loss": 3.2953, |
| "step": 46600 |
| }, |
| { |
| "epoch": 5.020988052954472, |
| "grad_norm": 0.6615280508995056, |
| "learning_rate": 0.00029918543260424517, |
| "loss": 3.2981, |
| "step": 46650 |
| }, |
| { |
| "epoch": 5.0263696049940805, |
| "grad_norm": 0.6606518030166626, |
| "learning_rate": 0.0002988621915741838, |
| "loss": 3.2933, |
| "step": 46700 |
| }, |
| { |
| "epoch": 5.031751157033688, |
| "grad_norm": 0.6257070302963257, |
| "learning_rate": 0.00029853895054412236, |
| "loss": 3.2914, |
| "step": 46750 |
| }, |
| { |
| "epoch": 5.037132709073297, |
| "grad_norm": 0.6961430311203003, |
| "learning_rate": 0.00029821570951406095, |
| "loss": 3.3039, |
| "step": 46800 |
| }, |
| { |
| "epoch": 5.042514261112905, |
| "grad_norm": 0.6646069288253784, |
| "learning_rate": 0.00029789246848399955, |
| "loss": 3.3104, |
| "step": 46850 |
| }, |
| { |
| "epoch": 5.047895813152513, |
| "grad_norm": 0.6522386074066162, |
| "learning_rate": 0.00029756922745393814, |
| "loss": 3.2756, |
| "step": 46900 |
| }, |
| { |
| "epoch": 5.0532773651921215, |
| "grad_norm": 0.6988029479980469, |
| "learning_rate": 0.00029724598642387674, |
| "loss": 3.32, |
| "step": 46950 |
| }, |
| { |
| "epoch": 5.05865891723173, |
| "grad_norm": 0.6169154644012451, |
| "learning_rate": 0.0002969227453938153, |
| "loss": 3.3194, |
| "step": 47000 |
| }, |
| { |
| "epoch": 5.05865891723173, |
| "eval_accuracy": 0.3803127051841878, |
| "eval_loss": 3.414651393890381, |
| "eval_runtime": 179.7017, |
| "eval_samples_per_second": 100.227, |
| "eval_steps_per_second": 6.266, |
| "step": 47000 |
| }, |
| { |
| "epoch": 5.064040469271338, |
| "grad_norm": 0.6657958030700684, |
| "learning_rate": 0.00029659950436375387, |
| "loss": 3.3005, |
| "step": 47050 |
| }, |
| { |
| "epoch": 5.069422021310946, |
| "grad_norm": 0.6399413347244263, |
| "learning_rate": 0.00029627626333369246, |
| "loss": 3.301, |
| "step": 47100 |
| }, |
| { |
| "epoch": 5.074803573350554, |
| "grad_norm": 0.6759291887283325, |
| "learning_rate": 0.00029595302230363106, |
| "loss": 3.3139, |
| "step": 47150 |
| }, |
| { |
| "epoch": 5.080185125390162, |
| "grad_norm": 0.6591566205024719, |
| "learning_rate": 0.0002956297812735696, |
| "loss": 3.3166, |
| "step": 47200 |
| }, |
| { |
| "epoch": 5.085566677429771, |
| "grad_norm": 0.6507174372673035, |
| "learning_rate": 0.00029530654024350825, |
| "loss": 3.3268, |
| "step": 47250 |
| }, |
| { |
| "epoch": 5.090948229469379, |
| "grad_norm": 0.6434475779533386, |
| "learning_rate": 0.0002949832992134468, |
| "loss": 3.2906, |
| "step": 47300 |
| }, |
| { |
| "epoch": 5.096329781508987, |
| "grad_norm": 0.6955178380012512, |
| "learning_rate": 0.0002946600581833854, |
| "loss": 3.3277, |
| "step": 47350 |
| }, |
| { |
| "epoch": 5.101711333548596, |
| "grad_norm": 0.6766684055328369, |
| "learning_rate": 0.000294336817153324, |
| "loss": 3.3073, |
| "step": 47400 |
| }, |
| { |
| "epoch": 5.107092885588203, |
| "grad_norm": 0.655519962310791, |
| "learning_rate": 0.0002940135761232626, |
| "loss": 3.3137, |
| "step": 47450 |
| }, |
| { |
| "epoch": 5.112474437627812, |
| "grad_norm": 0.6430268287658691, |
| "learning_rate": 0.0002936903350932011, |
| "loss": 3.3108, |
| "step": 47500 |
| }, |
| { |
| "epoch": 5.1178559896674205, |
| "grad_norm": 0.6219215989112854, |
| "learning_rate": 0.0002933670940631397, |
| "loss": 3.2839, |
| "step": 47550 |
| }, |
| { |
| "epoch": 5.123237541707028, |
| "grad_norm": 0.7049906253814697, |
| "learning_rate": 0.0002930438530330783, |
| "loss": 3.3125, |
| "step": 47600 |
| }, |
| { |
| "epoch": 5.128619093746637, |
| "grad_norm": 0.630456805229187, |
| "learning_rate": 0.0002927206120030169, |
| "loss": 3.3006, |
| "step": 47650 |
| }, |
| { |
| "epoch": 5.134000645786244, |
| "grad_norm": 0.676864504814148, |
| "learning_rate": 0.0002923973709729555, |
| "loss": 3.3178, |
| "step": 47700 |
| }, |
| { |
| "epoch": 5.139382197825853, |
| "grad_norm": 0.6149758696556091, |
| "learning_rate": 0.00029207412994289403, |
| "loss": 3.3064, |
| "step": 47750 |
| }, |
| { |
| "epoch": 5.1447637498654615, |
| "grad_norm": 0.6830527782440186, |
| "learning_rate": 0.0002917508889128327, |
| "loss": 3.3164, |
| "step": 47800 |
| }, |
| { |
| "epoch": 5.150145301905069, |
| "grad_norm": 0.657446563243866, |
| "learning_rate": 0.0002914276478827712, |
| "loss": 3.2956, |
| "step": 47850 |
| }, |
| { |
| "epoch": 5.155526853944678, |
| "grad_norm": 0.67881840467453, |
| "learning_rate": 0.0002911044068527098, |
| "loss": 3.3083, |
| "step": 47900 |
| }, |
| { |
| "epoch": 5.160908405984286, |
| "grad_norm": 0.6833150386810303, |
| "learning_rate": 0.0002907811658226484, |
| "loss": 3.3295, |
| "step": 47950 |
| }, |
| { |
| "epoch": 5.166289958023894, |
| "grad_norm": 0.7698062062263489, |
| "learning_rate": 0.000290457924792587, |
| "loss": 3.306, |
| "step": 48000 |
| }, |
| { |
| "epoch": 5.166289958023894, |
| "eval_accuracy": 0.38088324149810165, |
| "eval_loss": 3.413996934890747, |
| "eval_runtime": 180.0152, |
| "eval_samples_per_second": 100.053, |
| "eval_steps_per_second": 6.255, |
| "step": 48000 |
| }, |
| { |
| "epoch": 5.1716715100635025, |
| "grad_norm": 0.6625930666923523, |
| "learning_rate": 0.00029013468376252555, |
| "loss": 3.3147, |
| "step": 48050 |
| }, |
| { |
| "epoch": 5.17705306210311, |
| "grad_norm": 0.6521696448326111, |
| "learning_rate": 0.00028981144273246414, |
| "loss": 3.3087, |
| "step": 48100 |
| }, |
| { |
| "epoch": 5.182434614142719, |
| "grad_norm": 0.6896630525588989, |
| "learning_rate": 0.00028948820170240274, |
| "loss": 3.3195, |
| "step": 48150 |
| }, |
| { |
| "epoch": 5.187816166182327, |
| "grad_norm": 0.6332021951675415, |
| "learning_rate": 0.00028916496067234133, |
| "loss": 3.3183, |
| "step": 48200 |
| }, |
| { |
| "epoch": 5.193197718221935, |
| "grad_norm": 0.6617722511291504, |
| "learning_rate": 0.0002888417196422799, |
| "loss": 3.3017, |
| "step": 48250 |
| }, |
| { |
| "epoch": 5.198579270261543, |
| "grad_norm": 0.6539857983589172, |
| "learning_rate": 0.00028851847861221847, |
| "loss": 3.3423, |
| "step": 48300 |
| }, |
| { |
| "epoch": 5.203960822301152, |
| "grad_norm": 0.6800119280815125, |
| "learning_rate": 0.00028819523758215706, |
| "loss": 3.3171, |
| "step": 48350 |
| }, |
| { |
| "epoch": 5.20934237434076, |
| "grad_norm": 0.6877304315567017, |
| "learning_rate": 0.00028787199655209566, |
| "loss": 3.3214, |
| "step": 48400 |
| }, |
| { |
| "epoch": 5.214723926380368, |
| "grad_norm": 0.6638197302818298, |
| "learning_rate": 0.00028754875552203425, |
| "loss": 3.309, |
| "step": 48450 |
| }, |
| { |
| "epoch": 5.220105478419977, |
| "grad_norm": 0.6807978749275208, |
| "learning_rate": 0.0002872255144919728, |
| "loss": 3.3352, |
| "step": 48500 |
| }, |
| { |
| "epoch": 5.225487030459584, |
| "grad_norm": 0.6976240277290344, |
| "learning_rate": 0.00028690227346191144, |
| "loss": 3.3248, |
| "step": 48550 |
| }, |
| { |
| "epoch": 5.230868582499193, |
| "grad_norm": 0.713167130947113, |
| "learning_rate": 0.00028657903243185, |
| "loss": 3.3203, |
| "step": 48600 |
| }, |
| { |
| "epoch": 5.236250134538801, |
| "grad_norm": 0.7144914269447327, |
| "learning_rate": 0.0002862557914017886, |
| "loss": 3.3233, |
| "step": 48650 |
| }, |
| { |
| "epoch": 5.241631686578409, |
| "grad_norm": 0.679750919342041, |
| "learning_rate": 0.00028593255037172717, |
| "loss": 3.3142, |
| "step": 48700 |
| }, |
| { |
| "epoch": 5.247013238618018, |
| "grad_norm": 0.6690563559532166, |
| "learning_rate": 0.00028560930934166576, |
| "loss": 3.3248, |
| "step": 48750 |
| }, |
| { |
| "epoch": 5.252394790657625, |
| "grad_norm": 0.6724905967712402, |
| "learning_rate": 0.00028528606831160436, |
| "loss": 3.3239, |
| "step": 48800 |
| }, |
| { |
| "epoch": 5.257776342697234, |
| "grad_norm": 0.6915743947029114, |
| "learning_rate": 0.0002849628272815429, |
| "loss": 3.3224, |
| "step": 48850 |
| }, |
| { |
| "epoch": 5.2631578947368425, |
| "grad_norm": 0.7154117822647095, |
| "learning_rate": 0.0002846395862514815, |
| "loss": 3.3256, |
| "step": 48900 |
| }, |
| { |
| "epoch": 5.26853944677645, |
| "grad_norm": 0.6927023530006409, |
| "learning_rate": 0.0002843163452214201, |
| "loss": 3.332, |
| "step": 48950 |
| }, |
| { |
| "epoch": 5.273920998816059, |
| "grad_norm": 0.7423051595687866, |
| "learning_rate": 0.0002839931041913587, |
| "loss": 3.3269, |
| "step": 49000 |
| }, |
| { |
| "epoch": 5.273920998816059, |
| "eval_accuracy": 0.38093919773532614, |
| "eval_loss": 3.410620927810669, |
| "eval_runtime": 179.5074, |
| "eval_samples_per_second": 100.336, |
| "eval_steps_per_second": 6.273, |
| "step": 49000 |
| }, |
| { |
| "epoch": 5.279302550855666, |
| "grad_norm": 0.6497479677200317, |
| "learning_rate": 0.0002836698631612972, |
| "loss": 3.3177, |
| "step": 49050 |
| }, |
| { |
| "epoch": 5.284684102895275, |
| "grad_norm": 0.7746565937995911, |
| "learning_rate": 0.00028334662213123587, |
| "loss": 3.3411, |
| "step": 49100 |
| }, |
| { |
| "epoch": 5.2900656549348835, |
| "grad_norm": 0.702002763748169, |
| "learning_rate": 0.0002830233811011744, |
| "loss": 3.3387, |
| "step": 49150 |
| }, |
| { |
| "epoch": 5.295447206974491, |
| "grad_norm": 0.636340320110321, |
| "learning_rate": 0.0002827066048917142, |
| "loss": 3.3153, |
| "step": 49200 |
| }, |
| { |
| "epoch": 5.3008287590141, |
| "grad_norm": 0.7339147329330444, |
| "learning_rate": 0.0002823833638616528, |
| "loss": 3.3402, |
| "step": 49250 |
| }, |
| { |
| "epoch": 5.306210311053708, |
| "grad_norm": 0.6532171964645386, |
| "learning_rate": 0.0002820601228315914, |
| "loss": 3.3329, |
| "step": 49300 |
| }, |
| { |
| "epoch": 5.311591863093316, |
| "grad_norm": 0.7197509407997131, |
| "learning_rate": 0.00028173688180153, |
| "loss": 3.3392, |
| "step": 49350 |
| }, |
| { |
| "epoch": 5.316973415132924, |
| "grad_norm": 0.7137599587440491, |
| "learning_rate": 0.00028141364077146854, |
| "loss": 3.3024, |
| "step": 49400 |
| }, |
| { |
| "epoch": 5.322354967172533, |
| "grad_norm": 0.6693756580352783, |
| "learning_rate": 0.00028109039974140714, |
| "loss": 3.3333, |
| "step": 49450 |
| }, |
| { |
| "epoch": 5.327736519212141, |
| "grad_norm": 0.6450282335281372, |
| "learning_rate": 0.00028076715871134573, |
| "loss": 3.3379, |
| "step": 49500 |
| }, |
| { |
| "epoch": 5.333118071251749, |
| "grad_norm": 0.7221931219100952, |
| "learning_rate": 0.00028044391768128433, |
| "loss": 3.3119, |
| "step": 49550 |
| }, |
| { |
| "epoch": 5.338499623291357, |
| "grad_norm": 0.7065942287445068, |
| "learning_rate": 0.0002801206766512229, |
| "loss": 3.3233, |
| "step": 49600 |
| }, |
| { |
| "epoch": 5.343881175330965, |
| "grad_norm": 0.6537820100784302, |
| "learning_rate": 0.0002797974356211615, |
| "loss": 3.3253, |
| "step": 49650 |
| }, |
| { |
| "epoch": 5.349262727370574, |
| "grad_norm": 0.6890376210212708, |
| "learning_rate": 0.00027947419459110006, |
| "loss": 3.3457, |
| "step": 49700 |
| }, |
| { |
| "epoch": 5.354644279410182, |
| "grad_norm": 0.6945475339889526, |
| "learning_rate": 0.00027915095356103865, |
| "loss": 3.3181, |
| "step": 49750 |
| }, |
| { |
| "epoch": 5.36002583144979, |
| "grad_norm": 0.6715689301490784, |
| "learning_rate": 0.00027882771253097725, |
| "loss": 3.3325, |
| "step": 49800 |
| }, |
| { |
| "epoch": 5.365407383489399, |
| "grad_norm": 0.6894263625144958, |
| "learning_rate": 0.00027850447150091584, |
| "loss": 3.345, |
| "step": 49850 |
| }, |
| { |
| "epoch": 5.370788935529006, |
| "grad_norm": 0.6533727645874023, |
| "learning_rate": 0.00027818123047085444, |
| "loss": 3.347, |
| "step": 49900 |
| }, |
| { |
| "epoch": 5.376170487568615, |
| "grad_norm": 0.7649930715560913, |
| "learning_rate": 0.000277857989440793, |
| "loss": 3.339, |
| "step": 49950 |
| }, |
| { |
| "epoch": 5.3815520396082235, |
| "grad_norm": 0.6463519930839539, |
| "learning_rate": 0.00027753474841073157, |
| "loss": 3.3199, |
| "step": 50000 |
| }, |
| { |
| "epoch": 5.3815520396082235, |
| "eval_accuracy": 0.3816024149625457, |
| "eval_loss": 3.4057817459106445, |
| "eval_runtime": 179.819, |
| "eval_samples_per_second": 100.162, |
| "eval_steps_per_second": 6.262, |
| "step": 50000 |
| }, |
| { |
| "epoch": 5.386933591647831, |
| "grad_norm": 0.6719322800636292, |
| "learning_rate": 0.00027721150738067017, |
| "loss": 3.335, |
| "step": 50050 |
| }, |
| { |
| "epoch": 5.39231514368744, |
| "grad_norm": 0.7052873373031616, |
| "learning_rate": 0.00027688826635060876, |
| "loss": 3.3272, |
| "step": 50100 |
| }, |
| { |
| "epoch": 5.397696695727047, |
| "grad_norm": 0.6729222536087036, |
| "learning_rate": 0.00027657149014114857, |
| "loss": 3.3287, |
| "step": 50150 |
| }, |
| { |
| "epoch": 5.403078247766656, |
| "grad_norm": 0.6733786463737488, |
| "learning_rate": 0.00027624824911108716, |
| "loss": 3.3366, |
| "step": 50200 |
| }, |
| { |
| "epoch": 5.4084597998062645, |
| "grad_norm": 0.6847931146621704, |
| "learning_rate": 0.0002759250080810257, |
| "loss": 3.3309, |
| "step": 50250 |
| }, |
| { |
| "epoch": 5.413841351845872, |
| "grad_norm": 0.6446109414100647, |
| "learning_rate": 0.0002756017670509643, |
| "loss": 3.3295, |
| "step": 50300 |
| }, |
| { |
| "epoch": 5.419222903885481, |
| "grad_norm": 0.7692701816558838, |
| "learning_rate": 0.0002752785260209029, |
| "loss": 3.3344, |
| "step": 50350 |
| }, |
| { |
| "epoch": 5.424604455925088, |
| "grad_norm": 0.7050197720527649, |
| "learning_rate": 0.0002749552849908415, |
| "loss": 3.3242, |
| "step": 50400 |
| }, |
| { |
| "epoch": 5.429986007964697, |
| "grad_norm": 0.7438011765480042, |
| "learning_rate": 0.0002746320439607801, |
| "loss": 3.341, |
| "step": 50450 |
| }, |
| { |
| "epoch": 5.435367560004305, |
| "grad_norm": 0.6816525459289551, |
| "learning_rate": 0.0002743088029307186, |
| "loss": 3.3404, |
| "step": 50500 |
| }, |
| { |
| "epoch": 5.440749112043913, |
| "grad_norm": 0.6949660778045654, |
| "learning_rate": 0.00027398556190065727, |
| "loss": 3.3538, |
| "step": 50550 |
| }, |
| { |
| "epoch": 5.446130664083522, |
| "grad_norm": 0.7127841114997864, |
| "learning_rate": 0.0002736623208705958, |
| "loss": 3.323, |
| "step": 50600 |
| }, |
| { |
| "epoch": 5.45151221612313, |
| "grad_norm": 0.7076895833015442, |
| "learning_rate": 0.0002733390798405344, |
| "loss": 3.3439, |
| "step": 50650 |
| }, |
| { |
| "epoch": 5.456893768162738, |
| "grad_norm": 0.6947559714317322, |
| "learning_rate": 0.000273015838810473, |
| "loss": 3.3573, |
| "step": 50700 |
| }, |
| { |
| "epoch": 5.462275320202346, |
| "grad_norm": 0.7545173764228821, |
| "learning_rate": 0.0002726925977804116, |
| "loss": 3.3316, |
| "step": 50750 |
| }, |
| { |
| "epoch": 5.467656872241955, |
| "grad_norm": 0.6546699404716492, |
| "learning_rate": 0.00027236935675035013, |
| "loss": 3.3368, |
| "step": 50800 |
| }, |
| { |
| "epoch": 5.473038424281563, |
| "grad_norm": 0.6620163917541504, |
| "learning_rate": 0.00027204611572028873, |
| "loss": 3.3503, |
| "step": 50850 |
| }, |
| { |
| "epoch": 5.478419976321171, |
| "grad_norm": 0.6684191226959229, |
| "learning_rate": 0.0002717228746902273, |
| "loss": 3.3254, |
| "step": 50900 |
| }, |
| { |
| "epoch": 5.483801528360779, |
| "grad_norm": 0.6487999558448792, |
| "learning_rate": 0.0002713996336601659, |
| "loss": 3.3357, |
| "step": 50950 |
| }, |
| { |
| "epoch": 5.489183080400387, |
| "grad_norm": 0.6728455424308777, |
| "learning_rate": 0.0002710763926301045, |
| "loss": 3.3263, |
| "step": 51000 |
| }, |
| { |
| "epoch": 5.489183080400387, |
| "eval_accuracy": 0.38202127184507056, |
| "eval_loss": 3.400080680847168, |
| "eval_runtime": 179.9731, |
| "eval_samples_per_second": 100.076, |
| "eval_steps_per_second": 6.256, |
| "step": 51000 |
| }, |
| { |
| "epoch": 5.494564632439996, |
| "grad_norm": 0.6715853810310364, |
| "learning_rate": 0.00027075315160004305, |
| "loss": 3.3271, |
| "step": 51050 |
| }, |
| { |
| "epoch": 5.499946184479604, |
| "grad_norm": 0.6638278961181641, |
| "learning_rate": 0.00027042991056998165, |
| "loss": 3.3207, |
| "step": 51100 |
| }, |
| { |
| "epoch": 5.505327736519212, |
| "grad_norm": 0.683484673500061, |
| "learning_rate": 0.00027010666953992024, |
| "loss": 3.3383, |
| "step": 51150 |
| }, |
| { |
| "epoch": 5.510709288558821, |
| "grad_norm": 0.6461046934127808, |
| "learning_rate": 0.00026978342850985884, |
| "loss": 3.3287, |
| "step": 51200 |
| }, |
| { |
| "epoch": 5.516090840598428, |
| "grad_norm": 0.6551113724708557, |
| "learning_rate": 0.0002694601874797974, |
| "loss": 3.3486, |
| "step": 51250 |
| }, |
| { |
| "epoch": 5.521472392638037, |
| "grad_norm": 0.681959331035614, |
| "learning_rate": 0.000269136946449736, |
| "loss": 3.3301, |
| "step": 51300 |
| }, |
| { |
| "epoch": 5.5268539446776455, |
| "grad_norm": 0.7404993772506714, |
| "learning_rate": 0.00026881370541967457, |
| "loss": 3.3349, |
| "step": 51350 |
| }, |
| { |
| "epoch": 5.532235496717253, |
| "grad_norm": 0.6710616946220398, |
| "learning_rate": 0.00026849046438961316, |
| "loss": 3.3252, |
| "step": 51400 |
| }, |
| { |
| "epoch": 5.537617048756862, |
| "grad_norm": 0.670371949672699, |
| "learning_rate": 0.00026816722335955176, |
| "loss": 3.3355, |
| "step": 51450 |
| }, |
| { |
| "epoch": 5.542998600796469, |
| "grad_norm": 0.6416595578193665, |
| "learning_rate": 0.00026784398232949035, |
| "loss": 3.3277, |
| "step": 51500 |
| }, |
| { |
| "epoch": 5.548380152836078, |
| "grad_norm": 0.7234741449356079, |
| "learning_rate": 0.00026752074129942895, |
| "loss": 3.3327, |
| "step": 51550 |
| }, |
| { |
| "epoch": 5.553761704875686, |
| "grad_norm": 0.6543775796890259, |
| "learning_rate": 0.0002671975002693675, |
| "loss": 3.3349, |
| "step": 51600 |
| }, |
| { |
| "epoch": 5.559143256915294, |
| "grad_norm": 0.6863155961036682, |
| "learning_rate": 0.0002668742592393061, |
| "loss": 3.3377, |
| "step": 51650 |
| }, |
| { |
| "epoch": 5.564524808954903, |
| "grad_norm": 0.6523829102516174, |
| "learning_rate": 0.0002665510182092447, |
| "loss": 3.3378, |
| "step": 51700 |
| }, |
| { |
| "epoch": 5.569906360994511, |
| "grad_norm": 0.6852605938911438, |
| "learning_rate": 0.00026622777717918327, |
| "loss": 3.3293, |
| "step": 51750 |
| }, |
| { |
| "epoch": 5.575287913034119, |
| "grad_norm": 0.7273456454277039, |
| "learning_rate": 0.0002659045361491218, |
| "loss": 3.3322, |
| "step": 51800 |
| }, |
| { |
| "epoch": 5.580669465073727, |
| "grad_norm": 0.671544075012207, |
| "learning_rate": 0.00026558129511906046, |
| "loss": 3.326, |
| "step": 51850 |
| }, |
| { |
| "epoch": 5.586051017113336, |
| "grad_norm": 0.6995226740837097, |
| "learning_rate": 0.000265258054088999, |
| "loss": 3.3282, |
| "step": 51900 |
| }, |
| { |
| "epoch": 5.591432569152944, |
| "grad_norm": 0.6662087440490723, |
| "learning_rate": 0.0002649348130589376, |
| "loss": 3.3384, |
| "step": 51950 |
| }, |
| { |
| "epoch": 5.596814121192552, |
| "grad_norm": 0.90606689453125, |
| "learning_rate": 0.0002646115720288762, |
| "loss": 3.3439, |
| "step": 52000 |
| }, |
| { |
| "epoch": 5.596814121192552, |
| "eval_accuracy": 0.3824698996188565, |
| "eval_loss": 3.3940930366516113, |
| "eval_runtime": 180.0459, |
| "eval_samples_per_second": 100.036, |
| "eval_steps_per_second": 6.254, |
| "step": 52000 |
| }, |
| { |
| "epoch": 5.60219567323216, |
| "grad_norm": 0.670542299747467, |
| "learning_rate": 0.00026428833099881473, |
| "loss": 3.3299, |
| "step": 52050 |
| }, |
| { |
| "epoch": 5.607577225271768, |
| "grad_norm": 0.6939104795455933, |
| "learning_rate": 0.0002639650899687533, |
| "loss": 3.3458, |
| "step": 52100 |
| }, |
| { |
| "epoch": 5.612958777311377, |
| "grad_norm": 0.6510215997695923, |
| "learning_rate": 0.0002636418489386919, |
| "loss": 3.3294, |
| "step": 52150 |
| }, |
| { |
| "epoch": 5.618340329350985, |
| "grad_norm": 0.6817855834960938, |
| "learning_rate": 0.0002633186079086305, |
| "loss": 3.3445, |
| "step": 52200 |
| }, |
| { |
| "epoch": 5.623721881390593, |
| "grad_norm": 0.6868756413459778, |
| "learning_rate": 0.00026299536687856905, |
| "loss": 3.3462, |
| "step": 52250 |
| }, |
| { |
| "epoch": 5.629103433430201, |
| "grad_norm": 0.6777948141098022, |
| "learning_rate": 0.0002626721258485077, |
| "loss": 3.3293, |
| "step": 52300 |
| }, |
| { |
| "epoch": 5.634484985469809, |
| "grad_norm": 0.6776952743530273, |
| "learning_rate": 0.00026234888481844624, |
| "loss": 3.3273, |
| "step": 52350 |
| }, |
| { |
| "epoch": 5.639866537509418, |
| "grad_norm": 0.7230375409126282, |
| "learning_rate": 0.00026202564378838484, |
| "loss": 3.3376, |
| "step": 52400 |
| }, |
| { |
| "epoch": 5.645248089549026, |
| "grad_norm": 0.6413059830665588, |
| "learning_rate": 0.00026170240275832343, |
| "loss": 3.3498, |
| "step": 52450 |
| }, |
| { |
| "epoch": 5.650629641588634, |
| "grad_norm": 0.6590431332588196, |
| "learning_rate": 0.00026137916172826203, |
| "loss": 3.3422, |
| "step": 52500 |
| }, |
| { |
| "epoch": 5.656011193628243, |
| "grad_norm": 0.7001703977584839, |
| "learning_rate": 0.0002610559206982006, |
| "loss": 3.3453, |
| "step": 52550 |
| }, |
| { |
| "epoch": 5.66139274566785, |
| "grad_norm": 0.7514291405677795, |
| "learning_rate": 0.00026073267966813916, |
| "loss": 3.3427, |
| "step": 52600 |
| }, |
| { |
| "epoch": 5.666774297707459, |
| "grad_norm": 0.6938708424568176, |
| "learning_rate": 0.00026040943863807776, |
| "loss": 3.3422, |
| "step": 52650 |
| }, |
| { |
| "epoch": 5.672155849747067, |
| "grad_norm": 0.6692508459091187, |
| "learning_rate": 0.00026008619760801635, |
| "loss": 3.3259, |
| "step": 52700 |
| }, |
| { |
| "epoch": 5.677537401786675, |
| "grad_norm": 0.8175403475761414, |
| "learning_rate": 0.00025976295657795495, |
| "loss": 3.3254, |
| "step": 52750 |
| }, |
| { |
| "epoch": 5.682918953826284, |
| "grad_norm": 0.7053045034408569, |
| "learning_rate": 0.0002594397155478935, |
| "loss": 3.357, |
| "step": 52800 |
| }, |
| { |
| "epoch": 5.688300505865891, |
| "grad_norm": 0.7577919363975525, |
| "learning_rate": 0.00025911647451783214, |
| "loss": 3.3368, |
| "step": 52850 |
| }, |
| { |
| "epoch": 5.6936820579055, |
| "grad_norm": 0.6919559836387634, |
| "learning_rate": 0.0002587932334877707, |
| "loss": 3.3258, |
| "step": 52900 |
| }, |
| { |
| "epoch": 5.699063609945108, |
| "grad_norm": 0.7500218152999878, |
| "learning_rate": 0.00025846999245770927, |
| "loss": 3.3301, |
| "step": 52950 |
| }, |
| { |
| "epoch": 5.704445161984716, |
| "grad_norm": 0.7383402585983276, |
| "learning_rate": 0.00025814675142764787, |
| "loss": 3.3445, |
| "step": 53000 |
| }, |
| { |
| "epoch": 5.704445161984716, |
| "eval_accuracy": 0.38280150823246634, |
| "eval_loss": 3.3909990787506104, |
| "eval_runtime": 179.6204, |
| "eval_samples_per_second": 100.273, |
| "eval_steps_per_second": 6.269, |
| "step": 53000 |
| }, |
| { |
| "epoch": 5.709826714024325, |
| "grad_norm": 0.6965733170509338, |
| "learning_rate": 0.00025782351039758646, |
| "loss": 3.3443, |
| "step": 53050 |
| }, |
| { |
| "epoch": 5.715208266063933, |
| "grad_norm": 0.7147427797317505, |
| "learning_rate": 0.000257500269367525, |
| "loss": 3.3313, |
| "step": 53100 |
| }, |
| { |
| "epoch": 5.720589818103541, |
| "grad_norm": 0.692800760269165, |
| "learning_rate": 0.0002571770283374636, |
| "loss": 3.326, |
| "step": 53150 |
| }, |
| { |
| "epoch": 5.725971370143149, |
| "grad_norm": 0.6957950592041016, |
| "learning_rate": 0.0002568537873074022, |
| "loss": 3.338, |
| "step": 53200 |
| }, |
| { |
| "epoch": 5.731352922182758, |
| "grad_norm": 0.6565383076667786, |
| "learning_rate": 0.0002565305462773408, |
| "loss": 3.3311, |
| "step": 53250 |
| }, |
| { |
| "epoch": 5.736734474222366, |
| "grad_norm": 0.6763213872909546, |
| "learning_rate": 0.0002562073052472794, |
| "loss": 3.3413, |
| "step": 53300 |
| }, |
| { |
| "epoch": 5.742116026261974, |
| "grad_norm": 0.6820769309997559, |
| "learning_rate": 0.0002558840642172179, |
| "loss": 3.3296, |
| "step": 53350 |
| }, |
| { |
| "epoch": 5.747497578301582, |
| "grad_norm": 0.7125808000564575, |
| "learning_rate": 0.00025556082318715657, |
| "loss": 3.323, |
| "step": 53400 |
| }, |
| { |
| "epoch": 5.75287913034119, |
| "grad_norm": 0.6912208795547485, |
| "learning_rate": 0.0002552375821570951, |
| "loss": 3.3392, |
| "step": 53450 |
| }, |
| { |
| "epoch": 5.758260682380799, |
| "grad_norm": 0.6613236665725708, |
| "learning_rate": 0.0002549143411270337, |
| "loss": 3.3342, |
| "step": 53500 |
| }, |
| { |
| "epoch": 5.763642234420407, |
| "grad_norm": 0.681589663028717, |
| "learning_rate": 0.0002545911000969723, |
| "loss": 3.3559, |
| "step": 53550 |
| }, |
| { |
| "epoch": 5.769023786460015, |
| "grad_norm": 0.7078532576560974, |
| "learning_rate": 0.0002542678590669109, |
| "loss": 3.3391, |
| "step": 53600 |
| }, |
| { |
| "epoch": 5.774405338499624, |
| "grad_norm": 0.6990439891815186, |
| "learning_rate": 0.00025394461803684943, |
| "loss": 3.3236, |
| "step": 53650 |
| }, |
| { |
| "epoch": 5.779786890539231, |
| "grad_norm": 0.677062451839447, |
| "learning_rate": 0.00025362137700678803, |
| "loss": 3.3194, |
| "step": 53700 |
| }, |
| { |
| "epoch": 5.78516844257884, |
| "grad_norm": 0.8788602948188782, |
| "learning_rate": 0.0002532981359767266, |
| "loss": 3.3314, |
| "step": 53750 |
| }, |
| { |
| "epoch": 5.790549994618448, |
| "grad_norm": 0.7403194904327393, |
| "learning_rate": 0.0002529748949466652, |
| "loss": 3.3294, |
| "step": 53800 |
| }, |
| { |
| "epoch": 5.795931546658056, |
| "grad_norm": 0.7146193981170654, |
| "learning_rate": 0.0002526516539166038, |
| "loss": 3.3165, |
| "step": 53850 |
| }, |
| { |
| "epoch": 5.801313098697665, |
| "grad_norm": 0.7052546739578247, |
| "learning_rate": 0.00025232841288654235, |
| "loss": 3.3448, |
| "step": 53900 |
| }, |
| { |
| "epoch": 5.806694650737272, |
| "grad_norm": 0.7031681537628174, |
| "learning_rate": 0.00025200517185648095, |
| "loss": 3.3405, |
| "step": 53950 |
| }, |
| { |
| "epoch": 5.812076202776881, |
| "grad_norm": 0.7235373258590698, |
| "learning_rate": 0.00025168193082641954, |
| "loss": 3.3478, |
| "step": 54000 |
| }, |
| { |
| "epoch": 5.812076202776881, |
| "eval_accuracy": 0.38333303815965475, |
| "eval_loss": 3.3870346546173096, |
| "eval_runtime": 180.071, |
| "eval_samples_per_second": 100.022, |
| "eval_steps_per_second": 6.253, |
| "step": 54000 |
| }, |
| { |
| "epoch": 5.817457754816489, |
| "grad_norm": 0.6720332503318787, |
| "learning_rate": 0.00025136515461695935, |
| "loss": 3.3463, |
| "step": 54050 |
| }, |
| { |
| "epoch": 5.822839306856097, |
| "grad_norm": 0.6974610090255737, |
| "learning_rate": 0.00025104191358689794, |
| "loss": 3.3252, |
| "step": 54100 |
| }, |
| { |
| "epoch": 5.828220858895706, |
| "grad_norm": 0.679633378982544, |
| "learning_rate": 0.00025071867255683654, |
| "loss": 3.3429, |
| "step": 54150 |
| }, |
| { |
| "epoch": 5.833602410935313, |
| "grad_norm": 0.7177897095680237, |
| "learning_rate": 0.00025039543152677513, |
| "loss": 3.3452, |
| "step": 54200 |
| }, |
| { |
| "epoch": 5.838983962974922, |
| "grad_norm": 0.7383708357810974, |
| "learning_rate": 0.0002500721904967137, |
| "loss": 3.3431, |
| "step": 54250 |
| }, |
| { |
| "epoch": 5.84436551501453, |
| "grad_norm": 0.7053207159042358, |
| "learning_rate": 0.00024974894946665227, |
| "loss": 3.3192, |
| "step": 54300 |
| }, |
| { |
| "epoch": 5.849747067054138, |
| "grad_norm": 0.6786372065544128, |
| "learning_rate": 0.00024942570843659086, |
| "loss": 3.3387, |
| "step": 54350 |
| }, |
| { |
| "epoch": 5.855128619093747, |
| "grad_norm": 0.7125976085662842, |
| "learning_rate": 0.00024910246740652946, |
| "loss": 3.3389, |
| "step": 54400 |
| }, |
| { |
| "epoch": 5.860510171133355, |
| "grad_norm": 0.6816068887710571, |
| "learning_rate": 0.000248779226376468, |
| "loss": 3.3362, |
| "step": 54450 |
| }, |
| { |
| "epoch": 5.865891723172963, |
| "grad_norm": 0.691918671131134, |
| "learning_rate": 0.00024845598534640665, |
| "loss": 3.3445, |
| "step": 54500 |
| }, |
| { |
| "epoch": 5.871273275212571, |
| "grad_norm": 0.7392534017562866, |
| "learning_rate": 0.0002481327443163452, |
| "loss": 3.3357, |
| "step": 54550 |
| }, |
| { |
| "epoch": 5.87665482725218, |
| "grad_norm": 0.6493129730224609, |
| "learning_rate": 0.0002478095032862838, |
| "loss": 3.3564, |
| "step": 54600 |
| }, |
| { |
| "epoch": 5.882036379291788, |
| "grad_norm": 0.6863617300987244, |
| "learning_rate": 0.0002474862622562224, |
| "loss": 3.3392, |
| "step": 54650 |
| }, |
| { |
| "epoch": 5.887417931331396, |
| "grad_norm": 0.683229923248291, |
| "learning_rate": 0.00024716302122616097, |
| "loss": 3.3235, |
| "step": 54700 |
| }, |
| { |
| "epoch": 5.892799483371004, |
| "grad_norm": 0.6831369400024414, |
| "learning_rate": 0.0002468397801960995, |
| "loss": 3.3386, |
| "step": 54750 |
| }, |
| { |
| "epoch": 5.898181035410612, |
| "grad_norm": 0.7189830541610718, |
| "learning_rate": 0.0002465165391660381, |
| "loss": 3.3238, |
| "step": 54800 |
| }, |
| { |
| "epoch": 5.903562587450221, |
| "grad_norm": 0.7210323214530945, |
| "learning_rate": 0.0002461932981359767, |
| "loss": 3.3382, |
| "step": 54850 |
| }, |
| { |
| "epoch": 5.9089441394898286, |
| "grad_norm": 0.7270302772521973, |
| "learning_rate": 0.0002458700571059153, |
| "loss": 3.3289, |
| "step": 54900 |
| }, |
| { |
| "epoch": 5.914325691529437, |
| "grad_norm": 0.7537262439727783, |
| "learning_rate": 0.0002455468160758539, |
| "loss": 3.3261, |
| "step": 54950 |
| }, |
| { |
| "epoch": 5.919707243569046, |
| "grad_norm": 0.6871313452720642, |
| "learning_rate": 0.00024522357504579243, |
| "loss": 3.3193, |
| "step": 55000 |
| }, |
| { |
| "epoch": 5.919707243569046, |
| "eval_accuracy": 0.38383121165028417, |
| "eval_loss": 3.383720636367798, |
| "eval_runtime": 179.8106, |
| "eval_samples_per_second": 100.166, |
| "eval_steps_per_second": 6.262, |
| "step": 55000 |
| }, |
| { |
| "epoch": 5.925088795608653, |
| "grad_norm": 0.7151455879211426, |
| "learning_rate": 0.000244900334015731, |
| "loss": 3.3414, |
| "step": 55050 |
| }, |
| { |
| "epoch": 5.930470347648262, |
| "grad_norm": 0.6931750178337097, |
| "learning_rate": 0.0002445770929856696, |
| "loss": 3.3263, |
| "step": 55100 |
| }, |
| { |
| "epoch": 5.93585189968787, |
| "grad_norm": 0.6594939231872559, |
| "learning_rate": 0.0002442538519556082, |
| "loss": 3.326, |
| "step": 55150 |
| }, |
| { |
| "epoch": 5.941233451727478, |
| "grad_norm": 0.6955778002738953, |
| "learning_rate": 0.0002439306109255468, |
| "loss": 3.3455, |
| "step": 55200 |
| }, |
| { |
| "epoch": 5.946615003767087, |
| "grad_norm": 0.70442795753479, |
| "learning_rate": 0.00024360736989548538, |
| "loss": 3.3577, |
| "step": 55250 |
| }, |
| { |
| "epoch": 5.951996555806694, |
| "grad_norm": 0.7052885293960571, |
| "learning_rate": 0.00024328412886542394, |
| "loss": 3.3171, |
| "step": 55300 |
| }, |
| { |
| "epoch": 5.957378107846303, |
| "grad_norm": 0.7235458493232727, |
| "learning_rate": 0.00024296088783536257, |
| "loss": 3.3458, |
| "step": 55350 |
| }, |
| { |
| "epoch": 5.962759659885911, |
| "grad_norm": 0.7326828837394714, |
| "learning_rate": 0.00024263764680530113, |
| "loss": 3.3204, |
| "step": 55400 |
| }, |
| { |
| "epoch": 5.968141211925519, |
| "grad_norm": 0.724249005317688, |
| "learning_rate": 0.0002423144057752397, |
| "loss": 3.3413, |
| "step": 55450 |
| }, |
| { |
| "epoch": 5.973522763965128, |
| "grad_norm": 0.6884929537773132, |
| "learning_rate": 0.00024199116474517832, |
| "loss": 3.3452, |
| "step": 55500 |
| }, |
| { |
| "epoch": 5.978904316004736, |
| "grad_norm": 0.714123010635376, |
| "learning_rate": 0.0002416679237151169, |
| "loss": 3.3299, |
| "step": 55550 |
| }, |
| { |
| "epoch": 5.984285868044344, |
| "grad_norm": 0.7497240900993347, |
| "learning_rate": 0.00024134468268505546, |
| "loss": 3.3456, |
| "step": 55600 |
| }, |
| { |
| "epoch": 5.989667420083952, |
| "grad_norm": 0.7143298387527466, |
| "learning_rate": 0.00024102144165499405, |
| "loss": 3.3301, |
| "step": 55650 |
| }, |
| { |
| "epoch": 5.995048972123561, |
| "grad_norm": 0.6992262601852417, |
| "learning_rate": 0.00024069820062493265, |
| "loss": 3.3372, |
| "step": 55700 |
| }, |
| { |
| "epoch": 6.000430524163169, |
| "grad_norm": 0.6653108596801758, |
| "learning_rate": 0.00024037495959487121, |
| "loss": 3.318, |
| "step": 55750 |
| }, |
| { |
| "epoch": 6.005812076202777, |
| "grad_norm": 0.7049205303192139, |
| "learning_rate": 0.0002400517185648098, |
| "loss": 3.2475, |
| "step": 55800 |
| }, |
| { |
| "epoch": 6.011193628242385, |
| "grad_norm": 0.71421217918396, |
| "learning_rate": 0.00023972847753474838, |
| "loss": 3.2392, |
| "step": 55850 |
| }, |
| { |
| "epoch": 6.016575180281993, |
| "grad_norm": 0.6809349656105042, |
| "learning_rate": 0.00023940523650468697, |
| "loss": 3.2473, |
| "step": 55900 |
| }, |
| { |
| "epoch": 6.021956732321602, |
| "grad_norm": 0.7392157316207886, |
| "learning_rate": 0.00023908199547462557, |
| "loss": 3.25, |
| "step": 55950 |
| }, |
| { |
| "epoch": 6.0273382843612096, |
| "grad_norm": 0.7034648060798645, |
| "learning_rate": 0.00023875875444456413, |
| "loss": 3.2467, |
| "step": 56000 |
| }, |
| { |
| "epoch": 6.0273382843612096, |
| "eval_accuracy": 0.38408567671354565, |
| "eval_loss": 3.383943796157837, |
| "eval_runtime": 179.8915, |
| "eval_samples_per_second": 100.121, |
| "eval_steps_per_second": 6.259, |
| "step": 56000 |
| }, |
| { |
| "epoch": 6.032719836400818, |
| "grad_norm": 0.7213525176048279, |
| "learning_rate": 0.0002384355134145027, |
| "loss": 3.2419, |
| "step": 56050 |
| }, |
| { |
| "epoch": 6.038101388440427, |
| "grad_norm": 0.7486132979393005, |
| "learning_rate": 0.00023811227238444132, |
| "loss": 3.245, |
| "step": 56100 |
| }, |
| { |
| "epoch": 6.043482940480034, |
| "grad_norm": 0.6930773258209229, |
| "learning_rate": 0.0002377890313543799, |
| "loss": 3.2324, |
| "step": 56150 |
| }, |
| { |
| "epoch": 6.048864492519643, |
| "grad_norm": 0.7631097435951233, |
| "learning_rate": 0.00023746579032431849, |
| "loss": 3.253, |
| "step": 56200 |
| }, |
| { |
| "epoch": 6.0542460445592505, |
| "grad_norm": 0.7404524087905884, |
| "learning_rate": 0.00023714254929425708, |
| "loss": 3.2553, |
| "step": 56250 |
| }, |
| { |
| "epoch": 6.059627596598859, |
| "grad_norm": 0.7180728912353516, |
| "learning_rate": 0.00023681930826419565, |
| "loss": 3.2438, |
| "step": 56300 |
| }, |
| { |
| "epoch": 6.065009148638468, |
| "grad_norm": 0.7711812853813171, |
| "learning_rate": 0.00023649606723413424, |
| "loss": 3.254, |
| "step": 56350 |
| }, |
| { |
| "epoch": 6.070390700678075, |
| "grad_norm": 0.7722243666648865, |
| "learning_rate": 0.0002361728262040728, |
| "loss": 3.2458, |
| "step": 56400 |
| }, |
| { |
| "epoch": 6.075772252717684, |
| "grad_norm": 0.6931943893432617, |
| "learning_rate": 0.00023584958517401138, |
| "loss": 3.2818, |
| "step": 56450 |
| }, |
| { |
| "epoch": 6.081153804757292, |
| "grad_norm": 0.6928170323371887, |
| "learning_rate": 0.00023552634414395, |
| "loss": 3.258, |
| "step": 56500 |
| }, |
| { |
| "epoch": 6.0865353567969, |
| "grad_norm": 0.7071146368980408, |
| "learning_rate": 0.00023520310311388857, |
| "loss": 3.2763, |
| "step": 56550 |
| }, |
| { |
| "epoch": 6.091916908836509, |
| "grad_norm": 0.7333793044090271, |
| "learning_rate": 0.00023487986208382713, |
| "loss": 3.2601, |
| "step": 56600 |
| }, |
| { |
| "epoch": 6.097298460876116, |
| "grad_norm": 0.7094366550445557, |
| "learning_rate": 0.00023455662105376576, |
| "loss": 3.2547, |
| "step": 56650 |
| }, |
| { |
| "epoch": 6.102680012915725, |
| "grad_norm": 0.7214093208312988, |
| "learning_rate": 0.00023423338002370432, |
| "loss": 3.2599, |
| "step": 56700 |
| }, |
| { |
| "epoch": 6.108061564955333, |
| "grad_norm": 0.715020477771759, |
| "learning_rate": 0.0002339101389936429, |
| "loss": 3.2696, |
| "step": 56750 |
| }, |
| { |
| "epoch": 6.113443116994941, |
| "grad_norm": 0.7385733127593994, |
| "learning_rate": 0.00023358689796358149, |
| "loss": 3.2711, |
| "step": 56800 |
| }, |
| { |
| "epoch": 6.11882466903455, |
| "grad_norm": 0.7252424359321594, |
| "learning_rate": 0.00023326365693352008, |
| "loss": 3.2499, |
| "step": 56850 |
| }, |
| { |
| "epoch": 6.124206221074158, |
| "grad_norm": 0.7170380353927612, |
| "learning_rate": 0.00023294041590345865, |
| "loss": 3.2568, |
| "step": 56900 |
| }, |
| { |
| "epoch": 6.129587773113766, |
| "grad_norm": 0.7241175174713135, |
| "learning_rate": 0.00023261717487339724, |
| "loss": 3.2735, |
| "step": 56950 |
| }, |
| { |
| "epoch": 6.134969325153374, |
| "grad_norm": 0.6951444149017334, |
| "learning_rate": 0.0002322939338433358, |
| "loss": 3.2572, |
| "step": 57000 |
| }, |
| { |
| "epoch": 6.134969325153374, |
| "eval_accuracy": 0.38414315409119953, |
| "eval_loss": 3.3844387531280518, |
| "eval_runtime": 179.7181, |
| "eval_samples_per_second": 100.218, |
| "eval_steps_per_second": 6.265, |
| "step": 57000 |
| }, |
| { |
| "epoch": 6.140350877192983, |
| "grad_norm": 0.7219548225402832, |
| "learning_rate": 0.0002319706928132744, |
| "loss": 3.2531, |
| "step": 57050 |
| }, |
| { |
| "epoch": 6.1457324292325906, |
| "grad_norm": 0.6793873310089111, |
| "learning_rate": 0.000231647451783213, |
| "loss": 3.2661, |
| "step": 57100 |
| }, |
| { |
| "epoch": 6.151113981272199, |
| "grad_norm": 0.6982064247131348, |
| "learning_rate": 0.00023132421075315157, |
| "loss": 3.2649, |
| "step": 57150 |
| }, |
| { |
| "epoch": 6.156495533311807, |
| "grad_norm": 0.7467171549797058, |
| "learning_rate": 0.0002310009697230902, |
| "loss": 3.2689, |
| "step": 57200 |
| }, |
| { |
| "epoch": 6.161877085351415, |
| "grad_norm": 0.6914950609207153, |
| "learning_rate": 0.00023067772869302876, |
| "loss": 3.2709, |
| "step": 57250 |
| }, |
| { |
| "epoch": 6.167258637391024, |
| "grad_norm": 0.7428169250488281, |
| "learning_rate": 0.00023035448766296732, |
| "loss": 3.2955, |
| "step": 57300 |
| }, |
| { |
| "epoch": 6.1726401894306315, |
| "grad_norm": 0.7214995622634888, |
| "learning_rate": 0.00023003124663290592, |
| "loss": 3.269, |
| "step": 57350 |
| }, |
| { |
| "epoch": 6.17802174147024, |
| "grad_norm": 0.7301390767097473, |
| "learning_rate": 0.0002297080056028445, |
| "loss": 3.271, |
| "step": 57400 |
| }, |
| { |
| "epoch": 6.183403293509849, |
| "grad_norm": 0.7483002543449402, |
| "learning_rate": 0.00022938476457278308, |
| "loss": 3.2873, |
| "step": 57450 |
| }, |
| { |
| "epoch": 6.188784845549456, |
| "grad_norm": 0.7099791765213013, |
| "learning_rate": 0.00022906152354272168, |
| "loss": 3.2824, |
| "step": 57500 |
| }, |
| { |
| "epoch": 6.194166397589065, |
| "grad_norm": 0.7995650768280029, |
| "learning_rate": 0.00022873828251266024, |
| "loss": 3.2704, |
| "step": 57550 |
| }, |
| { |
| "epoch": 6.1995479496286725, |
| "grad_norm": 0.7128803730010986, |
| "learning_rate": 0.00022841504148259884, |
| "loss": 3.2638, |
| "step": 57600 |
| }, |
| { |
| "epoch": 6.204929501668281, |
| "grad_norm": 0.7057095170021057, |
| "learning_rate": 0.00022809180045253743, |
| "loss": 3.2757, |
| "step": 57650 |
| }, |
| { |
| "epoch": 6.21031105370789, |
| "grad_norm": 0.7325379252433777, |
| "learning_rate": 0.000227768559422476, |
| "loss": 3.2648, |
| "step": 57700 |
| }, |
| { |
| "epoch": 6.215692605747497, |
| "grad_norm": 0.710580050945282, |
| "learning_rate": 0.00022744531839241457, |
| "loss": 3.2753, |
| "step": 57750 |
| }, |
| { |
| "epoch": 6.221074157787106, |
| "grad_norm": 0.7522643208503723, |
| "learning_rate": 0.0002271220773623532, |
| "loss": 3.2799, |
| "step": 57800 |
| }, |
| { |
| "epoch": 6.226455709826714, |
| "grad_norm": 0.7007228136062622, |
| "learning_rate": 0.00022679883633229176, |
| "loss": 3.2747, |
| "step": 57850 |
| }, |
| { |
| "epoch": 6.231837261866322, |
| "grad_norm": 0.7241410613059998, |
| "learning_rate": 0.00022647559530223032, |
| "loss": 3.2759, |
| "step": 57900 |
| }, |
| { |
| "epoch": 6.237218813905931, |
| "grad_norm": 0.7449703216552734, |
| "learning_rate": 0.00022615235427216895, |
| "loss": 3.2641, |
| "step": 57950 |
| }, |
| { |
| "epoch": 6.242600365945538, |
| "grad_norm": 0.7371934652328491, |
| "learning_rate": 0.0002258291132421075, |
| "loss": 3.2651, |
| "step": 58000 |
| }, |
| { |
| "epoch": 6.242600365945538, |
| "eval_accuracy": 0.3842598472927123, |
| "eval_loss": 3.381721258163452, |
| "eval_runtime": 179.7671, |
| "eval_samples_per_second": 100.191, |
| "eval_steps_per_second": 6.264, |
| "step": 58000 |
| }, |
| { |
| "epoch": 6.247981917985147, |
| "grad_norm": 0.7152185440063477, |
| "learning_rate": 0.00022551233703264732, |
| "loss": 3.2728, |
| "step": 58050 |
| }, |
| { |
| "epoch": 6.253363470024755, |
| "grad_norm": 0.7568090558052063, |
| "learning_rate": 0.0002251890960025859, |
| "loss": 3.2583, |
| "step": 58100 |
| }, |
| { |
| "epoch": 6.258745022064363, |
| "grad_norm": 0.7317862510681152, |
| "learning_rate": 0.0002248658549725245, |
| "loss": 3.2633, |
| "step": 58150 |
| }, |
| { |
| "epoch": 6.264126574103972, |
| "grad_norm": 0.7428773045539856, |
| "learning_rate": 0.00022454261394246308, |
| "loss": 3.2919, |
| "step": 58200 |
| }, |
| { |
| "epoch": 6.26950812614358, |
| "grad_norm": 0.7505496144294739, |
| "learning_rate": 0.00022421937291240164, |
| "loss": 3.28, |
| "step": 58250 |
| }, |
| { |
| "epoch": 6.274889678183188, |
| "grad_norm": 0.7657462954521179, |
| "learning_rate": 0.00022389613188234027, |
| "loss": 3.2791, |
| "step": 58300 |
| }, |
| { |
| "epoch": 6.280271230222796, |
| "grad_norm": 0.7743219137191772, |
| "learning_rate": 0.00022357289085227883, |
| "loss": 3.2836, |
| "step": 58350 |
| }, |
| { |
| "epoch": 6.285652782262405, |
| "grad_norm": 0.7271228432655334, |
| "learning_rate": 0.0002232496498222174, |
| "loss": 3.2818, |
| "step": 58400 |
| }, |
| { |
| "epoch": 6.2910343343020125, |
| "grad_norm": 0.7528961896896362, |
| "learning_rate": 0.000222926408792156, |
| "loss": 3.2631, |
| "step": 58450 |
| }, |
| { |
| "epoch": 6.296415886341621, |
| "grad_norm": 0.8071222305297852, |
| "learning_rate": 0.0002226031677620946, |
| "loss": 3.2787, |
| "step": 58500 |
| }, |
| { |
| "epoch": 6.301797438381229, |
| "grad_norm": 0.7826973795890808, |
| "learning_rate": 0.00022227992673203316, |
| "loss": 3.2795, |
| "step": 58550 |
| }, |
| { |
| "epoch": 6.307178990420837, |
| "grad_norm": 0.7677108645439148, |
| "learning_rate": 0.00022195668570197175, |
| "loss": 3.2784, |
| "step": 58600 |
| }, |
| { |
| "epoch": 6.312560542460446, |
| "grad_norm": 0.7295692563056946, |
| "learning_rate": 0.00022163344467191032, |
| "loss": 3.2896, |
| "step": 58650 |
| }, |
| { |
| "epoch": 6.3179420945000535, |
| "grad_norm": 0.7121778130531311, |
| "learning_rate": 0.00022131020364184891, |
| "loss": 3.2813, |
| "step": 58700 |
| }, |
| { |
| "epoch": 6.323323646539662, |
| "grad_norm": 0.743169903755188, |
| "learning_rate": 0.0002209869626117875, |
| "loss": 3.2768, |
| "step": 58750 |
| }, |
| { |
| "epoch": 6.328705198579271, |
| "grad_norm": 0.7486587762832642, |
| "learning_rate": 0.00022066372158172608, |
| "loss": 3.2684, |
| "step": 58800 |
| }, |
| { |
| "epoch": 6.334086750618878, |
| "grad_norm": 0.738730788230896, |
| "learning_rate": 0.00022034048055166464, |
| "loss": 3.2697, |
| "step": 58850 |
| }, |
| { |
| "epoch": 6.339468302658487, |
| "grad_norm": 0.7581406831741333, |
| "learning_rate": 0.00022001723952160327, |
| "loss": 3.2812, |
| "step": 58900 |
| }, |
| { |
| "epoch": 6.344849854698095, |
| "grad_norm": 0.7450671792030334, |
| "learning_rate": 0.00021969399849154183, |
| "loss": 3.2936, |
| "step": 58950 |
| }, |
| { |
| "epoch": 6.350231406737703, |
| "grad_norm": 0.8056238293647766, |
| "learning_rate": 0.00021937075746148043, |
| "loss": 3.2875, |
| "step": 59000 |
| }, |
| { |
| "epoch": 6.350231406737703, |
| "eval_accuracy": 0.38507724296774065, |
| "eval_loss": 3.3782382011413574, |
| "eval_runtime": 179.996, |
| "eval_samples_per_second": 100.063, |
| "eval_steps_per_second": 6.256, |
| "step": 59000 |
| }, |
| { |
| "epoch": 6.355612958777312, |
| "grad_norm": 0.7694650888442993, |
| "learning_rate": 0.00021904751643141902, |
| "loss": 3.2736, |
| "step": 59050 |
| }, |
| { |
| "epoch": 6.360994510816919, |
| "grad_norm": 0.8028887510299683, |
| "learning_rate": 0.0002187242754013576, |
| "loss": 3.2917, |
| "step": 59100 |
| }, |
| { |
| "epoch": 6.366376062856528, |
| "grad_norm": 0.7098097205162048, |
| "learning_rate": 0.00021840103437129619, |
| "loss": 3.2609, |
| "step": 59150 |
| }, |
| { |
| "epoch": 6.371757614896136, |
| "grad_norm": 0.7649843096733093, |
| "learning_rate": 0.00021807779334123475, |
| "loss": 3.2861, |
| "step": 59200 |
| }, |
| { |
| "epoch": 6.377139166935744, |
| "grad_norm": 0.7714524865150452, |
| "learning_rate": 0.00021775455231117335, |
| "loss": 3.2928, |
| "step": 59250 |
| }, |
| { |
| "epoch": 6.382520718975353, |
| "grad_norm": 0.7555143237113953, |
| "learning_rate": 0.00021743131128111194, |
| "loss": 3.2814, |
| "step": 59300 |
| }, |
| { |
| "epoch": 6.387902271014961, |
| "grad_norm": 0.7391084432601929, |
| "learning_rate": 0.0002171080702510505, |
| "loss": 3.277, |
| "step": 59350 |
| }, |
| { |
| "epoch": 6.393283823054569, |
| "grad_norm": 0.7586864233016968, |
| "learning_rate": 0.00021678482922098908, |
| "loss": 3.2753, |
| "step": 59400 |
| }, |
| { |
| "epoch": 6.398665375094177, |
| "grad_norm": 0.7793105840682983, |
| "learning_rate": 0.0002164615881909277, |
| "loss": 3.283, |
| "step": 59450 |
| }, |
| { |
| "epoch": 6.404046927133785, |
| "grad_norm": 0.7611179351806641, |
| "learning_rate": 0.00021613834716086627, |
| "loss": 3.2817, |
| "step": 59500 |
| }, |
| { |
| "epoch": 6.4094284791733935, |
| "grad_norm": 0.7332265973091125, |
| "learning_rate": 0.00021581510613080483, |
| "loss": 3.2797, |
| "step": 59550 |
| }, |
| { |
| "epoch": 6.414810031213002, |
| "grad_norm": 0.7599790096282959, |
| "learning_rate": 0.00021549186510074346, |
| "loss": 3.2779, |
| "step": 59600 |
| }, |
| { |
| "epoch": 6.42019158325261, |
| "grad_norm": 0.7514798045158386, |
| "learning_rate": 0.00021516862407068202, |
| "loss": 3.2788, |
| "step": 59650 |
| }, |
| { |
| "epoch": 6.425573135292218, |
| "grad_norm": 0.7226225137710571, |
| "learning_rate": 0.0002148453830406206, |
| "loss": 3.3054, |
| "step": 59700 |
| }, |
| { |
| "epoch": 6.430954687331827, |
| "grad_norm": 0.7565598487854004, |
| "learning_rate": 0.00021452214201055919, |
| "loss": 3.2657, |
| "step": 59750 |
| }, |
| { |
| "epoch": 6.4363362393714345, |
| "grad_norm": 0.7199046015739441, |
| "learning_rate": 0.00021419890098049778, |
| "loss": 3.2656, |
| "step": 59800 |
| }, |
| { |
| "epoch": 6.441717791411043, |
| "grad_norm": 0.7737821340560913, |
| "learning_rate": 0.00021387565995043638, |
| "loss": 3.285, |
| "step": 59850 |
| }, |
| { |
| "epoch": 6.447099343450651, |
| "grad_norm": 0.7489638924598694, |
| "learning_rate": 0.00021355241892037494, |
| "loss": 3.2728, |
| "step": 59900 |
| }, |
| { |
| "epoch": 6.452480895490259, |
| "grad_norm": 0.8023089170455933, |
| "learning_rate": 0.0002132291778903135, |
| "loss": 3.2688, |
| "step": 59950 |
| }, |
| { |
| "epoch": 6.457862447529868, |
| "grad_norm": 0.7359650731086731, |
| "learning_rate": 0.00021290593686025213, |
| "loss": 3.2697, |
| "step": 60000 |
| }, |
| { |
| "epoch": 6.457862447529868, |
| "eval_accuracy": 0.38522142534987047, |
| "eval_loss": 3.3750526905059814, |
| "eval_runtime": 179.9044, |
| "eval_samples_per_second": 100.114, |
| "eval_steps_per_second": 6.259, |
| "step": 60000 |
| }, |
| { |
| "epoch": 6.4632439995694755, |
| "grad_norm": 0.739447832107544, |
| "learning_rate": 0.0002125891606507919, |
| "loss": 3.2756, |
| "step": 60050 |
| }, |
| { |
| "epoch": 6.468625551609084, |
| "grad_norm": 0.7175700068473816, |
| "learning_rate": 0.0002122659196207305, |
| "loss": 3.2523, |
| "step": 60100 |
| }, |
| { |
| "epoch": 6.474007103648693, |
| "grad_norm": 0.7286744713783264, |
| "learning_rate": 0.0002119426785906691, |
| "loss": 3.2838, |
| "step": 60150 |
| }, |
| { |
| "epoch": 6.4793886556883, |
| "grad_norm": 0.765244722366333, |
| "learning_rate": 0.0002116259023812089, |
| "loss": 3.2948, |
| "step": 60200 |
| }, |
| { |
| "epoch": 6.484770207727909, |
| "grad_norm": 0.7369528412818909, |
| "learning_rate": 0.00021130266135114747, |
| "loss": 3.2776, |
| "step": 60250 |
| }, |
| { |
| "epoch": 6.490151759767517, |
| "grad_norm": 0.7986508011817932, |
| "learning_rate": 0.00021097942032108607, |
| "loss": 3.3132, |
| "step": 60300 |
| }, |
| { |
| "epoch": 6.495533311807125, |
| "grad_norm": 0.7739220261573792, |
| "learning_rate": 0.00021065617929102466, |
| "loss": 3.2838, |
| "step": 60350 |
| }, |
| { |
| "epoch": 6.500914863846734, |
| "grad_norm": 0.7373184561729431, |
| "learning_rate": 0.00021033293826096323, |
| "loss": 3.297, |
| "step": 60400 |
| }, |
| { |
| "epoch": 6.506296415886341, |
| "grad_norm": 0.7406927347183228, |
| "learning_rate": 0.0002100096972309018, |
| "loss": 3.2618, |
| "step": 60450 |
| }, |
| { |
| "epoch": 6.51167796792595, |
| "grad_norm": 0.7355625033378601, |
| "learning_rate": 0.00020968645620084042, |
| "loss": 3.2744, |
| "step": 60500 |
| }, |
| { |
| "epoch": 6.517059519965558, |
| "grad_norm": 0.7376171946525574, |
| "learning_rate": 0.000209363215170779, |
| "loss": 3.2761, |
| "step": 60550 |
| }, |
| { |
| "epoch": 6.522441072005166, |
| "grad_norm": 0.76179039478302, |
| "learning_rate": 0.00020903997414071758, |
| "loss": 3.2946, |
| "step": 60600 |
| }, |
| { |
| "epoch": 6.5278226240447745, |
| "grad_norm": 0.7716966271400452, |
| "learning_rate": 0.00020871673311065618, |
| "loss": 3.2846, |
| "step": 60650 |
| }, |
| { |
| "epoch": 6.533204176084383, |
| "grad_norm": 0.7968785166740417, |
| "learning_rate": 0.00020839349208059475, |
| "loss": 3.2766, |
| "step": 60700 |
| }, |
| { |
| "epoch": 6.538585728123991, |
| "grad_norm": 0.7669464349746704, |
| "learning_rate": 0.00020807025105053334, |
| "loss": 3.2771, |
| "step": 60750 |
| }, |
| { |
| "epoch": 6.543967280163599, |
| "grad_norm": 0.7410077452659607, |
| "learning_rate": 0.0002077470100204719, |
| "loss": 3.2788, |
| "step": 60800 |
| }, |
| { |
| "epoch": 6.549348832203208, |
| "grad_norm": 0.7026287913322449, |
| "learning_rate": 0.00020742376899041048, |
| "loss": 3.2826, |
| "step": 60850 |
| }, |
| { |
| "epoch": 6.5547303842428155, |
| "grad_norm": 0.7268006205558777, |
| "learning_rate": 0.0002071005279603491, |
| "loss": 3.2779, |
| "step": 60900 |
| }, |
| { |
| "epoch": 6.560111936282424, |
| "grad_norm": 0.7330572605133057, |
| "learning_rate": 0.00020677728693028766, |
| "loss": 3.2736, |
| "step": 60950 |
| }, |
| { |
| "epoch": 6.565493488322032, |
| "grad_norm": 0.7642152905464172, |
| "learning_rate": 0.00020645404590022623, |
| "loss": 3.2909, |
| "step": 61000 |
| }, |
| { |
| "epoch": 6.565493488322032, |
| "eval_accuracy": 0.3857069951055134, |
| "eval_loss": 3.372335433959961, |
| "eval_runtime": 179.717, |
| "eval_samples_per_second": 100.219, |
| "eval_steps_per_second": 6.265, |
| "step": 61000 |
| }, |
| { |
| "epoch": 6.57087504036164, |
| "grad_norm": 0.7624441981315613, |
| "learning_rate": 0.00020613080487016485, |
| "loss": 3.2991, |
| "step": 61050 |
| }, |
| { |
| "epoch": 6.576256592401249, |
| "grad_norm": 0.7940060496330261, |
| "learning_rate": 0.00020580756384010342, |
| "loss": 3.302, |
| "step": 61100 |
| }, |
| { |
| "epoch": 6.5816381444408565, |
| "grad_norm": 0.6994228959083557, |
| "learning_rate": 0.000205484322810042, |
| "loss": 3.2817, |
| "step": 61150 |
| }, |
| { |
| "epoch": 6.587019696480465, |
| "grad_norm": 0.7420439124107361, |
| "learning_rate": 0.00020516108177998058, |
| "loss": 3.2781, |
| "step": 61200 |
| }, |
| { |
| "epoch": 6.592401248520073, |
| "grad_norm": 0.7529013156890869, |
| "learning_rate": 0.00020483784074991918, |
| "loss": 3.2846, |
| "step": 61250 |
| }, |
| { |
| "epoch": 6.597782800559681, |
| "grad_norm": 0.7393249273300171, |
| "learning_rate": 0.00020451459971985775, |
| "loss": 3.2994, |
| "step": 61300 |
| }, |
| { |
| "epoch": 6.60316435259929, |
| "grad_norm": 0.8019313216209412, |
| "learning_rate": 0.00020419135868979634, |
| "loss": 3.2977, |
| "step": 61350 |
| }, |
| { |
| "epoch": 6.608545904638898, |
| "grad_norm": 0.7454643845558167, |
| "learning_rate": 0.0002038681176597349, |
| "loss": 3.2705, |
| "step": 61400 |
| }, |
| { |
| "epoch": 6.613927456678506, |
| "grad_norm": 0.7411048412322998, |
| "learning_rate": 0.0002035448766296735, |
| "loss": 3.2781, |
| "step": 61450 |
| }, |
| { |
| "epoch": 6.619309008718115, |
| "grad_norm": 0.747560441493988, |
| "learning_rate": 0.0002032216355996121, |
| "loss": 3.2803, |
| "step": 61500 |
| }, |
| { |
| "epoch": 6.624690560757722, |
| "grad_norm": 0.7290646433830261, |
| "learning_rate": 0.00020289839456955067, |
| "loss": 3.2812, |
| "step": 61550 |
| }, |
| { |
| "epoch": 6.630072112797331, |
| "grad_norm": 0.757982611656189, |
| "learning_rate": 0.0002025751535394893, |
| "loss": 3.2954, |
| "step": 61600 |
| }, |
| { |
| "epoch": 6.635453664836939, |
| "grad_norm": 0.7968859672546387, |
| "learning_rate": 0.00020225191250942785, |
| "loss": 3.2833, |
| "step": 61650 |
| }, |
| { |
| "epoch": 6.640835216876547, |
| "grad_norm": 0.7631063461303711, |
| "learning_rate": 0.00020192867147936642, |
| "loss": 3.2753, |
| "step": 61700 |
| }, |
| { |
| "epoch": 6.6462167689161555, |
| "grad_norm": 0.781497061252594, |
| "learning_rate": 0.00020160543044930502, |
| "loss": 3.2842, |
| "step": 61750 |
| }, |
| { |
| "epoch": 6.651598320955763, |
| "grad_norm": 0.7126927375793457, |
| "learning_rate": 0.0002012821894192436, |
| "loss": 3.2704, |
| "step": 61800 |
| }, |
| { |
| "epoch": 6.656979872995372, |
| "grad_norm": 0.7944461703300476, |
| "learning_rate": 0.00020095894838918218, |
| "loss": 3.2639, |
| "step": 61850 |
| }, |
| { |
| "epoch": 6.66236142503498, |
| "grad_norm": 0.7860047817230225, |
| "learning_rate": 0.00020063570735912077, |
| "loss": 3.296, |
| "step": 61900 |
| }, |
| { |
| "epoch": 6.667742977074588, |
| "grad_norm": 0.8037759065628052, |
| "learning_rate": 0.00020031246632905934, |
| "loss": 3.2693, |
| "step": 61950 |
| }, |
| { |
| "epoch": 6.6731245291141965, |
| "grad_norm": 0.7314351797103882, |
| "learning_rate": 0.00019998922529899794, |
| "loss": 3.2797, |
| "step": 62000 |
| }, |
| { |
| "epoch": 6.6731245291141965, |
| "eval_accuracy": 0.38599720696886586, |
| "eval_loss": 3.369133234024048, |
| "eval_runtime": 180.0106, |
| "eval_samples_per_second": 100.055, |
| "eval_steps_per_second": 6.255, |
| "step": 62000 |
| }, |
| { |
| "epoch": 6.678506081153805, |
| "grad_norm": 0.761888861656189, |
| "learning_rate": 0.00019966598426893653, |
| "loss": 3.2843, |
| "step": 62050 |
| }, |
| { |
| "epoch": 6.683887633193413, |
| "grad_norm": 0.715152382850647, |
| "learning_rate": 0.0001993427432388751, |
| "loss": 3.2712, |
| "step": 62100 |
| }, |
| { |
| "epoch": 6.689269185233021, |
| "grad_norm": 0.7696418166160583, |
| "learning_rate": 0.00019901950220881367, |
| "loss": 3.3034, |
| "step": 62150 |
| }, |
| { |
| "epoch": 6.69465073727263, |
| "grad_norm": 0.7413962483406067, |
| "learning_rate": 0.0001986962611787523, |
| "loss": 3.2882, |
| "step": 62200 |
| }, |
| { |
| "epoch": 6.7000322893122375, |
| "grad_norm": 0.7677690386772156, |
| "learning_rate": 0.00019837302014869085, |
| "loss": 3.2869, |
| "step": 62250 |
| }, |
| { |
| "epoch": 6.705413841351846, |
| "grad_norm": 0.7569989562034607, |
| "learning_rate": 0.00019804977911862942, |
| "loss": 3.2869, |
| "step": 62300 |
| }, |
| { |
| "epoch": 6.710795393391454, |
| "grad_norm": 0.7461286783218384, |
| "learning_rate": 0.00019772653808856804, |
| "loss": 3.2899, |
| "step": 62350 |
| }, |
| { |
| "epoch": 6.716176945431062, |
| "grad_norm": 0.7389951944351196, |
| "learning_rate": 0.0001974032970585066, |
| "loss": 3.2635, |
| "step": 62400 |
| }, |
| { |
| "epoch": 6.721558497470671, |
| "grad_norm": 0.7195466756820679, |
| "learning_rate": 0.00019708005602844518, |
| "loss": 3.2589, |
| "step": 62450 |
| }, |
| { |
| "epoch": 6.7269400495102785, |
| "grad_norm": 0.7346585988998413, |
| "learning_rate": 0.00019675681499838377, |
| "loss": 3.2894, |
| "step": 62500 |
| }, |
| { |
| "epoch": 6.732321601549887, |
| "grad_norm": 0.7372649312019348, |
| "learning_rate": 0.00019643357396832237, |
| "loss": 3.2776, |
| "step": 62550 |
| }, |
| { |
| "epoch": 6.737703153589496, |
| "grad_norm": 0.7496547102928162, |
| "learning_rate": 0.00019611033293826096, |
| "loss": 3.2765, |
| "step": 62600 |
| }, |
| { |
| "epoch": 6.743084705629103, |
| "grad_norm": 0.7103299498558044, |
| "learning_rate": 0.00019578709190819953, |
| "loss": 3.3054, |
| "step": 62650 |
| }, |
| { |
| "epoch": 6.748466257668712, |
| "grad_norm": 0.7128199934959412, |
| "learning_rate": 0.0001954638508781381, |
| "loss": 3.2887, |
| "step": 62700 |
| }, |
| { |
| "epoch": 6.75384780970832, |
| "grad_norm": 0.7775223255157471, |
| "learning_rate": 0.00019514060984807672, |
| "loss": 3.2817, |
| "step": 62750 |
| }, |
| { |
| "epoch": 6.759229361747928, |
| "grad_norm": 0.7413020133972168, |
| "learning_rate": 0.0001948173688180153, |
| "loss": 3.2883, |
| "step": 62800 |
| }, |
| { |
| "epoch": 6.7646109137875365, |
| "grad_norm": 0.7511486411094666, |
| "learning_rate": 0.00019449412778795386, |
| "loss": 3.2796, |
| "step": 62850 |
| }, |
| { |
| "epoch": 6.769992465827144, |
| "grad_norm": 0.7225663661956787, |
| "learning_rate": 0.00019417088675789245, |
| "loss": 3.2843, |
| "step": 62900 |
| }, |
| { |
| "epoch": 6.775374017866753, |
| "grad_norm": 0.7489200830459595, |
| "learning_rate": 0.00019384764572783104, |
| "loss": 3.2977, |
| "step": 62950 |
| }, |
| { |
| "epoch": 6.780755569906361, |
| "grad_norm": 0.7403072714805603, |
| "learning_rate": 0.0001935244046977696, |
| "loss": 3.2781, |
| "step": 63000 |
| }, |
| { |
| "epoch": 6.780755569906361, |
| "eval_accuracy": 0.3861868062581021, |
| "eval_loss": 3.361095666885376, |
| "eval_runtime": 179.7674, |
| "eval_samples_per_second": 100.191, |
| "eval_steps_per_second": 6.264, |
| "step": 63000 |
| }, |
| { |
| "epoch": 6.786137121945969, |
| "grad_norm": 0.77875155210495, |
| "learning_rate": 0.0001932011636677082, |
| "loss": 3.3193, |
| "step": 63050 |
| }, |
| { |
| "epoch": 6.7915186739855775, |
| "grad_norm": 0.8084599375724792, |
| "learning_rate": 0.00019287792263764677, |
| "loss": 3.294, |
| "step": 63100 |
| }, |
| { |
| "epoch": 6.796900226025185, |
| "grad_norm": 0.7540778517723083, |
| "learning_rate": 0.00019255468160758537, |
| "loss": 3.284, |
| "step": 63150 |
| }, |
| { |
| "epoch": 6.802281778064794, |
| "grad_norm": 0.7482621669769287, |
| "learning_rate": 0.00019223144057752396, |
| "loss": 3.2912, |
| "step": 63200 |
| }, |
| { |
| "epoch": 6.807663330104402, |
| "grad_norm": 0.7529076933860779, |
| "learning_rate": 0.00019190819954746253, |
| "loss": 3.2768, |
| "step": 63250 |
| }, |
| { |
| "epoch": 6.813044882144011, |
| "grad_norm": 0.7714017033576965, |
| "learning_rate": 0.0001915849585174011, |
| "loss": 3.2763, |
| "step": 63300 |
| }, |
| { |
| "epoch": 6.8184264341836185, |
| "grad_norm": 0.7732032537460327, |
| "learning_rate": 0.00019126171748733972, |
| "loss": 3.2988, |
| "step": 63350 |
| }, |
| { |
| "epoch": 6.823807986223227, |
| "grad_norm": 0.7607737183570862, |
| "learning_rate": 0.0001909384764572783, |
| "loss": 3.2844, |
| "step": 63400 |
| }, |
| { |
| "epoch": 6.829189538262835, |
| "grad_norm": 0.7674480080604553, |
| "learning_rate": 0.00019061523542721688, |
| "loss": 3.2795, |
| "step": 63450 |
| }, |
| { |
| "epoch": 6.834571090302443, |
| "grad_norm": 0.7431223392486572, |
| "learning_rate": 0.00019029199439715548, |
| "loss": 3.2974, |
| "step": 63500 |
| }, |
| { |
| "epoch": 6.839952642342052, |
| "grad_norm": 0.7607661485671997, |
| "learning_rate": 0.00018996875336709404, |
| "loss": 3.2843, |
| "step": 63550 |
| }, |
| { |
| "epoch": 6.8453341943816595, |
| "grad_norm": 0.7642389535903931, |
| "learning_rate": 0.00018964551233703264, |
| "loss": 3.3047, |
| "step": 63600 |
| }, |
| { |
| "epoch": 6.850715746421268, |
| "grad_norm": 0.760985791683197, |
| "learning_rate": 0.0001893222713069712, |
| "loss": 3.285, |
| "step": 63650 |
| }, |
| { |
| "epoch": 6.856097298460876, |
| "grad_norm": 0.7435808777809143, |
| "learning_rate": 0.0001889990302769098, |
| "loss": 3.2766, |
| "step": 63700 |
| }, |
| { |
| "epoch": 6.861478850500484, |
| "grad_norm": 0.7285375595092773, |
| "learning_rate": 0.0001886757892468484, |
| "loss": 3.3006, |
| "step": 63750 |
| }, |
| { |
| "epoch": 6.866860402540093, |
| "grad_norm": 0.7190316915512085, |
| "learning_rate": 0.00018835254821678696, |
| "loss": 3.2846, |
| "step": 63800 |
| }, |
| { |
| "epoch": 6.8722419545797, |
| "grad_norm": 0.7786919474601746, |
| "learning_rate": 0.00018802930718672553, |
| "loss": 3.2925, |
| "step": 63850 |
| }, |
| { |
| "epoch": 6.877623506619309, |
| "grad_norm": 0.7862507104873657, |
| "learning_rate": 0.00018770606615666415, |
| "loss": 3.2824, |
| "step": 63900 |
| }, |
| { |
| "epoch": 6.8830050586589175, |
| "grad_norm": 0.8249383568763733, |
| "learning_rate": 0.00018738282512660272, |
| "loss": 3.284, |
| "step": 63950 |
| }, |
| { |
| "epoch": 6.888386610698525, |
| "grad_norm": 0.7981451749801636, |
| "learning_rate": 0.0001870595840965413, |
| "loss": 3.2945, |
| "step": 64000 |
| }, |
| { |
| "epoch": 6.888386610698525, |
| "eval_accuracy": 0.38657045960497505, |
| "eval_loss": 3.35958194732666, |
| "eval_runtime": 179.9583, |
| "eval_samples_per_second": 100.084, |
| "eval_steps_per_second": 6.257, |
| "step": 64000 |
| }, |
| { |
| "epoch": 6.893768162738134, |
| "grad_norm": 0.7628564834594727, |
| "learning_rate": 0.0001867363430664799, |
| "loss": 3.292, |
| "step": 64050 |
| }, |
| { |
| "epoch": 6.899149714777742, |
| "grad_norm": 0.8125451803207397, |
| "learning_rate": 0.00018641310203641848, |
| "loss": 3.2681, |
| "step": 64100 |
| }, |
| { |
| "epoch": 6.90453126681735, |
| "grad_norm": 0.7533065676689148, |
| "learning_rate": 0.00018608986100635705, |
| "loss": 3.2843, |
| "step": 64150 |
| }, |
| { |
| "epoch": 6.9099128188569585, |
| "grad_norm": 0.8026787638664246, |
| "learning_rate": 0.00018577308479689685, |
| "loss": 3.3001, |
| "step": 64200 |
| }, |
| { |
| "epoch": 6.915294370896566, |
| "grad_norm": 0.7561555504798889, |
| "learning_rate": 0.00018544984376683545, |
| "loss": 3.2767, |
| "step": 64250 |
| }, |
| { |
| "epoch": 6.920675922936175, |
| "grad_norm": 0.764043927192688, |
| "learning_rate": 0.00018512660273677404, |
| "loss": 3.2937, |
| "step": 64300 |
| }, |
| { |
| "epoch": 6.926057474975783, |
| "grad_norm": 0.7226858139038086, |
| "learning_rate": 0.0001848033617067126, |
| "loss": 3.2866, |
| "step": 64350 |
| }, |
| { |
| "epoch": 6.931439027015391, |
| "grad_norm": 0.7623168230056763, |
| "learning_rate": 0.00018448012067665123, |
| "loss": 3.2637, |
| "step": 64400 |
| }, |
| { |
| "epoch": 6.9368205790549995, |
| "grad_norm": 0.7732171416282654, |
| "learning_rate": 0.0001841568796465898, |
| "loss": 3.2752, |
| "step": 64450 |
| }, |
| { |
| "epoch": 6.942202131094608, |
| "grad_norm": 0.7387587428092957, |
| "learning_rate": 0.00018383363861652837, |
| "loss": 3.2894, |
| "step": 64500 |
| }, |
| { |
| "epoch": 6.947583683134216, |
| "grad_norm": 0.7186091542243958, |
| "learning_rate": 0.00018351039758646696, |
| "loss": 3.2944, |
| "step": 64550 |
| }, |
| { |
| "epoch": 6.952965235173824, |
| "grad_norm": 0.7569155097007751, |
| "learning_rate": 0.00018318715655640555, |
| "loss": 3.2855, |
| "step": 64600 |
| }, |
| { |
| "epoch": 6.958346787213433, |
| "grad_norm": 0.7623655200004578, |
| "learning_rate": 0.00018286391552634412, |
| "loss": 3.281, |
| "step": 64650 |
| }, |
| { |
| "epoch": 6.9637283392530405, |
| "grad_norm": 0.8205636143684387, |
| "learning_rate": 0.00018254067449628272, |
| "loss": 3.281, |
| "step": 64700 |
| }, |
| { |
| "epoch": 6.969109891292649, |
| "grad_norm": 0.7782757878303528, |
| "learning_rate": 0.00018221743346622128, |
| "loss": 3.2977, |
| "step": 64750 |
| }, |
| { |
| "epoch": 6.974491443332257, |
| "grad_norm": 0.7646986246109009, |
| "learning_rate": 0.00018189419243615988, |
| "loss": 3.2798, |
| "step": 64800 |
| }, |
| { |
| "epoch": 6.979872995371865, |
| "grad_norm": 0.7757244110107422, |
| "learning_rate": 0.00018157095140609847, |
| "loss": 3.287, |
| "step": 64850 |
| }, |
| { |
| "epoch": 6.985254547411474, |
| "grad_norm": 0.7472320199012756, |
| "learning_rate": 0.00018125417519663825, |
| "loss": 3.2609, |
| "step": 64900 |
| }, |
| { |
| "epoch": 6.990636099451081, |
| "grad_norm": 0.7466889023780823, |
| "learning_rate": 0.00018093093416657687, |
| "loss": 3.2883, |
| "step": 64950 |
| }, |
| { |
| "epoch": 6.99601765149069, |
| "grad_norm": 0.7409732937812805, |
| "learning_rate": 0.00018060769313651544, |
| "loss": 3.284, |
| "step": 65000 |
| }, |
| { |
| "epoch": 6.99601765149069, |
| "eval_accuracy": 0.3868794684179194, |
| "eval_loss": 3.35493803024292, |
| "eval_runtime": 179.8236, |
| "eval_samples_per_second": 100.159, |
| "eval_steps_per_second": 6.262, |
| "step": 65000 |
| }, |
| { |
| "epoch": 7.0013992035302985, |
| "grad_norm": 0.7562103867530823, |
| "learning_rate": 0.000180284452106454, |
| "loss": 3.2676, |
| "step": 65050 |
| }, |
| { |
| "epoch": 7.006780755569906, |
| "grad_norm": 0.7899980545043945, |
| "learning_rate": 0.00017996121107639263, |
| "loss": 3.223, |
| "step": 65100 |
| }, |
| { |
| "epoch": 7.012162307609515, |
| "grad_norm": 0.7686735391616821, |
| "learning_rate": 0.0001796379700463312, |
| "loss": 3.2053, |
| "step": 65150 |
| }, |
| { |
| "epoch": 7.017543859649122, |
| "grad_norm": 0.7950847744941711, |
| "learning_rate": 0.0001793147290162698, |
| "loss": 3.1919, |
| "step": 65200 |
| }, |
| { |
| "epoch": 7.022925411688731, |
| "grad_norm": 0.8973643183708191, |
| "learning_rate": 0.00017899148798620836, |
| "loss": 3.1952, |
| "step": 65250 |
| }, |
| { |
| "epoch": 7.0283069637283395, |
| "grad_norm": 0.7614500522613525, |
| "learning_rate": 0.00017866824695614696, |
| "loss": 3.1957, |
| "step": 65300 |
| }, |
| { |
| "epoch": 7.033688515767947, |
| "grad_norm": 0.8014223575592041, |
| "learning_rate": 0.00017834500592608555, |
| "loss": 3.1986, |
| "step": 65350 |
| }, |
| { |
| "epoch": 7.039070067807556, |
| "grad_norm": 0.7725815773010254, |
| "learning_rate": 0.00017802176489602412, |
| "loss": 3.1856, |
| "step": 65400 |
| }, |
| { |
| "epoch": 7.044451619847164, |
| "grad_norm": 0.7577409744262695, |
| "learning_rate": 0.00017769852386596269, |
| "loss": 3.2132, |
| "step": 65450 |
| }, |
| { |
| "epoch": 7.049833171886772, |
| "grad_norm": 0.7576920986175537, |
| "learning_rate": 0.0001773752828359013, |
| "loss": 3.1937, |
| "step": 65500 |
| }, |
| { |
| "epoch": 7.0552147239263805, |
| "grad_norm": 0.7937734723091125, |
| "learning_rate": 0.00017705204180583988, |
| "loss": 3.2073, |
| "step": 65550 |
| }, |
| { |
| "epoch": 7.060596275965988, |
| "grad_norm": 0.7605912685394287, |
| "learning_rate": 0.00017672880077577844, |
| "loss": 3.2075, |
| "step": 65600 |
| }, |
| { |
| "epoch": 7.065977828005597, |
| "grad_norm": 0.7895816564559937, |
| "learning_rate": 0.00017640555974571704, |
| "loss": 3.2088, |
| "step": 65650 |
| }, |
| { |
| "epoch": 7.071359380045205, |
| "grad_norm": 0.8300418853759766, |
| "learning_rate": 0.00017608231871565563, |
| "loss": 3.206, |
| "step": 65700 |
| }, |
| { |
| "epoch": 7.076740932084813, |
| "grad_norm": 0.7783611416816711, |
| "learning_rate": 0.0001757590776855942, |
| "loss": 3.2163, |
| "step": 65750 |
| }, |
| { |
| "epoch": 7.0821224841244215, |
| "grad_norm": 0.7723811864852905, |
| "learning_rate": 0.0001754358366555328, |
| "loss": 3.215, |
| "step": 65800 |
| }, |
| { |
| "epoch": 7.08750403616403, |
| "grad_norm": 0.7977986931800842, |
| "learning_rate": 0.00017511259562547136, |
| "loss": 3.2125, |
| "step": 65850 |
| }, |
| { |
| "epoch": 7.092885588203638, |
| "grad_norm": 0.7651707530021667, |
| "learning_rate": 0.00017478935459540996, |
| "loss": 3.2135, |
| "step": 65900 |
| }, |
| { |
| "epoch": 7.098267140243246, |
| "grad_norm": 0.812610924243927, |
| "learning_rate": 0.00017446611356534855, |
| "loss": 3.2242, |
| "step": 65950 |
| }, |
| { |
| "epoch": 7.103648692282855, |
| "grad_norm": 0.7896516919136047, |
| "learning_rate": 0.00017414287253528712, |
| "loss": 3.1994, |
| "step": 66000 |
| }, |
| { |
| "epoch": 7.103648692282855, |
| "eval_accuracy": 0.3869793204218209, |
| "eval_loss": 3.361632823944092, |
| "eval_runtime": 179.86, |
| "eval_samples_per_second": 100.139, |
| "eval_steps_per_second": 6.26, |
| "step": 66000 |
| }, |
| { |
| "epoch": 7.109030244322462, |
| "grad_norm": 0.7712588310241699, |
| "learning_rate": 0.00017381963150522569, |
| "loss": 3.2167, |
| "step": 66050 |
| }, |
| { |
| "epoch": 7.114411796362071, |
| "grad_norm": 0.7564050555229187, |
| "learning_rate": 0.0001734963904751643, |
| "loss": 3.2134, |
| "step": 66100 |
| }, |
| { |
| "epoch": 7.119793348401679, |
| "grad_norm": 0.7921330332756042, |
| "learning_rate": 0.00017317314944510288, |
| "loss": 3.2075, |
| "step": 66150 |
| }, |
| { |
| "epoch": 7.125174900441287, |
| "grad_norm": 0.8234257698059082, |
| "learning_rate": 0.00017284990841504147, |
| "loss": 3.2197, |
| "step": 66200 |
| }, |
| { |
| "epoch": 7.130556452480896, |
| "grad_norm": 0.7843002080917358, |
| "learning_rate": 0.00017252666738498007, |
| "loss": 3.2197, |
| "step": 66250 |
| }, |
| { |
| "epoch": 7.135938004520503, |
| "grad_norm": 0.7515401244163513, |
| "learning_rate": 0.00017220342635491863, |
| "loss": 3.209, |
| "step": 66300 |
| }, |
| { |
| "epoch": 7.141319556560112, |
| "grad_norm": 0.841936469078064, |
| "learning_rate": 0.00017188018532485723, |
| "loss": 3.2314, |
| "step": 66350 |
| }, |
| { |
| "epoch": 7.1467011085997205, |
| "grad_norm": 0.7923972010612488, |
| "learning_rate": 0.0001715569442947958, |
| "loss": 3.2195, |
| "step": 66400 |
| }, |
| { |
| "epoch": 7.152082660639328, |
| "grad_norm": 0.7993304133415222, |
| "learning_rate": 0.0001712337032647344, |
| "loss": 3.2214, |
| "step": 66450 |
| }, |
| { |
| "epoch": 7.157464212678937, |
| "grad_norm": 0.7993540167808533, |
| "learning_rate": 0.00017091046223467298, |
| "loss": 3.2377, |
| "step": 66500 |
| }, |
| { |
| "epoch": 7.162845764718545, |
| "grad_norm": 0.8128913640975952, |
| "learning_rate": 0.00017058722120461155, |
| "loss": 3.2127, |
| "step": 66550 |
| }, |
| { |
| "epoch": 7.168227316758153, |
| "grad_norm": 0.7577317357063293, |
| "learning_rate": 0.00017026398017455012, |
| "loss": 3.2401, |
| "step": 66600 |
| }, |
| { |
| "epoch": 7.1736088687977615, |
| "grad_norm": 0.8127763271331787, |
| "learning_rate": 0.00016994073914448874, |
| "loss": 3.2001, |
| "step": 66650 |
| }, |
| { |
| "epoch": 7.178990420837369, |
| "grad_norm": 0.7880443930625916, |
| "learning_rate": 0.0001696174981144273, |
| "loss": 3.2313, |
| "step": 66700 |
| }, |
| { |
| "epoch": 7.184371972876978, |
| "grad_norm": 0.7774691581726074, |
| "learning_rate": 0.00016929425708436588, |
| "loss": 3.2234, |
| "step": 66750 |
| }, |
| { |
| "epoch": 7.189753524916586, |
| "grad_norm": 0.8356835246086121, |
| "learning_rate": 0.0001689710160543045, |
| "loss": 3.203, |
| "step": 66800 |
| }, |
| { |
| "epoch": 7.195135076956194, |
| "grad_norm": 0.8748675584793091, |
| "learning_rate": 0.00016864777502424307, |
| "loss": 3.2245, |
| "step": 66850 |
| }, |
| { |
| "epoch": 7.2005166289958025, |
| "grad_norm": 0.7685278654098511, |
| "learning_rate": 0.00016832453399418163, |
| "loss": 3.225, |
| "step": 66900 |
| }, |
| { |
| "epoch": 7.205898181035411, |
| "grad_norm": 0.7593712210655212, |
| "learning_rate": 0.00016800129296412023, |
| "loss": 3.2192, |
| "step": 66950 |
| }, |
| { |
| "epoch": 7.211279733075019, |
| "grad_norm": 0.7982090711593628, |
| "learning_rate": 0.00016767805193405882, |
| "loss": 3.2195, |
| "step": 67000 |
| }, |
| { |
| "epoch": 7.211279733075019, |
| "eval_accuracy": 0.3874346846746514, |
| "eval_loss": 3.359243631362915, |
| "eval_runtime": 179.5957, |
| "eval_samples_per_second": 100.286, |
| "eval_steps_per_second": 6.27, |
| "step": 67000 |
| }, |
| { |
| "epoch": 7.216661285114627, |
| "grad_norm": 0.7449146509170532, |
| "learning_rate": 0.0001673548109039974, |
| "loss": 3.2246, |
| "step": 67050 |
| }, |
| { |
| "epoch": 7.222042837154235, |
| "grad_norm": 0.7682053446769714, |
| "learning_rate": 0.00016703156987393598, |
| "loss": 3.2388, |
| "step": 67100 |
| }, |
| { |
| "epoch": 7.2274243891938434, |
| "grad_norm": 0.7306789755821228, |
| "learning_rate": 0.00016670832884387455, |
| "loss": 3.2375, |
| "step": 67150 |
| }, |
| { |
| "epoch": 7.232805941233452, |
| "grad_norm": 0.7582760453224182, |
| "learning_rate": 0.00016638508781381317, |
| "loss": 3.2174, |
| "step": 67200 |
| }, |
| { |
| "epoch": 7.23818749327306, |
| "grad_norm": 0.7850392460823059, |
| "learning_rate": 0.00016606184678375174, |
| "loss": 3.2253, |
| "step": 67250 |
| }, |
| { |
| "epoch": 7.243569045312668, |
| "grad_norm": 0.7982510924339294, |
| "learning_rate": 0.0001657386057536903, |
| "loss": 3.2302, |
| "step": 67300 |
| }, |
| { |
| "epoch": 7.248950597352277, |
| "grad_norm": 0.8140314817428589, |
| "learning_rate": 0.00016541536472362893, |
| "loss": 3.2346, |
| "step": 67350 |
| }, |
| { |
| "epoch": 7.254332149391884, |
| "grad_norm": 0.7745968699455261, |
| "learning_rate": 0.0001650921236935675, |
| "loss": 3.2229, |
| "step": 67400 |
| }, |
| { |
| "epoch": 7.259713701431493, |
| "grad_norm": 0.7968515753746033, |
| "learning_rate": 0.00016476888266350607, |
| "loss": 3.2248, |
| "step": 67450 |
| }, |
| { |
| "epoch": 7.265095253471101, |
| "grad_norm": 0.7656205892562866, |
| "learning_rate": 0.00016444564163344466, |
| "loss": 3.2252, |
| "step": 67500 |
| }, |
| { |
| "epoch": 7.270476805510709, |
| "grad_norm": 0.8577672243118286, |
| "learning_rate": 0.00016412240060338326, |
| "loss": 3.2215, |
| "step": 67550 |
| }, |
| { |
| "epoch": 7.275858357550318, |
| "grad_norm": 0.795576274394989, |
| "learning_rate": 0.00016379915957332182, |
| "loss": 3.2365, |
| "step": 67600 |
| }, |
| { |
| "epoch": 7.281239909589925, |
| "grad_norm": 0.7914314866065979, |
| "learning_rate": 0.00016347591854326042, |
| "loss": 3.2276, |
| "step": 67650 |
| }, |
| { |
| "epoch": 7.286621461629534, |
| "grad_norm": 0.7388089299201965, |
| "learning_rate": 0.00016315267751319898, |
| "loss": 3.2328, |
| "step": 67700 |
| }, |
| { |
| "epoch": 7.2920030136691425, |
| "grad_norm": 0.8092725276947021, |
| "learning_rate": 0.00016282943648313755, |
| "loss": 3.221, |
| "step": 67750 |
| }, |
| { |
| "epoch": 7.29738456570875, |
| "grad_norm": 0.7859850525856018, |
| "learning_rate": 0.00016250619545307617, |
| "loss": 3.2308, |
| "step": 67800 |
| }, |
| { |
| "epoch": 7.302766117748359, |
| "grad_norm": 0.7894166111946106, |
| "learning_rate": 0.00016218295442301474, |
| "loss": 3.2362, |
| "step": 67850 |
| }, |
| { |
| "epoch": 7.308147669787967, |
| "grad_norm": 0.8145300149917603, |
| "learning_rate": 0.0001618597133929533, |
| "loss": 3.2318, |
| "step": 67900 |
| }, |
| { |
| "epoch": 7.313529221827575, |
| "grad_norm": 0.8098170757293701, |
| "learning_rate": 0.00016153647236289193, |
| "loss": 3.2284, |
| "step": 67950 |
| }, |
| { |
| "epoch": 7.3189107738671835, |
| "grad_norm": 0.8146171569824219, |
| "learning_rate": 0.0001612132313328305, |
| "loss": 3.2313, |
| "step": 68000 |
| }, |
| { |
| "epoch": 7.3189107738671835, |
| "eval_accuracy": 0.3875706094373073, |
| "eval_loss": 3.3579890727996826, |
| "eval_runtime": 179.851, |
| "eval_samples_per_second": 100.144, |
| "eval_steps_per_second": 6.261, |
| "step": 68000 |
| }, |
| { |
| "epoch": 7.324292325906791, |
| "grad_norm": 0.8207062482833862, |
| "learning_rate": 0.0001608899903027691, |
| "loss": 3.238, |
| "step": 68050 |
| }, |
| { |
| "epoch": 7.3296738779464, |
| "grad_norm": 1.0600287914276123, |
| "learning_rate": 0.00016056674927270766, |
| "loss": 3.2332, |
| "step": 68100 |
| }, |
| { |
| "epoch": 7.335055429986008, |
| "grad_norm": 0.7621212601661682, |
| "learning_rate": 0.00016024350824264626, |
| "loss": 3.2204, |
| "step": 68150 |
| }, |
| { |
| "epoch": 7.340436982025616, |
| "grad_norm": 0.804821789264679, |
| "learning_rate": 0.00015992026721258485, |
| "loss": 3.2355, |
| "step": 68200 |
| }, |
| { |
| "epoch": 7.3458185340652244, |
| "grad_norm": 0.7746620774269104, |
| "learning_rate": 0.00015959702618252342, |
| "loss": 3.2425, |
| "step": 68250 |
| }, |
| { |
| "epoch": 7.351200086104833, |
| "grad_norm": 0.7920762896537781, |
| "learning_rate": 0.00015927378515246199, |
| "loss": 3.2154, |
| "step": 68300 |
| }, |
| { |
| "epoch": 7.356581638144441, |
| "grad_norm": 0.8123354315757751, |
| "learning_rate": 0.0001589505441224006, |
| "loss": 3.2222, |
| "step": 68350 |
| }, |
| { |
| "epoch": 7.361963190184049, |
| "grad_norm": 0.7643717527389526, |
| "learning_rate": 0.00015862730309233917, |
| "loss": 3.2192, |
| "step": 68400 |
| }, |
| { |
| "epoch": 7.367344742223658, |
| "grad_norm": 0.7962055802345276, |
| "learning_rate": 0.00015830406206227774, |
| "loss": 3.2198, |
| "step": 68450 |
| }, |
| { |
| "epoch": 7.372726294263265, |
| "grad_norm": 0.7841677069664001, |
| "learning_rate": 0.00015798082103221636, |
| "loss": 3.2349, |
| "step": 68500 |
| }, |
| { |
| "epoch": 7.378107846302874, |
| "grad_norm": 0.8586899042129517, |
| "learning_rate": 0.00015765758000215493, |
| "loss": 3.2325, |
| "step": 68550 |
| }, |
| { |
| "epoch": 7.383489398342482, |
| "grad_norm": 0.7646048665046692, |
| "learning_rate": 0.0001573343389720935, |
| "loss": 3.2206, |
| "step": 68600 |
| }, |
| { |
| "epoch": 7.38887095038209, |
| "grad_norm": 0.7651141285896301, |
| "learning_rate": 0.0001570110979420321, |
| "loss": 3.2242, |
| "step": 68650 |
| }, |
| { |
| "epoch": 7.394252502421699, |
| "grad_norm": 0.7675405144691467, |
| "learning_rate": 0.0001566943217325719, |
| "loss": 3.221, |
| "step": 68700 |
| }, |
| { |
| "epoch": 7.399634054461306, |
| "grad_norm": 0.7907142639160156, |
| "learning_rate": 0.0001563710807025105, |
| "loss": 3.2205, |
| "step": 68750 |
| }, |
| { |
| "epoch": 7.405015606500915, |
| "grad_norm": 0.7915471792221069, |
| "learning_rate": 0.00015604783967244906, |
| "loss": 3.2347, |
| "step": 68800 |
| }, |
| { |
| "epoch": 7.4103971585405235, |
| "grad_norm": 0.8003562092781067, |
| "learning_rate": 0.00015572459864238763, |
| "loss": 3.2305, |
| "step": 68850 |
| }, |
| { |
| "epoch": 7.415778710580131, |
| "grad_norm": 0.7971240282058716, |
| "learning_rate": 0.00015540135761232625, |
| "loss": 3.2337, |
| "step": 68900 |
| }, |
| { |
| "epoch": 7.42116026261974, |
| "grad_norm": 0.7408497333526611, |
| "learning_rate": 0.00015507811658226482, |
| "loss": 3.2173, |
| "step": 68950 |
| }, |
| { |
| "epoch": 7.426541814659347, |
| "grad_norm": 0.803286075592041, |
| "learning_rate": 0.00015475487555220344, |
| "loss": 3.2322, |
| "step": 69000 |
| }, |
| { |
| "epoch": 7.426541814659347, |
| "eval_accuracy": 0.38788418167153993, |
| "eval_loss": 3.3515703678131104, |
| "eval_runtime": 179.8641, |
| "eval_samples_per_second": 100.137, |
| "eval_steps_per_second": 6.26, |
| "step": 69000 |
| }, |
| { |
| "epoch": 7.431923366698956, |
| "grad_norm": 0.7986598014831543, |
| "learning_rate": 0.000154431634522142, |
| "loss": 3.2336, |
| "step": 69050 |
| }, |
| { |
| "epoch": 7.4373049187385645, |
| "grad_norm": 0.7994623184204102, |
| "learning_rate": 0.00015410839349208058, |
| "loss": 3.2172, |
| "step": 69100 |
| }, |
| { |
| "epoch": 7.442686470778172, |
| "grad_norm": 0.802762508392334, |
| "learning_rate": 0.00015378515246201917, |
| "loss": 3.2326, |
| "step": 69150 |
| }, |
| { |
| "epoch": 7.448068022817781, |
| "grad_norm": 0.768039882183075, |
| "learning_rate": 0.00015346191143195774, |
| "loss": 3.2339, |
| "step": 69200 |
| }, |
| { |
| "epoch": 7.453449574857389, |
| "grad_norm": 0.818593442440033, |
| "learning_rate": 0.00015313867040189633, |
| "loss": 3.2376, |
| "step": 69250 |
| }, |
| { |
| "epoch": 7.458831126896997, |
| "grad_norm": 0.8104583621025085, |
| "learning_rate": 0.00015281542937183493, |
| "loss": 3.2267, |
| "step": 69300 |
| }, |
| { |
| "epoch": 7.4642126789366054, |
| "grad_norm": 0.806210458278656, |
| "learning_rate": 0.0001524921883417735, |
| "loss": 3.2389, |
| "step": 69350 |
| }, |
| { |
| "epoch": 7.469594230976213, |
| "grad_norm": 0.7950648665428162, |
| "learning_rate": 0.00015216894731171206, |
| "loss": 3.2193, |
| "step": 69400 |
| }, |
| { |
| "epoch": 7.474975783015822, |
| "grad_norm": 0.8381013870239258, |
| "learning_rate": 0.00015184570628165068, |
| "loss": 3.2313, |
| "step": 69450 |
| }, |
| { |
| "epoch": 7.48035733505543, |
| "grad_norm": 0.7694802284240723, |
| "learning_rate": 0.00015152246525158925, |
| "loss": 3.2385, |
| "step": 69500 |
| }, |
| { |
| "epoch": 7.485738887095038, |
| "grad_norm": 0.8459978699684143, |
| "learning_rate": 0.00015119922422152782, |
| "loss": 3.2132, |
| "step": 69550 |
| }, |
| { |
| "epoch": 7.491120439134646, |
| "grad_norm": 0.8075374960899353, |
| "learning_rate": 0.00015087598319146644, |
| "loss": 3.2399, |
| "step": 69600 |
| }, |
| { |
| "epoch": 7.496501991174255, |
| "grad_norm": 0.7973178029060364, |
| "learning_rate": 0.000150552742161405, |
| "loss": 3.2294, |
| "step": 69650 |
| }, |
| { |
| "epoch": 7.501883543213863, |
| "grad_norm": 0.7813896536827087, |
| "learning_rate": 0.00015022950113134358, |
| "loss": 3.228, |
| "step": 69700 |
| }, |
| { |
| "epoch": 7.507265095253471, |
| "grad_norm": 0.8545538187026978, |
| "learning_rate": 0.00014990626010128217, |
| "loss": 3.2443, |
| "step": 69750 |
| }, |
| { |
| "epoch": 7.51264664729308, |
| "grad_norm": 0.7499383091926575, |
| "learning_rate": 0.00014958301907122077, |
| "loss": 3.2401, |
| "step": 69800 |
| }, |
| { |
| "epoch": 7.518028199332687, |
| "grad_norm": 0.8156957030296326, |
| "learning_rate": 0.00014925977804115933, |
| "loss": 3.2466, |
| "step": 69850 |
| }, |
| { |
| "epoch": 7.523409751372296, |
| "grad_norm": 0.8415682315826416, |
| "learning_rate": 0.00014893653701109793, |
| "loss": 3.2477, |
| "step": 69900 |
| }, |
| { |
| "epoch": 7.528791303411904, |
| "grad_norm": 0.7899500131607056, |
| "learning_rate": 0.0001486132959810365, |
| "loss": 3.2335, |
| "step": 69950 |
| }, |
| { |
| "epoch": 7.534172855451512, |
| "grad_norm": 0.819366991519928, |
| "learning_rate": 0.0001482900549509751, |
| "loss": 3.2278, |
| "step": 70000 |
| }, |
| { |
| "epoch": 7.534172855451512, |
| "eval_accuracy": 0.38836833893964123, |
| "eval_loss": 3.3499035835266113, |
| "eval_runtime": 180.0283, |
| "eval_samples_per_second": 100.045, |
| "eval_steps_per_second": 6.255, |
| "step": 70000 |
| }, |
| { |
| "epoch": 7.539554407491121, |
| "grad_norm": 0.8849138021469116, |
| "learning_rate": 0.00014796681392091368, |
| "loss": 3.2443, |
| "step": 70050 |
| }, |
| { |
| "epoch": 7.544935959530728, |
| "grad_norm": 0.8181607723236084, |
| "learning_rate": 0.00014764357289085228, |
| "loss": 3.2384, |
| "step": 70100 |
| }, |
| { |
| "epoch": 7.550317511570337, |
| "grad_norm": 0.9902865290641785, |
| "learning_rate": 0.00014732033186079085, |
| "loss": 3.2333, |
| "step": 70150 |
| }, |
| { |
| "epoch": 7.5556990636099455, |
| "grad_norm": 0.7719454169273376, |
| "learning_rate": 0.00014699709083072944, |
| "loss": 3.2358, |
| "step": 70200 |
| }, |
| { |
| "epoch": 7.561080615649553, |
| "grad_norm": 0.830987274646759, |
| "learning_rate": 0.00014667384980066804, |
| "loss": 3.2325, |
| "step": 70250 |
| }, |
| { |
| "epoch": 7.566462167689162, |
| "grad_norm": 0.8358805775642395, |
| "learning_rate": 0.0001463506087706066, |
| "loss": 3.2332, |
| "step": 70300 |
| }, |
| { |
| "epoch": 7.57184371972877, |
| "grad_norm": 0.7925933003425598, |
| "learning_rate": 0.0001460273677405452, |
| "loss": 3.2502, |
| "step": 70350 |
| }, |
| { |
| "epoch": 7.577225271768378, |
| "grad_norm": 0.8156459927558899, |
| "learning_rate": 0.00014570412671048377, |
| "loss": 3.2425, |
| "step": 70400 |
| }, |
| { |
| "epoch": 7.5826068238079865, |
| "grad_norm": 0.8065808415412903, |
| "learning_rate": 0.00014538088568042236, |
| "loss": 3.2272, |
| "step": 70450 |
| }, |
| { |
| "epoch": 7.587988375847594, |
| "grad_norm": 0.8279649019241333, |
| "learning_rate": 0.00014505764465036093, |
| "loss": 3.232, |
| "step": 70500 |
| }, |
| { |
| "epoch": 7.593369927887203, |
| "grad_norm": 0.8306601047515869, |
| "learning_rate": 0.00014473440362029952, |
| "loss": 3.2414, |
| "step": 70550 |
| }, |
| { |
| "epoch": 7.598751479926811, |
| "grad_norm": 0.8502399325370789, |
| "learning_rate": 0.00014441116259023812, |
| "loss": 3.2204, |
| "step": 70600 |
| }, |
| { |
| "epoch": 7.604133031966419, |
| "grad_norm": 0.8230718970298767, |
| "learning_rate": 0.00014408792156017669, |
| "loss": 3.2361, |
| "step": 70650 |
| }, |
| { |
| "epoch": 7.609514584006027, |
| "grad_norm": 0.8219791650772095, |
| "learning_rate": 0.00014376468053011528, |
| "loss": 3.2281, |
| "step": 70700 |
| }, |
| { |
| "epoch": 7.614896136045635, |
| "grad_norm": 0.8464500904083252, |
| "learning_rate": 0.00014344143950005387, |
| "loss": 3.2384, |
| "step": 70750 |
| }, |
| { |
| "epoch": 7.620277688085244, |
| "grad_norm": 0.8252779245376587, |
| "learning_rate": 0.00014311819846999244, |
| "loss": 3.2316, |
| "step": 70800 |
| }, |
| { |
| "epoch": 7.625659240124852, |
| "grad_norm": 0.808069109916687, |
| "learning_rate": 0.00014279495743993104, |
| "loss": 3.2265, |
| "step": 70850 |
| }, |
| { |
| "epoch": 7.63104079216446, |
| "grad_norm": 0.8365920782089233, |
| "learning_rate": 0.00014247171640986963, |
| "loss": 3.2415, |
| "step": 70900 |
| }, |
| { |
| "epoch": 7.636422344204068, |
| "grad_norm": 0.7884583473205566, |
| "learning_rate": 0.0001421484753798082, |
| "loss": 3.2226, |
| "step": 70950 |
| }, |
| { |
| "epoch": 7.641803896243677, |
| "grad_norm": 0.8387513160705566, |
| "learning_rate": 0.0001418252343497468, |
| "loss": 3.253, |
| "step": 71000 |
| }, |
| { |
| "epoch": 7.641803896243677, |
| "eval_accuracy": 0.3887576422366805, |
| "eval_loss": 3.345111608505249, |
| "eval_runtime": 179.568, |
| "eval_samples_per_second": 100.302, |
| "eval_steps_per_second": 6.271, |
| "step": 71000 |
| }, |
| { |
| "epoch": 7.647185448283285, |
| "grad_norm": 0.7808358669281006, |
| "learning_rate": 0.00014150199331968536, |
| "loss": 3.2328, |
| "step": 71050 |
| }, |
| { |
| "epoch": 7.652567000322893, |
| "grad_norm": 0.7849861979484558, |
| "learning_rate": 0.00014117875228962396, |
| "loss": 3.2317, |
| "step": 71100 |
| }, |
| { |
| "epoch": 7.657948552362502, |
| "grad_norm": 0.8119670152664185, |
| "learning_rate": 0.00014085551125956252, |
| "loss": 3.233, |
| "step": 71150 |
| }, |
| { |
| "epoch": 7.663330104402109, |
| "grad_norm": 0.8060752749443054, |
| "learning_rate": 0.00014053227022950112, |
| "loss": 3.2309, |
| "step": 71200 |
| }, |
| { |
| "epoch": 7.668711656441718, |
| "grad_norm": 0.8241363167762756, |
| "learning_rate": 0.0001402090291994397, |
| "loss": 3.238, |
| "step": 71250 |
| }, |
| { |
| "epoch": 7.674093208481326, |
| "grad_norm": 0.8252966403961182, |
| "learning_rate": 0.00013988578816937828, |
| "loss": 3.2444, |
| "step": 71300 |
| }, |
| { |
| "epoch": 7.679474760520934, |
| "grad_norm": 0.8020745515823364, |
| "learning_rate": 0.00013956254713931687, |
| "loss": 3.2285, |
| "step": 71350 |
| }, |
| { |
| "epoch": 7.684856312560543, |
| "grad_norm": 0.8435476422309875, |
| "learning_rate": 0.00013923930610925547, |
| "loss": 3.2405, |
| "step": 71400 |
| }, |
| { |
| "epoch": 7.69023786460015, |
| "grad_norm": 0.7923038005828857, |
| "learning_rate": 0.00013891606507919404, |
| "loss": 3.2201, |
| "step": 71450 |
| }, |
| { |
| "epoch": 7.695619416639759, |
| "grad_norm": 0.8090280890464783, |
| "learning_rate": 0.00013859282404913263, |
| "loss": 3.2247, |
| "step": 71500 |
| }, |
| { |
| "epoch": 7.7010009686793675, |
| "grad_norm": 0.8997794985771179, |
| "learning_rate": 0.0001382695830190712, |
| "loss": 3.2194, |
| "step": 71550 |
| }, |
| { |
| "epoch": 7.706382520718975, |
| "grad_norm": 0.8104215264320374, |
| "learning_rate": 0.0001379463419890098, |
| "loss": 3.2425, |
| "step": 71600 |
| }, |
| { |
| "epoch": 7.711764072758584, |
| "grad_norm": 0.8430519104003906, |
| "learning_rate": 0.00013762310095894836, |
| "loss": 3.242, |
| "step": 71650 |
| }, |
| { |
| "epoch": 7.717145624798192, |
| "grad_norm": 0.8335681557655334, |
| "learning_rate": 0.00013729985992888696, |
| "loss": 3.237, |
| "step": 71700 |
| }, |
| { |
| "epoch": 7.7225271768378, |
| "grad_norm": 0.8195766806602478, |
| "learning_rate": 0.00013697661889882555, |
| "loss": 3.2369, |
| "step": 71750 |
| }, |
| { |
| "epoch": 7.727908728877408, |
| "grad_norm": 0.8362144231796265, |
| "learning_rate": 0.00013665337786876412, |
| "loss": 3.226, |
| "step": 71800 |
| }, |
| { |
| "epoch": 7.733290280917016, |
| "grad_norm": 0.8281188011169434, |
| "learning_rate": 0.0001363301368387027, |
| "loss": 3.2287, |
| "step": 71850 |
| }, |
| { |
| "epoch": 7.738671832956625, |
| "grad_norm": 0.7755012512207031, |
| "learning_rate": 0.0001360068958086413, |
| "loss": 3.2165, |
| "step": 71900 |
| }, |
| { |
| "epoch": 7.744053384996233, |
| "grad_norm": 0.8271076083183289, |
| "learning_rate": 0.00013568365477857988, |
| "loss": 3.234, |
| "step": 71950 |
| }, |
| { |
| "epoch": 7.749434937035841, |
| "grad_norm": 0.7885873913764954, |
| "learning_rate": 0.00013536041374851847, |
| "loss": 3.2342, |
| "step": 72000 |
| }, |
| { |
| "epoch": 7.749434937035841, |
| "eval_accuracy": 0.38911706598957085, |
| "eval_loss": 3.3421249389648438, |
| "eval_runtime": 179.9591, |
| "eval_samples_per_second": 100.084, |
| "eval_steps_per_second": 6.257, |
| "step": 72000 |
| }, |
| { |
| "epoch": 7.754816489075449, |
| "grad_norm": 0.7660197615623474, |
| "learning_rate": 0.00013503717271845706, |
| "loss": 3.2277, |
| "step": 72050 |
| }, |
| { |
| "epoch": 7.760198041115058, |
| "grad_norm": 0.9116640090942383, |
| "learning_rate": 0.00013471393168839563, |
| "loss": 3.2477, |
| "step": 72100 |
| }, |
| { |
| "epoch": 7.765579593154666, |
| "grad_norm": 0.8500056862831116, |
| "learning_rate": 0.00013439069065833423, |
| "loss": 3.2316, |
| "step": 72150 |
| }, |
| { |
| "epoch": 7.770961145194274, |
| "grad_norm": 0.8157951831817627, |
| "learning_rate": 0.0001340674496282728, |
| "loss": 3.2313, |
| "step": 72200 |
| }, |
| { |
| "epoch": 7.776342697233883, |
| "grad_norm": 0.7920417785644531, |
| "learning_rate": 0.0001337442085982114, |
| "loss": 3.2379, |
| "step": 72250 |
| }, |
| { |
| "epoch": 7.78172424927349, |
| "grad_norm": 0.8302222490310669, |
| "learning_rate": 0.00013342096756814996, |
| "loss": 3.2365, |
| "step": 72300 |
| }, |
| { |
| "epoch": 7.787105801313099, |
| "grad_norm": 0.8150511980056763, |
| "learning_rate": 0.00013309772653808855, |
| "loss": 3.2492, |
| "step": 72350 |
| }, |
| { |
| "epoch": 7.792487353352707, |
| "grad_norm": 0.8173815608024597, |
| "learning_rate": 0.00013277448550802715, |
| "loss": 3.2356, |
| "step": 72400 |
| }, |
| { |
| "epoch": 7.797868905392315, |
| "grad_norm": 0.7868077754974365, |
| "learning_rate": 0.0001324512444779657, |
| "loss": 3.2376, |
| "step": 72450 |
| }, |
| { |
| "epoch": 7.803250457431924, |
| "grad_norm": 0.7838640213012695, |
| "learning_rate": 0.0001321280034479043, |
| "loss": 3.2176, |
| "step": 72500 |
| }, |
| { |
| "epoch": 7.808632009471531, |
| "grad_norm": 0.8046959042549133, |
| "learning_rate": 0.0001318047624178429, |
| "loss": 3.2453, |
| "step": 72550 |
| }, |
| { |
| "epoch": 7.81401356151114, |
| "grad_norm": 0.8591763377189636, |
| "learning_rate": 0.0001314815213877815, |
| "loss": 3.233, |
| "step": 72600 |
| }, |
| { |
| "epoch": 7.819395113550748, |
| "grad_norm": 0.7944203019142151, |
| "learning_rate": 0.00013115828035772007, |
| "loss": 3.2246, |
| "step": 72650 |
| }, |
| { |
| "epoch": 7.824776665590356, |
| "grad_norm": 0.8305500149726868, |
| "learning_rate": 0.00013084150414825987, |
| "loss": 3.2299, |
| "step": 72700 |
| }, |
| { |
| "epoch": 7.830158217629965, |
| "grad_norm": 0.8165892362594604, |
| "learning_rate": 0.00013051826311819844, |
| "loss": 3.247, |
| "step": 72750 |
| }, |
| { |
| "epoch": 7.835539769669572, |
| "grad_norm": 0.8104371428489685, |
| "learning_rate": 0.00013019502208813703, |
| "loss": 3.2239, |
| "step": 72800 |
| }, |
| { |
| "epoch": 7.840921321709181, |
| "grad_norm": 0.808006227016449, |
| "learning_rate": 0.00012987178105807563, |
| "loss": 3.231, |
| "step": 72850 |
| }, |
| { |
| "epoch": 7.846302873748789, |
| "grad_norm": 0.7946688532829285, |
| "learning_rate": 0.00012954854002801422, |
| "loss": 3.2496, |
| "step": 72900 |
| }, |
| { |
| "epoch": 7.851684425788397, |
| "grad_norm": 0.7898575663566589, |
| "learning_rate": 0.00012923176381855403, |
| "loss": 3.2305, |
| "step": 72950 |
| }, |
| { |
| "epoch": 7.857065977828006, |
| "grad_norm": 0.8683642148971558, |
| "learning_rate": 0.00012890852278849262, |
| "loss": 3.2291, |
| "step": 73000 |
| }, |
| { |
| "epoch": 7.857065977828006, |
| "eval_accuracy": 0.38946323409014777, |
| "eval_loss": 3.33805513381958, |
| "eval_runtime": 179.9574, |
| "eval_samples_per_second": 100.085, |
| "eval_steps_per_second": 6.257, |
| "step": 73000 |
| }, |
| { |
| "epoch": 7.862447529867614, |
| "grad_norm": 0.8216552138328552, |
| "learning_rate": 0.0001285852817584312, |
| "loss": 3.2354, |
| "step": 73050 |
| }, |
| { |
| "epoch": 7.867829081907222, |
| "grad_norm": 0.8108968138694763, |
| "learning_rate": 0.0001282620407283698, |
| "loss": 3.2364, |
| "step": 73100 |
| }, |
| { |
| "epoch": 7.87321063394683, |
| "grad_norm": 0.8014196157455444, |
| "learning_rate": 0.00012793879969830835, |
| "loss": 3.2347, |
| "step": 73150 |
| }, |
| { |
| "epoch": 7.878592185986438, |
| "grad_norm": 0.7955536246299744, |
| "learning_rate": 0.00012761555866824695, |
| "loss": 3.2308, |
| "step": 73200 |
| }, |
| { |
| "epoch": 7.883973738026047, |
| "grad_norm": 0.9082058072090149, |
| "learning_rate": 0.00012729231763818552, |
| "loss": 3.2434, |
| "step": 73250 |
| }, |
| { |
| "epoch": 7.889355290065655, |
| "grad_norm": 0.842169463634491, |
| "learning_rate": 0.0001269690766081241, |
| "loss": 3.2563, |
| "step": 73300 |
| }, |
| { |
| "epoch": 7.894736842105263, |
| "grad_norm": 0.8316313028335571, |
| "learning_rate": 0.0001266458355780627, |
| "loss": 3.2582, |
| "step": 73350 |
| }, |
| { |
| "epoch": 7.900118394144871, |
| "grad_norm": 0.8240712285041809, |
| "learning_rate": 0.00012632259454800127, |
| "loss": 3.2216, |
| "step": 73400 |
| }, |
| { |
| "epoch": 7.90549994618448, |
| "grad_norm": 0.803352952003479, |
| "learning_rate": 0.00012599935351793987, |
| "loss": 3.2502, |
| "step": 73450 |
| }, |
| { |
| "epoch": 7.910881498224088, |
| "grad_norm": 0.8391147255897522, |
| "learning_rate": 0.00012567611248787846, |
| "loss": 3.2302, |
| "step": 73500 |
| }, |
| { |
| "epoch": 7.916263050263696, |
| "grad_norm": 0.8049103617668152, |
| "learning_rate": 0.00012535287145781703, |
| "loss": 3.2284, |
| "step": 73550 |
| }, |
| { |
| "epoch": 7.921644602303305, |
| "grad_norm": 0.8465777635574341, |
| "learning_rate": 0.00012502963042775562, |
| "loss": 3.2373, |
| "step": 73600 |
| }, |
| { |
| "epoch": 7.927026154342912, |
| "grad_norm": 0.8268304467201233, |
| "learning_rate": 0.00012470638939769422, |
| "loss": 3.2596, |
| "step": 73650 |
| }, |
| { |
| "epoch": 7.932407706382521, |
| "grad_norm": 0.8311086297035217, |
| "learning_rate": 0.0001243831483676328, |
| "loss": 3.2529, |
| "step": 73700 |
| }, |
| { |
| "epoch": 7.937789258422129, |
| "grad_norm": 0.8166625499725342, |
| "learning_rate": 0.00012405990733757138, |
| "loss": 3.2289, |
| "step": 73750 |
| }, |
| { |
| "epoch": 7.943170810461737, |
| "grad_norm": 0.805748462677002, |
| "learning_rate": 0.00012373666630750995, |
| "loss": 3.251, |
| "step": 73800 |
| }, |
| { |
| "epoch": 7.948552362501346, |
| "grad_norm": 0.8548930287361145, |
| "learning_rate": 0.00012341342527744854, |
| "loss": 3.2249, |
| "step": 73850 |
| }, |
| { |
| "epoch": 7.953933914540953, |
| "grad_norm": 0.8273134827613831, |
| "learning_rate": 0.0001230901842473871, |
| "loss": 3.2495, |
| "step": 73900 |
| }, |
| { |
| "epoch": 7.959315466580562, |
| "grad_norm": 0.782123327255249, |
| "learning_rate": 0.0001227669432173257, |
| "loss": 3.2184, |
| "step": 73950 |
| }, |
| { |
| "epoch": 7.96469701862017, |
| "grad_norm": 0.8412731885910034, |
| "learning_rate": 0.0001224437021872643, |
| "loss": 3.2333, |
| "step": 74000 |
| }, |
| { |
| "epoch": 7.96469701862017, |
| "eval_accuracy": 0.3899255521277984, |
| "eval_loss": 3.3341968059539795, |
| "eval_runtime": 180.0793, |
| "eval_samples_per_second": 100.017, |
| "eval_steps_per_second": 6.253, |
| "step": 74000 |
| }, |
| { |
| "epoch": 7.970078570659778, |
| "grad_norm": 0.8382527232170105, |
| "learning_rate": 0.00012212046115720287, |
| "loss": 3.2202, |
| "step": 74050 |
| }, |
| { |
| "epoch": 7.975460122699387, |
| "grad_norm": 0.8564316630363464, |
| "learning_rate": 0.00012179722012714146, |
| "loss": 3.2181, |
| "step": 74100 |
| }, |
| { |
| "epoch": 7.980841674738995, |
| "grad_norm": 0.9008880257606506, |
| "learning_rate": 0.00012147397909708004, |
| "loss": 3.2314, |
| "step": 74150 |
| }, |
| { |
| "epoch": 7.986223226778603, |
| "grad_norm": 0.7987130284309387, |
| "learning_rate": 0.00012115073806701864, |
| "loss": 3.2211, |
| "step": 74200 |
| }, |
| { |
| "epoch": 7.991604778818211, |
| "grad_norm": 0.8617604970932007, |
| "learning_rate": 0.0001208274970369572, |
| "loss": 3.2507, |
| "step": 74250 |
| }, |
| { |
| "epoch": 7.996986330857819, |
| "grad_norm": 0.8378741145133972, |
| "learning_rate": 0.0001205042560068958, |
| "loss": 3.2028, |
| "step": 74300 |
| }, |
| { |
| "epoch": 8.002367882897428, |
| "grad_norm": 0.8566402196884155, |
| "learning_rate": 0.0001201810149768344, |
| "loss": 3.2122, |
| "step": 74350 |
| }, |
| { |
| "epoch": 8.007749434937036, |
| "grad_norm": 0.8603155016899109, |
| "learning_rate": 0.00011985777394677296, |
| "loss": 3.1329, |
| "step": 74400 |
| }, |
| { |
| "epoch": 8.013130986976645, |
| "grad_norm": 0.8972188830375671, |
| "learning_rate": 0.00011953453291671154, |
| "loss": 3.162, |
| "step": 74450 |
| }, |
| { |
| "epoch": 8.018512539016251, |
| "grad_norm": 0.8691694140434265, |
| "learning_rate": 0.00011921129188665014, |
| "loss": 3.1586, |
| "step": 74500 |
| }, |
| { |
| "epoch": 8.02389409105586, |
| "grad_norm": 0.8346155285835266, |
| "learning_rate": 0.0001188880508565887, |
| "loss": 3.1576, |
| "step": 74550 |
| }, |
| { |
| "epoch": 8.029275643095469, |
| "grad_norm": 0.8023788332939148, |
| "learning_rate": 0.0001185648098265273, |
| "loss": 3.1625, |
| "step": 74600 |
| }, |
| { |
| "epoch": 8.034657195135077, |
| "grad_norm": 0.8025177121162415, |
| "learning_rate": 0.0001182415687964659, |
| "loss": 3.1463, |
| "step": 74650 |
| }, |
| { |
| "epoch": 8.040038747174686, |
| "grad_norm": 0.827228844165802, |
| "learning_rate": 0.00011791832776640448, |
| "loss": 3.1827, |
| "step": 74700 |
| }, |
| { |
| "epoch": 8.045420299214294, |
| "grad_norm": 0.8884971737861633, |
| "learning_rate": 0.00011759508673634306, |
| "loss": 3.1788, |
| "step": 74750 |
| }, |
| { |
| "epoch": 8.050801851253901, |
| "grad_norm": 0.8682385087013245, |
| "learning_rate": 0.00011727184570628164, |
| "loss": 3.1658, |
| "step": 74800 |
| }, |
| { |
| "epoch": 8.05618340329351, |
| "grad_norm": 0.847655713558197, |
| "learning_rate": 0.00011694860467622023, |
| "loss": 3.1655, |
| "step": 74850 |
| }, |
| { |
| "epoch": 8.061564955333118, |
| "grad_norm": 0.8776028156280518, |
| "learning_rate": 0.0001166253636461588, |
| "loss": 3.1699, |
| "step": 74900 |
| }, |
| { |
| "epoch": 8.066946507372727, |
| "grad_norm": 0.8304228782653809, |
| "learning_rate": 0.0001163021226160974, |
| "loss": 3.1682, |
| "step": 74950 |
| }, |
| { |
| "epoch": 8.072328059412335, |
| "grad_norm": 0.799704909324646, |
| "learning_rate": 0.00011597888158603598, |
| "loss": 3.1684, |
| "step": 75000 |
| }, |
| { |
| "epoch": 8.072328059412335, |
| "eval_accuracy": 0.3898515595111967, |
| "eval_loss": 3.3377063274383545, |
| "eval_runtime": 180.2546, |
| "eval_samples_per_second": 99.92, |
| "eval_steps_per_second": 6.247, |
| "step": 75000 |
| }, |
| { |
| "epoch": 8.077709611451942, |
| "grad_norm": 0.8407004475593567, |
| "learning_rate": 0.00011565564055597456, |
| "loss": 3.1608, |
| "step": 75050 |
| }, |
| { |
| "epoch": 8.08309116349155, |
| "grad_norm": 0.878447949886322, |
| "learning_rate": 0.00011533239952591314, |
| "loss": 3.1724, |
| "step": 75100 |
| }, |
| { |
| "epoch": 8.088472715531159, |
| "grad_norm": 0.8164194822311401, |
| "learning_rate": 0.00011500915849585173, |
| "loss": 3.1569, |
| "step": 75150 |
| }, |
| { |
| "epoch": 8.093854267570768, |
| "grad_norm": 0.8630013465881348, |
| "learning_rate": 0.00011469238228639154, |
| "loss": 3.1556, |
| "step": 75200 |
| }, |
| { |
| "epoch": 8.099235819610376, |
| "grad_norm": 0.8045172095298767, |
| "learning_rate": 0.00011436914125633012, |
| "loss": 3.1652, |
| "step": 75250 |
| }, |
| { |
| "epoch": 8.104617371649983, |
| "grad_norm": 0.8801988959312439, |
| "learning_rate": 0.00011404590022626872, |
| "loss": 3.1672, |
| "step": 75300 |
| }, |
| { |
| "epoch": 8.109998923689592, |
| "grad_norm": 0.8112406730651855, |
| "learning_rate": 0.00011372265919620728, |
| "loss": 3.1884, |
| "step": 75350 |
| }, |
| { |
| "epoch": 8.1153804757292, |
| "grad_norm": 0.8461372256278992, |
| "learning_rate": 0.00011339941816614588, |
| "loss": 3.1602, |
| "step": 75400 |
| }, |
| { |
| "epoch": 8.120762027768809, |
| "grad_norm": 0.9137732982635498, |
| "learning_rate": 0.00011307617713608447, |
| "loss": 3.1663, |
| "step": 75450 |
| }, |
| { |
| "epoch": 8.126143579808417, |
| "grad_norm": 0.8456742763519287, |
| "learning_rate": 0.00011275293610602305, |
| "loss": 3.1589, |
| "step": 75500 |
| }, |
| { |
| "epoch": 8.131525131848026, |
| "grad_norm": 0.8249267935752869, |
| "learning_rate": 0.00011242969507596164, |
| "loss": 3.1508, |
| "step": 75550 |
| }, |
| { |
| "epoch": 8.136906683887632, |
| "grad_norm": 0.8262121081352234, |
| "learning_rate": 0.00011210645404590022, |
| "loss": 3.1733, |
| "step": 75600 |
| }, |
| { |
| "epoch": 8.142288235927241, |
| "grad_norm": 0.8334835171699524, |
| "learning_rate": 0.00011178321301583881, |
| "loss": 3.1677, |
| "step": 75650 |
| }, |
| { |
| "epoch": 8.14766978796685, |
| "grad_norm": 0.8308705687522888, |
| "learning_rate": 0.00011145997198577738, |
| "loss": 3.1778, |
| "step": 75700 |
| }, |
| { |
| "epoch": 8.153051340006458, |
| "grad_norm": 0.8552451729774475, |
| "learning_rate": 0.00011113673095571597, |
| "loss": 3.1685, |
| "step": 75750 |
| }, |
| { |
| "epoch": 8.158432892046067, |
| "grad_norm": 0.8798450231552124, |
| "learning_rate": 0.00011081348992565455, |
| "loss": 3.1637, |
| "step": 75800 |
| }, |
| { |
| "epoch": 8.163814444085673, |
| "grad_norm": 0.839418888092041, |
| "learning_rate": 0.00011049024889559314, |
| "loss": 3.165, |
| "step": 75850 |
| }, |
| { |
| "epoch": 8.169195996125282, |
| "grad_norm": 0.8186075091362, |
| "learning_rate": 0.00011016700786553172, |
| "loss": 3.1782, |
| "step": 75900 |
| }, |
| { |
| "epoch": 8.17457754816489, |
| "grad_norm": 0.836857795715332, |
| "learning_rate": 0.00010984376683547031, |
| "loss": 3.1717, |
| "step": 75950 |
| }, |
| { |
| "epoch": 8.1799591002045, |
| "grad_norm": 0.78592449426651, |
| "learning_rate": 0.00010952052580540889, |
| "loss": 3.1849, |
| "step": 76000 |
| }, |
| { |
| "epoch": 8.1799591002045, |
| "eval_accuracy": 0.39025553094809096, |
| "eval_loss": 3.335026979446411, |
| "eval_runtime": 179.5655, |
| "eval_samples_per_second": 100.303, |
| "eval_steps_per_second": 6.271, |
| "step": 76000 |
| }, |
| { |
| "epoch": 8.185340652244108, |
| "grad_norm": 0.8572103381156921, |
| "learning_rate": 0.00010919728477534747, |
| "loss": 3.18, |
| "step": 76050 |
| }, |
| { |
| "epoch": 8.190722204283716, |
| "grad_norm": 0.8487206101417542, |
| "learning_rate": 0.0001088805085658873, |
| "loss": 3.165, |
| "step": 76100 |
| }, |
| { |
| "epoch": 8.196103756323323, |
| "grad_norm": 0.8511477112770081, |
| "learning_rate": 0.00010855726753582586, |
| "loss": 3.1883, |
| "step": 76150 |
| }, |
| { |
| "epoch": 8.201485308362932, |
| "grad_norm": 0.8475496172904968, |
| "learning_rate": 0.00010823402650576446, |
| "loss": 3.1702, |
| "step": 76200 |
| }, |
| { |
| "epoch": 8.20686686040254, |
| "grad_norm": 0.8378956317901611, |
| "learning_rate": 0.00010791078547570304, |
| "loss": 3.1585, |
| "step": 76250 |
| }, |
| { |
| "epoch": 8.212248412442149, |
| "grad_norm": 0.8379483222961426, |
| "learning_rate": 0.00010758754444564162, |
| "loss": 3.1581, |
| "step": 76300 |
| }, |
| { |
| "epoch": 8.217629964481757, |
| "grad_norm": 0.8493615388870239, |
| "learning_rate": 0.0001072643034155802, |
| "loss": 3.1709, |
| "step": 76350 |
| }, |
| { |
| "epoch": 8.223011516521364, |
| "grad_norm": 0.8749557137489319, |
| "learning_rate": 0.0001069410623855188, |
| "loss": 3.1636, |
| "step": 76400 |
| }, |
| { |
| "epoch": 8.228393068560973, |
| "grad_norm": 0.8558868765830994, |
| "learning_rate": 0.00010661782135545739, |
| "loss": 3.1669, |
| "step": 76450 |
| }, |
| { |
| "epoch": 8.233774620600581, |
| "grad_norm": 0.813459575176239, |
| "learning_rate": 0.00010629458032539596, |
| "loss": 3.1802, |
| "step": 76500 |
| }, |
| { |
| "epoch": 8.23915617264019, |
| "grad_norm": 0.8157198429107666, |
| "learning_rate": 0.00010597133929533455, |
| "loss": 3.1721, |
| "step": 76550 |
| }, |
| { |
| "epoch": 8.244537724679798, |
| "grad_norm": 0.8524532318115234, |
| "learning_rate": 0.00010564809826527313, |
| "loss": 3.1772, |
| "step": 76600 |
| }, |
| { |
| "epoch": 8.249919276719407, |
| "grad_norm": 0.813799262046814, |
| "learning_rate": 0.00010532485723521171, |
| "loss": 3.1585, |
| "step": 76650 |
| }, |
| { |
| "epoch": 8.255300828759013, |
| "grad_norm": 0.8357850909233093, |
| "learning_rate": 0.0001050016162051503, |
| "loss": 3.1791, |
| "step": 76700 |
| }, |
| { |
| "epoch": 8.260682380798622, |
| "grad_norm": 0.8634746670722961, |
| "learning_rate": 0.00010467837517508889, |
| "loss": 3.1795, |
| "step": 76750 |
| }, |
| { |
| "epoch": 8.26606393283823, |
| "grad_norm": 0.8267971277236938, |
| "learning_rate": 0.00010435513414502746, |
| "loss": 3.1719, |
| "step": 76800 |
| }, |
| { |
| "epoch": 8.27144548487784, |
| "grad_norm": 0.8415201306343079, |
| "learning_rate": 0.00010403189311496605, |
| "loss": 3.19, |
| "step": 76850 |
| }, |
| { |
| "epoch": 8.276827036917448, |
| "grad_norm": 0.8072708249092102, |
| "learning_rate": 0.00010370865208490463, |
| "loss": 3.1761, |
| "step": 76900 |
| }, |
| { |
| "epoch": 8.282208588957054, |
| "grad_norm": 0.8550540208816528, |
| "learning_rate": 0.00010338541105484323, |
| "loss": 3.1753, |
| "step": 76950 |
| }, |
| { |
| "epoch": 8.287590140996663, |
| "grad_norm": 0.8370460271835327, |
| "learning_rate": 0.0001030621700247818, |
| "loss": 3.1775, |
| "step": 77000 |
| }, |
| { |
| "epoch": 8.287590140996663, |
| "eval_accuracy": 0.3904935894252925, |
| "eval_loss": 3.3347671031951904, |
| "eval_runtime": 179.9264, |
| "eval_samples_per_second": 100.102, |
| "eval_steps_per_second": 6.258, |
| "step": 77000 |
| }, |
| { |
| "epoch": 8.292971693036272, |
| "grad_norm": 0.8297492265701294, |
| "learning_rate": 0.00010273892899472039, |
| "loss": 3.1716, |
| "step": 77050 |
| }, |
| { |
| "epoch": 8.29835324507588, |
| "grad_norm": 0.8193721175193787, |
| "learning_rate": 0.00010241568796465898, |
| "loss": 3.1875, |
| "step": 77100 |
| }, |
| { |
| "epoch": 8.303734797115489, |
| "grad_norm": 0.8413053154945374, |
| "learning_rate": 0.00010209244693459755, |
| "loss": 3.1671, |
| "step": 77150 |
| }, |
| { |
| "epoch": 8.309116349155097, |
| "grad_norm": 0.8616971969604492, |
| "learning_rate": 0.00010176920590453613, |
| "loss": 3.1864, |
| "step": 77200 |
| }, |
| { |
| "epoch": 8.314497901194704, |
| "grad_norm": 0.8818926811218262, |
| "learning_rate": 0.00010144596487447473, |
| "loss": 3.175, |
| "step": 77250 |
| }, |
| { |
| "epoch": 8.319879453234313, |
| "grad_norm": 0.8248533606529236, |
| "learning_rate": 0.00010112272384441332, |
| "loss": 3.1903, |
| "step": 77300 |
| }, |
| { |
| "epoch": 8.325261005273921, |
| "grad_norm": 0.8652777075767517, |
| "learning_rate": 0.00010079948281435189, |
| "loss": 3.1636, |
| "step": 77350 |
| }, |
| { |
| "epoch": 8.33064255731353, |
| "grad_norm": 0.819579541683197, |
| "learning_rate": 0.00010047624178429048, |
| "loss": 3.1682, |
| "step": 77400 |
| }, |
| { |
| "epoch": 8.336024109353138, |
| "grad_norm": 0.795547366142273, |
| "learning_rate": 0.00010015300075422906, |
| "loss": 3.1727, |
| "step": 77450 |
| }, |
| { |
| "epoch": 8.341405661392745, |
| "grad_norm": 0.8604868650436401, |
| "learning_rate": 9.982975972416765e-05, |
| "loss": 3.1878, |
| "step": 77500 |
| }, |
| { |
| "epoch": 8.346787213432354, |
| "grad_norm": 0.8407930135726929, |
| "learning_rate": 9.950651869410623e-05, |
| "loss": 3.1736, |
| "step": 77550 |
| }, |
| { |
| "epoch": 8.352168765471962, |
| "grad_norm": 0.8258522748947144, |
| "learning_rate": 9.918327766404482e-05, |
| "loss": 3.1694, |
| "step": 77600 |
| }, |
| { |
| "epoch": 8.35755031751157, |
| "grad_norm": 0.8680301308631897, |
| "learning_rate": 9.886003663398339e-05, |
| "loss": 3.1864, |
| "step": 77650 |
| }, |
| { |
| "epoch": 8.36293186955118, |
| "grad_norm": 0.8190805912017822, |
| "learning_rate": 9.853679560392198e-05, |
| "loss": 3.1861, |
| "step": 77700 |
| }, |
| { |
| "epoch": 8.368313421590786, |
| "grad_norm": 0.8155598044395447, |
| "learning_rate": 9.821355457386056e-05, |
| "loss": 3.1661, |
| "step": 77750 |
| }, |
| { |
| "epoch": 8.373694973630395, |
| "grad_norm": 0.850091814994812, |
| "learning_rate": 9.789031354379916e-05, |
| "loss": 3.1893, |
| "step": 77800 |
| }, |
| { |
| "epoch": 8.379076525670003, |
| "grad_norm": 0.8463276028633118, |
| "learning_rate": 9.756707251373773e-05, |
| "loss": 3.1879, |
| "step": 77850 |
| }, |
| { |
| "epoch": 8.384458077709612, |
| "grad_norm": 0.8920957446098328, |
| "learning_rate": 9.724383148367632e-05, |
| "loss": 3.1779, |
| "step": 77900 |
| }, |
| { |
| "epoch": 8.38983962974922, |
| "grad_norm": 0.8446083068847656, |
| "learning_rate": 9.692059045361492e-05, |
| "loss": 3.1744, |
| "step": 77950 |
| }, |
| { |
| "epoch": 8.395221181788829, |
| "grad_norm": 0.864314079284668, |
| "learning_rate": 9.659734942355348e-05, |
| "loss": 3.1833, |
| "step": 78000 |
| }, |
| { |
| "epoch": 8.395221181788829, |
| "eval_accuracy": 0.3908043366844419, |
| "eval_loss": 3.33040189743042, |
| "eval_runtime": 179.9825, |
| "eval_samples_per_second": 100.071, |
| "eval_steps_per_second": 6.256, |
| "step": 78000 |
| }, |
| { |
| "epoch": 8.400602733828435, |
| "grad_norm": 0.8862897157669067, |
| "learning_rate": 9.627410839349208e-05, |
| "loss": 3.1755, |
| "step": 78050 |
| }, |
| { |
| "epoch": 8.405984285868044, |
| "grad_norm": 0.819072961807251, |
| "learning_rate": 9.595086736343066e-05, |
| "loss": 3.176, |
| "step": 78100 |
| }, |
| { |
| "epoch": 8.411365837907653, |
| "grad_norm": 0.8221203088760376, |
| "learning_rate": 9.562762633336923e-05, |
| "loss": 3.1816, |
| "step": 78150 |
| }, |
| { |
| "epoch": 8.416747389947261, |
| "grad_norm": 0.8711651563644409, |
| "learning_rate": 9.530438530330782e-05, |
| "loss": 3.1838, |
| "step": 78200 |
| }, |
| { |
| "epoch": 8.42212894198687, |
| "grad_norm": 0.8410494923591614, |
| "learning_rate": 9.498114427324642e-05, |
| "loss": 3.1768, |
| "step": 78250 |
| }, |
| { |
| "epoch": 8.427510494026476, |
| "grad_norm": 0.8514099717140198, |
| "learning_rate": 9.4657903243185e-05, |
| "loss": 3.166, |
| "step": 78300 |
| }, |
| { |
| "epoch": 8.432892046066085, |
| "grad_norm": 0.8536844253540039, |
| "learning_rate": 9.433466221312358e-05, |
| "loss": 3.172, |
| "step": 78350 |
| }, |
| { |
| "epoch": 8.438273598105694, |
| "grad_norm": 0.8886139988899231, |
| "learning_rate": 9.401142118306216e-05, |
| "loss": 3.1935, |
| "step": 78400 |
| }, |
| { |
| "epoch": 8.443655150145302, |
| "grad_norm": 0.8296248912811279, |
| "learning_rate": 9.368818015300075e-05, |
| "loss": 3.1878, |
| "step": 78450 |
| }, |
| { |
| "epoch": 8.44903670218491, |
| "grad_norm": 0.88595050573349, |
| "learning_rate": 9.336493912293932e-05, |
| "loss": 3.1633, |
| "step": 78500 |
| }, |
| { |
| "epoch": 8.45441825422452, |
| "grad_norm": 0.8764588832855225, |
| "learning_rate": 9.304169809287792e-05, |
| "loss": 3.1798, |
| "step": 78550 |
| }, |
| { |
| "epoch": 8.459799806264126, |
| "grad_norm": 0.830272912979126, |
| "learning_rate": 9.27184570628165e-05, |
| "loss": 3.1857, |
| "step": 78600 |
| }, |
| { |
| "epoch": 8.465181358303735, |
| "grad_norm": 0.8409690856933594, |
| "learning_rate": 9.239521603275508e-05, |
| "loss": 3.1858, |
| "step": 78650 |
| }, |
| { |
| "epoch": 8.470562910343343, |
| "grad_norm": 0.8955550789833069, |
| "learning_rate": 9.207197500269366e-05, |
| "loss": 3.1948, |
| "step": 78700 |
| }, |
| { |
| "epoch": 8.475944462382952, |
| "grad_norm": 0.8606041073799133, |
| "learning_rate": 9.174873397263225e-05, |
| "loss": 3.2018, |
| "step": 78750 |
| }, |
| { |
| "epoch": 8.48132601442256, |
| "grad_norm": 0.8329320549964905, |
| "learning_rate": 9.142549294257085e-05, |
| "loss": 3.1797, |
| "step": 78800 |
| }, |
| { |
| "epoch": 8.486707566462167, |
| "grad_norm": 0.8341664671897888, |
| "learning_rate": 9.110225191250942e-05, |
| "loss": 3.1785, |
| "step": 78850 |
| }, |
| { |
| "epoch": 8.492089118501776, |
| "grad_norm": 0.8955467939376831, |
| "learning_rate": 9.077901088244801e-05, |
| "loss": 3.1911, |
| "step": 78900 |
| }, |
| { |
| "epoch": 8.497470670541384, |
| "grad_norm": 0.8430368304252625, |
| "learning_rate": 9.045576985238659e-05, |
| "loss": 3.1755, |
| "step": 78950 |
| }, |
| { |
| "epoch": 8.502852222580993, |
| "grad_norm": 0.8482922315597534, |
| "learning_rate": 9.013252882232517e-05, |
| "loss": 3.183, |
| "step": 79000 |
| }, |
| { |
| "epoch": 8.502852222580993, |
| "eval_accuracy": 0.39098600431286773, |
| "eval_loss": 3.328646183013916, |
| "eval_runtime": 180.2849, |
| "eval_samples_per_second": 99.903, |
| "eval_steps_per_second": 6.246, |
| "step": 79000 |
| }, |
| { |
| "epoch": 8.508233774620601, |
| "grad_norm": 0.860719621181488, |
| "learning_rate": 8.980928779226375e-05, |
| "loss": 3.2104, |
| "step": 79050 |
| }, |
| { |
| "epoch": 8.513615326660208, |
| "grad_norm": 0.836053729057312, |
| "learning_rate": 8.948604676220235e-05, |
| "loss": 3.1859, |
| "step": 79100 |
| }, |
| { |
| "epoch": 8.518996878699816, |
| "grad_norm": 0.8130183815956116, |
| "learning_rate": 8.916280573214092e-05, |
| "loss": 3.1714, |
| "step": 79150 |
| }, |
| { |
| "epoch": 8.524378430739425, |
| "grad_norm": 0.927423357963562, |
| "learning_rate": 8.883956470207951e-05, |
| "loss": 3.2046, |
| "step": 79200 |
| }, |
| { |
| "epoch": 8.529759982779034, |
| "grad_norm": 0.8596153855323792, |
| "learning_rate": 8.851632367201809e-05, |
| "loss": 3.183, |
| "step": 79250 |
| }, |
| { |
| "epoch": 8.535141534818642, |
| "grad_norm": 0.8415372371673584, |
| "learning_rate": 8.819308264195669e-05, |
| "loss": 3.1731, |
| "step": 79300 |
| }, |
| { |
| "epoch": 8.54052308685825, |
| "grad_norm": 0.8734657168388367, |
| "learning_rate": 8.786984161189526e-05, |
| "loss": 3.1889, |
| "step": 79350 |
| }, |
| { |
| "epoch": 8.545904638897857, |
| "grad_norm": 0.838965654373169, |
| "learning_rate": 8.754660058183385e-05, |
| "loss": 3.1763, |
| "step": 79400 |
| }, |
| { |
| "epoch": 8.551286190937466, |
| "grad_norm": 0.8917363286018372, |
| "learning_rate": 8.722335955177243e-05, |
| "loss": 3.1875, |
| "step": 79450 |
| }, |
| { |
| "epoch": 8.556667742977075, |
| "grad_norm": 0.8633471727371216, |
| "learning_rate": 8.690011852171101e-05, |
| "loss": 3.1781, |
| "step": 79500 |
| }, |
| { |
| "epoch": 8.562049295016683, |
| "grad_norm": 0.8390756845474243, |
| "learning_rate": 8.657687749164959e-05, |
| "loss": 3.1836, |
| "step": 79550 |
| }, |
| { |
| "epoch": 8.567430847056292, |
| "grad_norm": 0.8166258931159973, |
| "learning_rate": 8.625363646158819e-05, |
| "loss": 3.1666, |
| "step": 79600 |
| }, |
| { |
| "epoch": 8.572812399095898, |
| "grad_norm": 0.857069194316864, |
| "learning_rate": 8.593039543152676e-05, |
| "loss": 3.1912, |
| "step": 79650 |
| }, |
| { |
| "epoch": 8.578193951135507, |
| "grad_norm": 0.8506762385368347, |
| "learning_rate": 8.560715440146535e-05, |
| "loss": 3.1866, |
| "step": 79700 |
| }, |
| { |
| "epoch": 8.583575503175116, |
| "grad_norm": 0.8568733334541321, |
| "learning_rate": 8.528391337140394e-05, |
| "loss": 3.1747, |
| "step": 79750 |
| }, |
| { |
| "epoch": 8.588957055214724, |
| "grad_norm": 0.8337421417236328, |
| "learning_rate": 8.496067234134253e-05, |
| "loss": 3.1741, |
| "step": 79800 |
| }, |
| { |
| "epoch": 8.594338607254333, |
| "grad_norm": 0.8349370360374451, |
| "learning_rate": 8.46374313112811e-05, |
| "loss": 3.1783, |
| "step": 79850 |
| }, |
| { |
| "epoch": 8.599720159293941, |
| "grad_norm": 0.9050174355506897, |
| "learning_rate": 8.431419028121969e-05, |
| "loss": 3.1898, |
| "step": 79900 |
| }, |
| { |
| "epoch": 8.605101711333548, |
| "grad_norm": 0.8692119717597961, |
| "learning_rate": 8.399094925115828e-05, |
| "loss": 3.1776, |
| "step": 79950 |
| }, |
| { |
| "epoch": 8.610483263373157, |
| "grad_norm": 0.8102073669433594, |
| "learning_rate": 8.366770822109685e-05, |
| "loss": 3.1628, |
| "step": 80000 |
| }, |
| { |
| "epoch": 8.610483263373157, |
| "eval_accuracy": 0.39150384397619153, |
| "eval_loss": 3.32488751411438, |
| "eval_runtime": 179.9701, |
| "eval_samples_per_second": 100.078, |
| "eval_steps_per_second": 6.257, |
| "step": 80000 |
| }, |
| { |
| "epoch": 8.615864815412765, |
| "grad_norm": 0.8268591165542603, |
| "learning_rate": 8.334446719103544e-05, |
| "loss": 3.1842, |
| "step": 80050 |
| }, |
| { |
| "epoch": 8.621246367452374, |
| "grad_norm": 0.877962052822113, |
| "learning_rate": 8.302122616097403e-05, |
| "loss": 3.2023, |
| "step": 80100 |
| }, |
| { |
| "epoch": 8.626627919491982, |
| "grad_norm": 0.8843813538551331, |
| "learning_rate": 8.269798513091261e-05, |
| "loss": 3.1682, |
| "step": 80150 |
| }, |
| { |
| "epoch": 8.632009471531589, |
| "grad_norm": 0.8113887906074524, |
| "learning_rate": 8.237474410085119e-05, |
| "loss": 3.1886, |
| "step": 80200 |
| }, |
| { |
| "epoch": 8.637391023571197, |
| "grad_norm": 0.8340319991111755, |
| "learning_rate": 8.205150307078978e-05, |
| "loss": 3.2039, |
| "step": 80250 |
| }, |
| { |
| "epoch": 8.642772575610806, |
| "grad_norm": 0.8976853489875793, |
| "learning_rate": 8.172826204072838e-05, |
| "loss": 3.2052, |
| "step": 80300 |
| }, |
| { |
| "epoch": 8.648154127650415, |
| "grad_norm": 0.8994436264038086, |
| "learning_rate": 8.140502101066694e-05, |
| "loss": 3.1815, |
| "step": 80350 |
| }, |
| { |
| "epoch": 8.653535679690023, |
| "grad_norm": 0.8481301665306091, |
| "learning_rate": 8.108177998060553e-05, |
| "loss": 3.1876, |
| "step": 80400 |
| }, |
| { |
| "epoch": 8.658917231729632, |
| "grad_norm": 0.8830190896987915, |
| "learning_rate": 8.075853895054412e-05, |
| "loss": 3.1799, |
| "step": 80450 |
| }, |
| { |
| "epoch": 8.664298783769238, |
| "grad_norm": 0.8312429785728455, |
| "learning_rate": 8.043529792048269e-05, |
| "loss": 3.1841, |
| "step": 80500 |
| }, |
| { |
| "epoch": 8.669680335808847, |
| "grad_norm": 0.8653261661529541, |
| "learning_rate": 8.011205689042128e-05, |
| "loss": 3.1789, |
| "step": 80550 |
| }, |
| { |
| "epoch": 8.675061887848456, |
| "grad_norm": 0.8662141561508179, |
| "learning_rate": 7.978881586035988e-05, |
| "loss": 3.171, |
| "step": 80600 |
| }, |
| { |
| "epoch": 8.680443439888064, |
| "grad_norm": 0.9127115607261658, |
| "learning_rate": 7.946557483029845e-05, |
| "loss": 3.1847, |
| "step": 80650 |
| }, |
| { |
| "epoch": 8.685824991927673, |
| "grad_norm": 0.8367604613304138, |
| "learning_rate": 7.914233380023704e-05, |
| "loss": 3.1722, |
| "step": 80700 |
| }, |
| { |
| "epoch": 8.69120654396728, |
| "grad_norm": 0.8278328776359558, |
| "learning_rate": 7.881909277017562e-05, |
| "loss": 3.1743, |
| "step": 80750 |
| }, |
| { |
| "epoch": 8.696588096006888, |
| "grad_norm": 0.8759754300117493, |
| "learning_rate": 7.849585174011422e-05, |
| "loss": 3.1851, |
| "step": 80800 |
| }, |
| { |
| "epoch": 8.701969648046497, |
| "grad_norm": 0.8897731304168701, |
| "learning_rate": 7.817261071005278e-05, |
| "loss": 3.1884, |
| "step": 80850 |
| }, |
| { |
| "epoch": 8.707351200086105, |
| "grad_norm": 0.846344530582428, |
| "learning_rate": 7.784936967999138e-05, |
| "loss": 3.2009, |
| "step": 80900 |
| }, |
| { |
| "epoch": 8.712732752125714, |
| "grad_norm": 0.832398533821106, |
| "learning_rate": 7.752612864992996e-05, |
| "loss": 3.188, |
| "step": 80950 |
| }, |
| { |
| "epoch": 8.718114304165322, |
| "grad_norm": 0.8839730024337769, |
| "learning_rate": 7.720288761986854e-05, |
| "loss": 3.1654, |
| "step": 81000 |
| }, |
| { |
| "epoch": 8.718114304165322, |
| "eval_accuracy": 0.3917791703939134, |
| "eval_loss": 3.3209238052368164, |
| "eval_runtime": 179.8859, |
| "eval_samples_per_second": 100.125, |
| "eval_steps_per_second": 6.26, |
| "step": 81000 |
| }, |
| { |
| "epoch": 8.723495856204929, |
| "grad_norm": 0.8395397067070007, |
| "learning_rate": 7.687964658980712e-05, |
| "loss": 3.1755, |
| "step": 81050 |
| }, |
| { |
| "epoch": 8.728877408244538, |
| "grad_norm": 0.8440877199172974, |
| "learning_rate": 7.655640555974572e-05, |
| "loss": 3.2005, |
| "step": 81100 |
| }, |
| { |
| "epoch": 8.734258960284146, |
| "grad_norm": 0.8579280376434326, |
| "learning_rate": 7.623316452968428e-05, |
| "loss": 3.1813, |
| "step": 81150 |
| }, |
| { |
| "epoch": 8.739640512323755, |
| "grad_norm": 0.8807401657104492, |
| "learning_rate": 7.590992349962288e-05, |
| "loss": 3.1917, |
| "step": 81200 |
| }, |
| { |
| "epoch": 8.745022064363363, |
| "grad_norm": 0.8908937573432922, |
| "learning_rate": 7.558668246956147e-05, |
| "loss": 3.1825, |
| "step": 81250 |
| }, |
| { |
| "epoch": 8.75040361640297, |
| "grad_norm": 0.9015775322914124, |
| "learning_rate": 7.526990626010127e-05, |
| "loss": 3.1864, |
| "step": 81300 |
| }, |
| { |
| "epoch": 8.755785168442578, |
| "grad_norm": 0.823549211025238, |
| "learning_rate": 7.494666523003986e-05, |
| "loss": 3.1703, |
| "step": 81350 |
| }, |
| { |
| "epoch": 8.761166720482187, |
| "grad_norm": 0.9204948544502258, |
| "learning_rate": 7.462342419997844e-05, |
| "loss": 3.1872, |
| "step": 81400 |
| }, |
| { |
| "epoch": 8.766548272521796, |
| "grad_norm": 0.8299639225006104, |
| "learning_rate": 7.430018316991704e-05, |
| "loss": 3.2054, |
| "step": 81450 |
| }, |
| { |
| "epoch": 8.771929824561404, |
| "grad_norm": 0.8888871669769287, |
| "learning_rate": 7.397694213985562e-05, |
| "loss": 3.1921, |
| "step": 81500 |
| }, |
| { |
| "epoch": 8.777311376601011, |
| "grad_norm": 0.8606129288673401, |
| "learning_rate": 7.36537011097942e-05, |
| "loss": 3.1933, |
| "step": 81550 |
| }, |
| { |
| "epoch": 8.78269292864062, |
| "grad_norm": 0.8673796653747559, |
| "learning_rate": 7.333046007973278e-05, |
| "loss": 3.1882, |
| "step": 81600 |
| }, |
| { |
| "epoch": 8.788074480680228, |
| "grad_norm": 0.8238900899887085, |
| "learning_rate": 7.300721904967136e-05, |
| "loss": 3.1728, |
| "step": 81650 |
| }, |
| { |
| "epoch": 8.793456032719837, |
| "grad_norm": 0.8531633615493774, |
| "learning_rate": 7.268397801960996e-05, |
| "loss": 3.1705, |
| "step": 81700 |
| }, |
| { |
| "epoch": 8.798837584759445, |
| "grad_norm": 0.84257972240448, |
| "learning_rate": 7.236073698954854e-05, |
| "loss": 3.1883, |
| "step": 81750 |
| }, |
| { |
| "epoch": 8.804219136799054, |
| "grad_norm": 0.9469799399375916, |
| "learning_rate": 7.203749595948712e-05, |
| "loss": 3.1883, |
| "step": 81800 |
| }, |
| { |
| "epoch": 8.80960068883866, |
| "grad_norm": 0.9426635503768921, |
| "learning_rate": 7.17142549294257e-05, |
| "loss": 3.1985, |
| "step": 81850 |
| }, |
| { |
| "epoch": 8.814982240878269, |
| "grad_norm": 0.8460460305213928, |
| "learning_rate": 7.139101389936428e-05, |
| "loss": 3.1817, |
| "step": 81900 |
| }, |
| { |
| "epoch": 8.820363792917878, |
| "grad_norm": 0.8010488748550415, |
| "learning_rate": 7.106777286930287e-05, |
| "loss": 3.1964, |
| "step": 81950 |
| }, |
| { |
| "epoch": 8.825745344957486, |
| "grad_norm": 0.9217504262924194, |
| "learning_rate": 7.074453183924146e-05, |
| "loss": 3.1983, |
| "step": 82000 |
| }, |
| { |
| "epoch": 8.825745344957486, |
| "eval_accuracy": 0.3922471383817304, |
| "eval_loss": 3.3171520233154297, |
| "eval_runtime": 179.8992, |
| "eval_samples_per_second": 100.117, |
| "eval_steps_per_second": 6.259, |
| "step": 82000 |
| }, |
| { |
| "epoch": 8.831126896997095, |
| "grad_norm": 0.8669100403785706, |
| "learning_rate": 7.042129080918004e-05, |
| "loss": 3.1705, |
| "step": 82050 |
| }, |
| { |
| "epoch": 8.836508449036701, |
| "grad_norm": 0.9114564657211304, |
| "learning_rate": 7.009804977911862e-05, |
| "loss": 3.1815, |
| "step": 82100 |
| }, |
| { |
| "epoch": 8.84189000107631, |
| "grad_norm": 0.8626256585121155, |
| "learning_rate": 6.97748087490572e-05, |
| "loss": 3.175, |
| "step": 82150 |
| }, |
| { |
| "epoch": 8.847271553115919, |
| "grad_norm": 0.842830240726471, |
| "learning_rate": 6.945156771899579e-05, |
| "loss": 3.1945, |
| "step": 82200 |
| }, |
| { |
| "epoch": 8.852653105155527, |
| "grad_norm": 0.8612439036369324, |
| "learning_rate": 6.912832668893437e-05, |
| "loss": 3.1973, |
| "step": 82250 |
| }, |
| { |
| "epoch": 8.858034657195136, |
| "grad_norm": 0.8703390955924988, |
| "learning_rate": 6.880508565887297e-05, |
| "loss": 3.1825, |
| "step": 82300 |
| }, |
| { |
| "epoch": 8.863416209234742, |
| "grad_norm": 0.9062275290489197, |
| "learning_rate": 6.848184462881155e-05, |
| "loss": 3.1814, |
| "step": 82350 |
| }, |
| { |
| "epoch": 8.868797761274351, |
| "grad_norm": 0.895692765712738, |
| "learning_rate": 6.815860359875013e-05, |
| "loss": 3.1771, |
| "step": 82400 |
| }, |
| { |
| "epoch": 8.87417931331396, |
| "grad_norm": 0.8932182788848877, |
| "learning_rate": 6.783536256868871e-05, |
| "loss": 3.1825, |
| "step": 82450 |
| }, |
| { |
| "epoch": 8.879560865353568, |
| "grad_norm": 0.8720111846923828, |
| "learning_rate": 6.75121215386273e-05, |
| "loss": 3.1972, |
| "step": 82500 |
| }, |
| { |
| "epoch": 8.884942417393177, |
| "grad_norm": 0.8333724737167358, |
| "learning_rate": 6.718888050856589e-05, |
| "loss": 3.199, |
| "step": 82550 |
| }, |
| { |
| "epoch": 8.890323969432785, |
| "grad_norm": 0.8801641464233398, |
| "learning_rate": 6.686563947850447e-05, |
| "loss": 3.1722, |
| "step": 82600 |
| }, |
| { |
| "epoch": 8.895705521472392, |
| "grad_norm": 0.8496854305267334, |
| "learning_rate": 6.654239844844305e-05, |
| "loss": 3.1837, |
| "step": 82650 |
| }, |
| { |
| "epoch": 8.901087073512, |
| "grad_norm": 0.888050377368927, |
| "learning_rate": 6.621915741838163e-05, |
| "loss": 3.1976, |
| "step": 82700 |
| }, |
| { |
| "epoch": 8.906468625551609, |
| "grad_norm": 0.8081518411636353, |
| "learning_rate": 6.589591638832021e-05, |
| "loss": 3.188, |
| "step": 82750 |
| }, |
| { |
| "epoch": 8.911850177591218, |
| "grad_norm": 0.8638512492179871, |
| "learning_rate": 6.557267535825881e-05, |
| "loss": 3.1785, |
| "step": 82800 |
| }, |
| { |
| "epoch": 8.917231729630826, |
| "grad_norm": 0.8664565682411194, |
| "learning_rate": 6.524943432819739e-05, |
| "loss": 3.1917, |
| "step": 82850 |
| }, |
| { |
| "epoch": 8.922613281670433, |
| "grad_norm": 0.8508527874946594, |
| "learning_rate": 6.492619329813597e-05, |
| "loss": 3.1835, |
| "step": 82900 |
| }, |
| { |
| "epoch": 8.927994833710041, |
| "grad_norm": 0.8651978373527527, |
| "learning_rate": 6.460295226807455e-05, |
| "loss": 3.1777, |
| "step": 82950 |
| }, |
| { |
| "epoch": 8.93337638574965, |
| "grad_norm": 0.8856733441352844, |
| "learning_rate": 6.427971123801313e-05, |
| "loss": 3.1908, |
| "step": 83000 |
| }, |
| { |
| "epoch": 8.93337638574965, |
| "eval_accuracy": 0.3923557912695449, |
| "eval_loss": 3.314030170440674, |
| "eval_runtime": 179.6808, |
| "eval_samples_per_second": 100.239, |
| "eval_steps_per_second": 6.267, |
| "step": 83000 |
| }, |
| { |
| "epoch": 8.938757937789259, |
| "grad_norm": 0.8681096434593201, |
| "learning_rate": 6.395647020795173e-05, |
| "loss": 3.1862, |
| "step": 83050 |
| }, |
| { |
| "epoch": 8.944139489828867, |
| "grad_norm": 0.8370834589004517, |
| "learning_rate": 6.363322917789031e-05, |
| "loss": 3.178, |
| "step": 83100 |
| }, |
| { |
| "epoch": 8.949521041868476, |
| "grad_norm": 0.8382720947265625, |
| "learning_rate": 6.330998814782889e-05, |
| "loss": 3.1773, |
| "step": 83150 |
| }, |
| { |
| "epoch": 8.954902593908082, |
| "grad_norm": 0.918395459651947, |
| "learning_rate": 6.298674711776748e-05, |
| "loss": 3.1927, |
| "step": 83200 |
| }, |
| { |
| "epoch": 8.960284145947691, |
| "grad_norm": 0.8420078754425049, |
| "learning_rate": 6.266350608770606e-05, |
| "loss": 3.1879, |
| "step": 83250 |
| }, |
| { |
| "epoch": 8.9656656979873, |
| "grad_norm": 0.8565245866775513, |
| "learning_rate": 6.234026505764465e-05, |
| "loss": 3.1709, |
| "step": 83300 |
| }, |
| { |
| "epoch": 8.971047250026908, |
| "grad_norm": 0.8625103235244751, |
| "learning_rate": 6.201702402758323e-05, |
| "loss": 3.1964, |
| "step": 83350 |
| }, |
| { |
| "epoch": 8.976428802066517, |
| "grad_norm": 0.8943954706192017, |
| "learning_rate": 6.169378299752181e-05, |
| "loss": 3.2013, |
| "step": 83400 |
| }, |
| { |
| "epoch": 8.981810354106123, |
| "grad_norm": 0.9224780201911926, |
| "learning_rate": 6.13705419674604e-05, |
| "loss": 3.1733, |
| "step": 83450 |
| }, |
| { |
| "epoch": 8.987191906145732, |
| "grad_norm": 0.8462356925010681, |
| "learning_rate": 6.104730093739898e-05, |
| "loss": 3.1706, |
| "step": 83500 |
| }, |
| { |
| "epoch": 8.99257345818534, |
| "grad_norm": 0.8774244785308838, |
| "learning_rate": 6.072405990733757e-05, |
| "loss": 3.1717, |
| "step": 83550 |
| }, |
| { |
| "epoch": 8.997955010224949, |
| "grad_norm": 0.8515488505363464, |
| "learning_rate": 6.040081887727615e-05, |
| "loss": 3.1694, |
| "step": 83600 |
| }, |
| { |
| "epoch": 9.003336562264558, |
| "grad_norm": 0.8341154456138611, |
| "learning_rate": 6.007757784721473e-05, |
| "loss": 3.1465, |
| "step": 83650 |
| }, |
| { |
| "epoch": 9.008718114304166, |
| "grad_norm": 0.8942939043045044, |
| "learning_rate": 5.975433681715332e-05, |
| "loss": 3.1279, |
| "step": 83700 |
| }, |
| { |
| "epoch": 9.014099666343773, |
| "grad_norm": 0.8541099429130554, |
| "learning_rate": 5.94310957870919e-05, |
| "loss": 3.0992, |
| "step": 83750 |
| }, |
| { |
| "epoch": 9.019481218383381, |
| "grad_norm": 0.8721352219581604, |
| "learning_rate": 5.910785475703049e-05, |
| "loss": 3.1244, |
| "step": 83800 |
| }, |
| { |
| "epoch": 9.02486277042299, |
| "grad_norm": 0.877285897731781, |
| "learning_rate": 5.878461372696907e-05, |
| "loss": 3.1187, |
| "step": 83850 |
| }, |
| { |
| "epoch": 9.030244322462599, |
| "grad_norm": 0.9021511673927307, |
| "learning_rate": 5.846137269690765e-05, |
| "loss": 3.1431, |
| "step": 83900 |
| }, |
| { |
| "epoch": 9.035625874502207, |
| "grad_norm": 0.9125701785087585, |
| "learning_rate": 5.813813166684624e-05, |
| "loss": 3.1139, |
| "step": 83950 |
| }, |
| { |
| "epoch": 9.041007426541814, |
| "grad_norm": 0.8349764347076416, |
| "learning_rate": 5.781489063678482e-05, |
| "loss": 3.126, |
| "step": 84000 |
| }, |
| { |
| "epoch": 9.041007426541814, |
| "eval_accuracy": 0.3922756054383378, |
| "eval_loss": 3.317336320877075, |
| "eval_runtime": 180.0088, |
| "eval_samples_per_second": 100.056, |
| "eval_steps_per_second": 6.255, |
| "step": 84000 |
| }, |
| { |
| "epoch": 9.046388978581422, |
| "grad_norm": 0.9072763323783875, |
| "learning_rate": 5.749164960672341e-05, |
| "loss": 3.1405, |
| "step": 84050 |
| }, |
| { |
| "epoch": 9.051770530621031, |
| "grad_norm": 0.9203513264656067, |
| "learning_rate": 5.716840857666199e-05, |
| "loss": 3.1331, |
| "step": 84100 |
| }, |
| { |
| "epoch": 9.05715208266064, |
| "grad_norm": 0.8729878664016724, |
| "learning_rate": 5.684516754660057e-05, |
| "loss": 3.1335, |
| "step": 84150 |
| }, |
| { |
| "epoch": 9.062533634700248, |
| "grad_norm": 0.9056902527809143, |
| "learning_rate": 5.652192651653916e-05, |
| "loss": 3.1256, |
| "step": 84200 |
| }, |
| { |
| "epoch": 9.067915186739857, |
| "grad_norm": 0.8653237223625183, |
| "learning_rate": 5.619868548647774e-05, |
| "loss": 3.1253, |
| "step": 84250 |
| }, |
| { |
| "epoch": 9.073296738779463, |
| "grad_norm": 0.8897874355316162, |
| "learning_rate": 5.5875444456416335e-05, |
| "loss": 3.1121, |
| "step": 84300 |
| }, |
| { |
| "epoch": 9.078678290819072, |
| "grad_norm": 0.873859167098999, |
| "learning_rate": 5.5552203426354916e-05, |
| "loss": 3.1191, |
| "step": 84350 |
| }, |
| { |
| "epoch": 9.08405984285868, |
| "grad_norm": 0.8796614408493042, |
| "learning_rate": 5.52289623962935e-05, |
| "loss": 3.1252, |
| "step": 84400 |
| }, |
| { |
| "epoch": 9.089441394898289, |
| "grad_norm": 0.9063519835472107, |
| "learning_rate": 5.4905721366232085e-05, |
| "loss": 3.1188, |
| "step": 84450 |
| }, |
| { |
| "epoch": 9.094822946937898, |
| "grad_norm": 0.8301732540130615, |
| "learning_rate": 5.4582480336170666e-05, |
| "loss": 3.117, |
| "step": 84500 |
| }, |
| { |
| "epoch": 9.100204498977504, |
| "grad_norm": 0.8899036049842834, |
| "learning_rate": 5.4259239306109254e-05, |
| "loss": 3.1182, |
| "step": 84550 |
| }, |
| { |
| "epoch": 9.105586051017113, |
| "grad_norm": 0.876559317111969, |
| "learning_rate": 5.3935998276047835e-05, |
| "loss": 3.1158, |
| "step": 84600 |
| }, |
| { |
| "epoch": 9.110967603056721, |
| "grad_norm": 0.8589257597923279, |
| "learning_rate": 5.3612757245986416e-05, |
| "loss": 3.1303, |
| "step": 84650 |
| }, |
| { |
| "epoch": 9.11634915509633, |
| "grad_norm": 0.9055981636047363, |
| "learning_rate": 5.3289516215925004e-05, |
| "loss": 3.113, |
| "step": 84700 |
| }, |
| { |
| "epoch": 9.121730707135939, |
| "grad_norm": 0.898881196975708, |
| "learning_rate": 5.2966275185863585e-05, |
| "loss": 3.1289, |
| "step": 84750 |
| }, |
| { |
| "epoch": 9.127112259175545, |
| "grad_norm": 0.8705615401268005, |
| "learning_rate": 5.264303415580217e-05, |
| "loss": 3.1179, |
| "step": 84800 |
| }, |
| { |
| "epoch": 9.132493811215154, |
| "grad_norm": 0.8401497006416321, |
| "learning_rate": 5.2319793125740754e-05, |
| "loss": 3.1123, |
| "step": 84850 |
| }, |
| { |
| "epoch": 9.137875363254762, |
| "grad_norm": 0.8974107503890991, |
| "learning_rate": 5.1996552095679336e-05, |
| "loss": 3.1348, |
| "step": 84900 |
| }, |
| { |
| "epoch": 9.143256915294371, |
| "grad_norm": 0.8535527586936951, |
| "learning_rate": 5.1673311065617923e-05, |
| "loss": 3.1261, |
| "step": 84950 |
| }, |
| { |
| "epoch": 9.14863846733398, |
| "grad_norm": 0.8549533486366272, |
| "learning_rate": 5.135653485615774e-05, |
| "loss": 3.1465, |
| "step": 85000 |
| }, |
| { |
| "epoch": 9.14863846733398, |
| "eval_accuracy": 0.39245216638103636, |
| "eval_loss": 3.317115545272827, |
| "eval_runtime": 179.8863, |
| "eval_samples_per_second": 100.124, |
| "eval_steps_per_second": 6.26, |
| "step": 85000 |
| }, |
| { |
| "epoch": 9.154020019373588, |
| "grad_norm": 0.8901803493499756, |
| "learning_rate": 5.103329382609632e-05, |
| "loss": 3.1413, |
| "step": 85050 |
| }, |
| { |
| "epoch": 9.159401571413195, |
| "grad_norm": 0.8346094489097595, |
| "learning_rate": 5.0710052796034906e-05, |
| "loss": 3.1247, |
| "step": 85100 |
| }, |
| { |
| "epoch": 9.164783123452803, |
| "grad_norm": 0.8812398910522461, |
| "learning_rate": 5.038681176597349e-05, |
| "loss": 3.137, |
| "step": 85150 |
| }, |
| { |
| "epoch": 9.170164675492412, |
| "grad_norm": 0.8596518635749817, |
| "learning_rate": 5.006357073591207e-05, |
| "loss": 3.1351, |
| "step": 85200 |
| }, |
| { |
| "epoch": 9.17554622753202, |
| "grad_norm": 0.8846608400344849, |
| "learning_rate": 4.974032970585066e-05, |
| "loss": 3.1143, |
| "step": 85250 |
| }, |
| { |
| "epoch": 9.180927779571629, |
| "grad_norm": 0.8773012161254883, |
| "learning_rate": 4.9417088675789244e-05, |
| "loss": 3.1483, |
| "step": 85300 |
| }, |
| { |
| "epoch": 9.186309331611236, |
| "grad_norm": 0.855322003364563, |
| "learning_rate": 4.909384764572783e-05, |
| "loss": 3.1371, |
| "step": 85350 |
| }, |
| { |
| "epoch": 9.191690883650844, |
| "grad_norm": 0.8355333209037781, |
| "learning_rate": 4.877060661566641e-05, |
| "loss": 3.1179, |
| "step": 85400 |
| }, |
| { |
| "epoch": 9.197072435690453, |
| "grad_norm": 0.8519471883773804, |
| "learning_rate": 4.8447365585604994e-05, |
| "loss": 3.1286, |
| "step": 85450 |
| }, |
| { |
| "epoch": 9.202453987730062, |
| "grad_norm": 0.8353000283241272, |
| "learning_rate": 4.812412455554358e-05, |
| "loss": 3.118, |
| "step": 85500 |
| }, |
| { |
| "epoch": 9.20783553976967, |
| "grad_norm": 0.8422732353210449, |
| "learning_rate": 4.780088352548216e-05, |
| "loss": 3.1536, |
| "step": 85550 |
| }, |
| { |
| "epoch": 9.213217091809279, |
| "grad_norm": 0.8970634937286377, |
| "learning_rate": 4.747764249542075e-05, |
| "loss": 3.1394, |
| "step": 85600 |
| }, |
| { |
| "epoch": 9.218598643848885, |
| "grad_norm": 0.8779487013816833, |
| "learning_rate": 4.715440146535933e-05, |
| "loss": 3.1236, |
| "step": 85650 |
| }, |
| { |
| "epoch": 9.223980195888494, |
| "grad_norm": 0.887233316898346, |
| "learning_rate": 4.683116043529791e-05, |
| "loss": 3.1297, |
| "step": 85700 |
| }, |
| { |
| "epoch": 9.229361747928102, |
| "grad_norm": 0.8759692907333374, |
| "learning_rate": 4.65079194052365e-05, |
| "loss": 3.1114, |
| "step": 85750 |
| }, |
| { |
| "epoch": 9.234743299967711, |
| "grad_norm": 0.8891520500183105, |
| "learning_rate": 4.618467837517508e-05, |
| "loss": 3.1221, |
| "step": 85800 |
| }, |
| { |
| "epoch": 9.24012485200732, |
| "grad_norm": 0.867538571357727, |
| "learning_rate": 4.586143734511367e-05, |
| "loss": 3.1237, |
| "step": 85850 |
| }, |
| { |
| "epoch": 9.245506404046926, |
| "grad_norm": 0.8855520486831665, |
| "learning_rate": 4.553819631505225e-05, |
| "loss": 3.1306, |
| "step": 85900 |
| }, |
| { |
| "epoch": 9.250887956086535, |
| "grad_norm": 0.8678146600723267, |
| "learning_rate": 4.521495528499083e-05, |
| "loss": 3.1267, |
| "step": 85950 |
| }, |
| { |
| "epoch": 9.256269508126143, |
| "grad_norm": 0.8817727565765381, |
| "learning_rate": 4.489171425492942e-05, |
| "loss": 3.1352, |
| "step": 86000 |
| }, |
| { |
| "epoch": 9.256269508126143, |
| "eval_accuracy": 0.39275030990519927, |
| "eval_loss": 3.3147568702697754, |
| "eval_runtime": 179.4526, |
| "eval_samples_per_second": 100.366, |
| "eval_steps_per_second": 6.275, |
| "step": 86000 |
| }, |
| { |
| "epoch": 9.261651060165752, |
| "grad_norm": 0.8901002407073975, |
| "learning_rate": 4.4568473224868e-05, |
| "loss": 3.1272, |
| "step": 86050 |
| }, |
| { |
| "epoch": 9.26703261220536, |
| "grad_norm": 0.882002592086792, |
| "learning_rate": 4.4245232194806596e-05, |
| "loss": 3.1305, |
| "step": 86100 |
| }, |
| { |
| "epoch": 9.272414164244967, |
| "grad_norm": 0.8344738483428955, |
| "learning_rate": 4.392199116474518e-05, |
| "loss": 3.1374, |
| "step": 86150 |
| }, |
| { |
| "epoch": 9.277795716284576, |
| "grad_norm": 0.8456497192382812, |
| "learning_rate": 4.359875013468376e-05, |
| "loss": 3.1337, |
| "step": 86200 |
| }, |
| { |
| "epoch": 9.283177268324184, |
| "grad_norm": 0.8751317262649536, |
| "learning_rate": 4.3275509104622346e-05, |
| "loss": 3.1329, |
| "step": 86250 |
| }, |
| { |
| "epoch": 9.288558820363793, |
| "grad_norm": 0.8995939493179321, |
| "learning_rate": 4.295226807456093e-05, |
| "loss": 3.1315, |
| "step": 86300 |
| }, |
| { |
| "epoch": 9.293940372403402, |
| "grad_norm": 0.8733546137809753, |
| "learning_rate": 4.2629027044499515e-05, |
| "loss": 3.1177, |
| "step": 86350 |
| }, |
| { |
| "epoch": 9.29932192444301, |
| "grad_norm": 0.837860643863678, |
| "learning_rate": 4.2305786014438096e-05, |
| "loss": 3.1346, |
| "step": 86400 |
| }, |
| { |
| "epoch": 9.304703476482617, |
| "grad_norm": 0.8571630120277405, |
| "learning_rate": 4.198254498437668e-05, |
| "loss": 3.1312, |
| "step": 86450 |
| }, |
| { |
| "epoch": 9.310085028522225, |
| "grad_norm": 0.8765683174133301, |
| "learning_rate": 4.1659303954315265e-05, |
| "loss": 3.1339, |
| "step": 86500 |
| }, |
| { |
| "epoch": 9.315466580561834, |
| "grad_norm": 0.8633087873458862, |
| "learning_rate": 4.1336062924253846e-05, |
| "loss": 3.146, |
| "step": 86550 |
| }, |
| { |
| "epoch": 9.320848132601443, |
| "grad_norm": 0.8428789377212524, |
| "learning_rate": 4.1012821894192434e-05, |
| "loss": 3.1403, |
| "step": 86600 |
| }, |
| { |
| "epoch": 9.326229684641051, |
| "grad_norm": 0.815321683883667, |
| "learning_rate": 4.0689580864131015e-05, |
| "loss": 3.1163, |
| "step": 86650 |
| }, |
| { |
| "epoch": 9.331611236680658, |
| "grad_norm": 0.8932182788848877, |
| "learning_rate": 4.0366339834069596e-05, |
| "loss": 3.142, |
| "step": 86700 |
| }, |
| { |
| "epoch": 9.336992788720266, |
| "grad_norm": 0.8550369143486023, |
| "learning_rate": 4.0043098804008184e-05, |
| "loss": 3.1437, |
| "step": 86750 |
| }, |
| { |
| "epoch": 9.342374340759875, |
| "grad_norm": 0.8777996301651001, |
| "learning_rate": 3.9719857773946765e-05, |
| "loss": 3.1413, |
| "step": 86800 |
| }, |
| { |
| "epoch": 9.347755892799483, |
| "grad_norm": 0.8686240315437317, |
| "learning_rate": 3.939661674388536e-05, |
| "loss": 3.1265, |
| "step": 86850 |
| }, |
| { |
| "epoch": 9.353137444839092, |
| "grad_norm": 0.8639189600944519, |
| "learning_rate": 3.9073375713823934e-05, |
| "loss": 3.1314, |
| "step": 86900 |
| }, |
| { |
| "epoch": 9.3585189968787, |
| "grad_norm": 0.8951987028121948, |
| "learning_rate": 3.8750134683762515e-05, |
| "loss": 3.1329, |
| "step": 86950 |
| }, |
| { |
| "epoch": 9.363900548918307, |
| "grad_norm": 0.8524070382118225, |
| "learning_rate": 3.842689365370111e-05, |
| "loss": 3.129, |
| "step": 87000 |
| }, |
| { |
| "epoch": 9.363900548918307, |
| "eval_accuracy": 0.39313265941741843, |
| "eval_loss": 3.3121142387390137, |
| "eval_runtime": 179.955, |
| "eval_samples_per_second": 100.086, |
| "eval_steps_per_second": 6.257, |
| "step": 87000 |
| }, |
| { |
| "epoch": 9.369282100957916, |
| "grad_norm": 0.8834142088890076, |
| "learning_rate": 3.811011744424092e-05, |
| "loss": 3.1444, |
| "step": 87050 |
| }, |
| { |
| "epoch": 9.374663652997524, |
| "grad_norm": 0.8924848437309265, |
| "learning_rate": 3.7786876414179504e-05, |
| "loss": 3.1386, |
| "step": 87100 |
| }, |
| { |
| "epoch": 9.380045205037133, |
| "grad_norm": 0.8910598158836365, |
| "learning_rate": 3.7463635384118085e-05, |
| "loss": 3.1267, |
| "step": 87150 |
| }, |
| { |
| "epoch": 9.385426757076742, |
| "grad_norm": 0.9576981067657471, |
| "learning_rate": 3.714039435405667e-05, |
| "loss": 3.1285, |
| "step": 87200 |
| }, |
| { |
| "epoch": 9.390808309116348, |
| "grad_norm": 0.8550928235054016, |
| "learning_rate": 3.6817153323995254e-05, |
| "loss": 3.1348, |
| "step": 87250 |
| }, |
| { |
| "epoch": 9.396189861155957, |
| "grad_norm": 0.8894127607345581, |
| "learning_rate": 3.649391229393384e-05, |
| "loss": 3.1176, |
| "step": 87300 |
| }, |
| { |
| "epoch": 9.401571413195565, |
| "grad_norm": 0.8873329758644104, |
| "learning_rate": 3.617067126387242e-05, |
| "loss": 3.1366, |
| "step": 87350 |
| }, |
| { |
| "epoch": 9.406952965235174, |
| "grad_norm": 0.916664719581604, |
| "learning_rate": 3.5847430233811004e-05, |
| "loss": 3.1222, |
| "step": 87400 |
| }, |
| { |
| "epoch": 9.412334517274783, |
| "grad_norm": 0.8714802265167236, |
| "learning_rate": 3.552418920374959e-05, |
| "loss": 3.1415, |
| "step": 87450 |
| }, |
| { |
| "epoch": 9.417716069314391, |
| "grad_norm": 0.8508152961730957, |
| "learning_rate": 3.520094817368818e-05, |
| "loss": 3.134, |
| "step": 87500 |
| }, |
| { |
| "epoch": 9.423097621353998, |
| "grad_norm": 0.8734935522079468, |
| "learning_rate": 3.487770714362676e-05, |
| "loss": 3.1225, |
| "step": 87550 |
| }, |
| { |
| "epoch": 9.428479173393606, |
| "grad_norm": 0.8918778300285339, |
| "learning_rate": 3.455446611356535e-05, |
| "loss": 3.1418, |
| "step": 87600 |
| }, |
| { |
| "epoch": 9.433860725433215, |
| "grad_norm": 0.848052442073822, |
| "learning_rate": 3.423122508350393e-05, |
| "loss": 3.1324, |
| "step": 87650 |
| }, |
| { |
| "epoch": 9.439242277472824, |
| "grad_norm": 0.9284157156944275, |
| "learning_rate": 3.390798405344251e-05, |
| "loss": 3.1299, |
| "step": 87700 |
| }, |
| { |
| "epoch": 9.444623829512432, |
| "grad_norm": 0.8494699597358704, |
| "learning_rate": 3.35847430233811e-05, |
| "loss": 3.1307, |
| "step": 87750 |
| }, |
| { |
| "epoch": 9.450005381552039, |
| "grad_norm": 0.8794162273406982, |
| "learning_rate": 3.326150199331968e-05, |
| "loss": 3.1361, |
| "step": 87800 |
| }, |
| { |
| "epoch": 9.455386933591647, |
| "grad_norm": 0.8918963670730591, |
| "learning_rate": 3.293826096325827e-05, |
| "loss": 3.1246, |
| "step": 87850 |
| }, |
| { |
| "epoch": 9.460768485631256, |
| "grad_norm": 0.8786701560020447, |
| "learning_rate": 3.261501993319685e-05, |
| "loss": 3.1148, |
| "step": 87900 |
| }, |
| { |
| "epoch": 9.466150037670864, |
| "grad_norm": 0.8807184100151062, |
| "learning_rate": 3.229177890313544e-05, |
| "loss": 3.135, |
| "step": 87950 |
| }, |
| { |
| "epoch": 9.471531589710473, |
| "grad_norm": 0.8800942301750183, |
| "learning_rate": 3.196853787307402e-05, |
| "loss": 3.1379, |
| "step": 88000 |
| }, |
| { |
| "epoch": 9.471531589710473, |
| "eval_accuracy": 0.39338103991896234, |
| "eval_loss": 3.3090343475341797, |
| "eval_runtime": 179.8302, |
| "eval_samples_per_second": 100.156, |
| "eval_steps_per_second": 6.261, |
| "step": 88000 |
| }, |
| { |
| "epoch": 9.476913141750082, |
| "grad_norm": 0.9230080246925354, |
| "learning_rate": 3.1645296843012606e-05, |
| "loss": 3.1339, |
| "step": 88050 |
| }, |
| { |
| "epoch": 9.482294693789688, |
| "grad_norm": 0.8794991970062256, |
| "learning_rate": 3.132205581295119e-05, |
| "loss": 3.128, |
| "step": 88100 |
| }, |
| { |
| "epoch": 9.487676245829297, |
| "grad_norm": 0.8440155386924744, |
| "learning_rate": 3.099881478288977e-05, |
| "loss": 3.1349, |
| "step": 88150 |
| }, |
| { |
| "epoch": 9.493057797868905, |
| "grad_norm": 0.8837097883224487, |
| "learning_rate": 3.0675573752828356e-05, |
| "loss": 3.1294, |
| "step": 88200 |
| }, |
| { |
| "epoch": 9.498439349908514, |
| "grad_norm": 0.8591768741607666, |
| "learning_rate": 3.035233272276694e-05, |
| "loss": 3.1204, |
| "step": 88250 |
| }, |
| { |
| "epoch": 9.503820901948123, |
| "grad_norm": 0.9089881181716919, |
| "learning_rate": 3.0029091692705525e-05, |
| "loss": 3.1381, |
| "step": 88300 |
| }, |
| { |
| "epoch": 9.50920245398773, |
| "grad_norm": 0.8778650760650635, |
| "learning_rate": 2.970585066264411e-05, |
| "loss": 3.1344, |
| "step": 88350 |
| }, |
| { |
| "epoch": 9.514584006027338, |
| "grad_norm": 0.8694146871566772, |
| "learning_rate": 2.938260963258269e-05, |
| "loss": 3.1233, |
| "step": 88400 |
| }, |
| { |
| "epoch": 9.519965558066946, |
| "grad_norm": 0.8793097734451294, |
| "learning_rate": 2.9059368602521275e-05, |
| "loss": 3.1229, |
| "step": 88450 |
| }, |
| { |
| "epoch": 9.525347110106555, |
| "grad_norm": 0.896824300289154, |
| "learning_rate": 2.8736127572459863e-05, |
| "loss": 3.1412, |
| "step": 88500 |
| }, |
| { |
| "epoch": 9.530728662146164, |
| "grad_norm": 0.8718636631965637, |
| "learning_rate": 2.8412886542398448e-05, |
| "loss": 3.1335, |
| "step": 88550 |
| }, |
| { |
| "epoch": 9.536110214185772, |
| "grad_norm": 0.8667758703231812, |
| "learning_rate": 2.8089645512337032e-05, |
| "loss": 3.13, |
| "step": 88600 |
| }, |
| { |
| "epoch": 9.541491766225379, |
| "grad_norm": 0.8904845118522644, |
| "learning_rate": 2.7766404482275613e-05, |
| "loss": 3.1138, |
| "step": 88650 |
| }, |
| { |
| "epoch": 9.546873318264987, |
| "grad_norm": 0.9374863505363464, |
| "learning_rate": 2.7443163452214198e-05, |
| "loss": 3.1137, |
| "step": 88700 |
| }, |
| { |
| "epoch": 9.552254870304596, |
| "grad_norm": 0.9184094071388245, |
| "learning_rate": 2.7119922422152782e-05, |
| "loss": 3.148, |
| "step": 88750 |
| }, |
| { |
| "epoch": 9.557636422344205, |
| "grad_norm": 0.8584219217300415, |
| "learning_rate": 2.6796681392091367e-05, |
| "loss": 3.1272, |
| "step": 88800 |
| }, |
| { |
| "epoch": 9.563017974383813, |
| "grad_norm": 0.8915684819221497, |
| "learning_rate": 2.6473440362029955e-05, |
| "loss": 3.1465, |
| "step": 88850 |
| }, |
| { |
| "epoch": 9.56839952642342, |
| "grad_norm": 0.9296562671661377, |
| "learning_rate": 2.6150199331968536e-05, |
| "loss": 3.1323, |
| "step": 88900 |
| }, |
| { |
| "epoch": 9.573781078463028, |
| "grad_norm": 0.888725757598877, |
| "learning_rate": 2.582695830190712e-05, |
| "loss": 3.134, |
| "step": 88950 |
| }, |
| { |
| "epoch": 9.579162630502637, |
| "grad_norm": 0.9244557619094849, |
| "learning_rate": 2.5503717271845705e-05, |
| "loss": 3.1373, |
| "step": 89000 |
| }, |
| { |
| "epoch": 9.579162630502637, |
| "eval_accuracy": 0.393602039892777, |
| "eval_loss": 3.307893753051758, |
| "eval_runtime": 180.0866, |
| "eval_samples_per_second": 100.013, |
| "eval_steps_per_second": 6.253, |
| "step": 89000 |
| }, |
| { |
| "epoch": 9.584544182542245, |
| "grad_norm": 0.8654752373695374, |
| "learning_rate": 2.518047624178429e-05, |
| "loss": 3.1271, |
| "step": 89050 |
| }, |
| { |
| "epoch": 9.589925734581854, |
| "grad_norm": 0.8975775837898254, |
| "learning_rate": 2.4857235211722874e-05, |
| "loss": 3.1469, |
| "step": 89100 |
| }, |
| { |
| "epoch": 9.59530728662146, |
| "grad_norm": 0.8581560254096985, |
| "learning_rate": 2.4533994181661455e-05, |
| "loss": 3.1602, |
| "step": 89150 |
| }, |
| { |
| "epoch": 9.60068883866107, |
| "grad_norm": 0.8734593391418457, |
| "learning_rate": 2.421075315160004e-05, |
| "loss": 3.121, |
| "step": 89200 |
| }, |
| { |
| "epoch": 9.606070390700678, |
| "grad_norm": 0.8599057197570801, |
| "learning_rate": 2.3887512121538624e-05, |
| "loss": 3.1476, |
| "step": 89250 |
| }, |
| { |
| "epoch": 9.611451942740286, |
| "grad_norm": 0.8804906606674194, |
| "learning_rate": 2.3564271091477212e-05, |
| "loss": 3.1037, |
| "step": 89300 |
| }, |
| { |
| "epoch": 9.616833494779895, |
| "grad_norm": 0.8924012184143066, |
| "learning_rate": 2.3241030061415796e-05, |
| "loss": 3.1147, |
| "step": 89350 |
| }, |
| { |
| "epoch": 9.622215046819504, |
| "grad_norm": 0.8790477514266968, |
| "learning_rate": 2.2917789031354377e-05, |
| "loss": 3.1474, |
| "step": 89400 |
| }, |
| { |
| "epoch": 9.62759659885911, |
| "grad_norm": 0.8938375115394592, |
| "learning_rate": 2.2594548001292962e-05, |
| "loss": 3.1409, |
| "step": 89450 |
| }, |
| { |
| "epoch": 9.632978150898719, |
| "grad_norm": 0.849459171295166, |
| "learning_rate": 2.2271306971231546e-05, |
| "loss": 3.1288, |
| "step": 89500 |
| }, |
| { |
| "epoch": 9.638359702938327, |
| "grad_norm": 0.8806934952735901, |
| "learning_rate": 2.194806594117013e-05, |
| "loss": 3.128, |
| "step": 89550 |
| }, |
| { |
| "epoch": 9.643741254977936, |
| "grad_norm": 0.8296778202056885, |
| "learning_rate": 2.1624824911108715e-05, |
| "loss": 3.1499, |
| "step": 89600 |
| }, |
| { |
| "epoch": 9.649122807017545, |
| "grad_norm": 0.8698211908340454, |
| "learning_rate": 2.1301583881047296e-05, |
| "loss": 3.131, |
| "step": 89650 |
| }, |
| { |
| "epoch": 9.654504359057151, |
| "grad_norm": 0.9121602177619934, |
| "learning_rate": 2.097834285098588e-05, |
| "loss": 3.1383, |
| "step": 89700 |
| }, |
| { |
| "epoch": 9.65988591109676, |
| "grad_norm": 0.8963391780853271, |
| "learning_rate": 2.065510182092447e-05, |
| "loss": 3.1249, |
| "step": 89750 |
| }, |
| { |
| "epoch": 9.665267463136368, |
| "grad_norm": 0.9141267538070679, |
| "learning_rate": 2.0331860790863053e-05, |
| "loss": 3.1462, |
| "step": 89800 |
| }, |
| { |
| "epoch": 9.670649015175977, |
| "grad_norm": 0.869062602519989, |
| "learning_rate": 2.0008619760801638e-05, |
| "loss": 3.1401, |
| "step": 89850 |
| }, |
| { |
| "epoch": 9.676030567215586, |
| "grad_norm": 0.897432804107666, |
| "learning_rate": 1.968537873074022e-05, |
| "loss": 3.1304, |
| "step": 89900 |
| }, |
| { |
| "epoch": 9.681412119255192, |
| "grad_norm": 0.8654459118843079, |
| "learning_rate": 1.9362137700678803e-05, |
| "loss": 3.1216, |
| "step": 89950 |
| }, |
| { |
| "epoch": 9.6867936712948, |
| "grad_norm": 0.9050065875053406, |
| "learning_rate": 1.9038896670617388e-05, |
| "loss": 3.1356, |
| "step": 90000 |
| }, |
| { |
| "epoch": 9.6867936712948, |
| "eval_accuracy": 0.39384498774993015, |
| "eval_loss": 3.3056321144104004, |
| "eval_runtime": 179.7547, |
| "eval_samples_per_second": 100.198, |
| "eval_steps_per_second": 6.264, |
| "step": 90000 |
| }, |
| { |
| "epoch": 9.69217522333441, |
| "grad_norm": 0.9278258681297302, |
| "learning_rate": 1.8715655640555972e-05, |
| "loss": 3.15, |
| "step": 90050 |
| }, |
| { |
| "epoch": 9.697556775374018, |
| "grad_norm": 0.8931335806846619, |
| "learning_rate": 1.8392414610494557e-05, |
| "loss": 3.1421, |
| "step": 90100 |
| }, |
| { |
| "epoch": 9.702938327413626, |
| "grad_norm": 0.8307539820671082, |
| "learning_rate": 1.806917358043314e-05, |
| "loss": 3.1121, |
| "step": 90150 |
| }, |
| { |
| "epoch": 9.708319879453235, |
| "grad_norm": 0.8717964887619019, |
| "learning_rate": 1.7745932550371726e-05, |
| "loss": 3.1036, |
| "step": 90200 |
| }, |
| { |
| "epoch": 9.713701431492842, |
| "grad_norm": 0.8405609726905823, |
| "learning_rate": 1.742269152031031e-05, |
| "loss": 3.1154, |
| "step": 90250 |
| }, |
| { |
| "epoch": 9.71908298353245, |
| "grad_norm": 0.8634082078933716, |
| "learning_rate": 1.7099450490248895e-05, |
| "loss": 3.1389, |
| "step": 90300 |
| }, |
| { |
| "epoch": 9.724464535572059, |
| "grad_norm": 0.905585527420044, |
| "learning_rate": 1.677620946018748e-05, |
| "loss": 3.1538, |
| "step": 90350 |
| }, |
| { |
| "epoch": 9.729846087611667, |
| "grad_norm": 0.8903700709342957, |
| "learning_rate": 1.6452968430126064e-05, |
| "loss": 3.1403, |
| "step": 90400 |
| }, |
| { |
| "epoch": 9.735227639651276, |
| "grad_norm": 0.8525581359863281, |
| "learning_rate": 1.6129727400064645e-05, |
| "loss": 3.1337, |
| "step": 90450 |
| }, |
| { |
| "epoch": 9.740609191690883, |
| "grad_norm": 0.8256161212921143, |
| "learning_rate": 1.580648637000323e-05, |
| "loss": 3.1202, |
| "step": 90500 |
| }, |
| { |
| "epoch": 9.745990743730491, |
| "grad_norm": 0.8859636783599854, |
| "learning_rate": 1.5483245339941817e-05, |
| "loss": 3.1281, |
| "step": 90550 |
| }, |
| { |
| "epoch": 9.7513722957701, |
| "grad_norm": 0.879533052444458, |
| "learning_rate": 1.5160004309880398e-05, |
| "loss": 3.1193, |
| "step": 90600 |
| }, |
| { |
| "epoch": 9.756753847809708, |
| "grad_norm": 0.8815825581550598, |
| "learning_rate": 1.4836763279818985e-05, |
| "loss": 3.148, |
| "step": 90650 |
| }, |
| { |
| "epoch": 9.762135399849317, |
| "grad_norm": 0.8480821847915649, |
| "learning_rate": 1.4513522249757567e-05, |
| "loss": 3.1323, |
| "step": 90700 |
| }, |
| { |
| "epoch": 9.767516951888926, |
| "grad_norm": 0.9032825827598572, |
| "learning_rate": 1.4190281219696152e-05, |
| "loss": 3.1242, |
| "step": 90750 |
| }, |
| { |
| "epoch": 9.772898503928532, |
| "grad_norm": 0.8777044415473938, |
| "learning_rate": 1.3867040189634736e-05, |
| "loss": 3.1248, |
| "step": 90800 |
| }, |
| { |
| "epoch": 9.77828005596814, |
| "grad_norm": 0.8846954107284546, |
| "learning_rate": 1.3543799159573321e-05, |
| "loss": 3.1226, |
| "step": 90850 |
| }, |
| { |
| "epoch": 9.78366160800775, |
| "grad_norm": 0.9044235944747925, |
| "learning_rate": 1.3220558129511905e-05, |
| "loss": 3.1235, |
| "step": 90900 |
| }, |
| { |
| "epoch": 9.789043160047358, |
| "grad_norm": 0.8546289205551147, |
| "learning_rate": 1.2897317099450488e-05, |
| "loss": 3.1282, |
| "step": 90950 |
| }, |
| { |
| "epoch": 9.794424712086967, |
| "grad_norm": 0.8554105758666992, |
| "learning_rate": 1.2574076069389073e-05, |
| "loss": 3.1494, |
| "step": 91000 |
| }, |
| { |
| "epoch": 9.794424712086967, |
| "eval_accuracy": 0.39412846313423816, |
| "eval_loss": 3.3036842346191406, |
| "eval_runtime": 179.8656, |
| "eval_samples_per_second": 100.136, |
| "eval_steps_per_second": 6.26, |
| "step": 91000 |
| }, |
| { |
| "epoch": 9.799806264126573, |
| "grad_norm": 0.8401182293891907, |
| "learning_rate": 1.2257299859928886e-05, |
| "loss": 3.1188, |
| "step": 91050 |
| }, |
| { |
| "epoch": 9.805187816166182, |
| "grad_norm": 0.8869813084602356, |
| "learning_rate": 1.193405882986747e-05, |
| "loss": 3.1306, |
| "step": 91100 |
| }, |
| { |
| "epoch": 9.81056936820579, |
| "grad_norm": 0.9146585464477539, |
| "learning_rate": 1.1610817799806053e-05, |
| "loss": 3.1238, |
| "step": 91150 |
| }, |
| { |
| "epoch": 9.815950920245399, |
| "grad_norm": 0.8608843088150024, |
| "learning_rate": 1.128757676974464e-05, |
| "loss": 3.1311, |
| "step": 91200 |
| }, |
| { |
| "epoch": 9.821332472285007, |
| "grad_norm": 0.8796128630638123, |
| "learning_rate": 1.0964335739683222e-05, |
| "loss": 3.1301, |
| "step": 91250 |
| }, |
| { |
| "epoch": 9.826714024324616, |
| "grad_norm": 0.8590572476387024, |
| "learning_rate": 1.0641094709621807e-05, |
| "loss": 3.1268, |
| "step": 91300 |
| }, |
| { |
| "epoch": 9.832095576364223, |
| "grad_norm": 0.841092586517334, |
| "learning_rate": 1.0317853679560393e-05, |
| "loss": 3.1468, |
| "step": 91350 |
| }, |
| { |
| "epoch": 9.837477128403831, |
| "grad_norm": 0.879798412322998, |
| "learning_rate": 9.994612649498976e-06, |
| "loss": 3.124, |
| "step": 91400 |
| }, |
| { |
| "epoch": 9.84285868044344, |
| "grad_norm": 0.8939313292503357, |
| "learning_rate": 9.67137161943756e-06, |
| "loss": 3.1264, |
| "step": 91450 |
| }, |
| { |
| "epoch": 9.848240232483048, |
| "grad_norm": 0.8775787949562073, |
| "learning_rate": 9.348130589376145e-06, |
| "loss": 3.1319, |
| "step": 91500 |
| }, |
| { |
| "epoch": 9.853621784522657, |
| "grad_norm": 0.8580185174942017, |
| "learning_rate": 9.024889559314727e-06, |
| "loss": 3.1292, |
| "step": 91550 |
| }, |
| { |
| "epoch": 9.859003336562264, |
| "grad_norm": 0.9018335342407227, |
| "learning_rate": 8.701648529253312e-06, |
| "loss": 3.1408, |
| "step": 91600 |
| }, |
| { |
| "epoch": 9.864384888601872, |
| "grad_norm": 0.8868499398231506, |
| "learning_rate": 8.378407499191896e-06, |
| "loss": 3.1549, |
| "step": 91650 |
| }, |
| { |
| "epoch": 9.869766440641481, |
| "grad_norm": 0.8963901400566101, |
| "learning_rate": 8.055166469130481e-06, |
| "loss": 3.1332, |
| "step": 91700 |
| }, |
| { |
| "epoch": 9.87514799268109, |
| "grad_norm": 0.8725941777229309, |
| "learning_rate": 7.731925439069065e-06, |
| "loss": 3.1298, |
| "step": 91750 |
| }, |
| { |
| "epoch": 9.880529544720698, |
| "grad_norm": 0.8279412984848022, |
| "learning_rate": 7.408684409007649e-06, |
| "loss": 3.1264, |
| "step": 91800 |
| }, |
| { |
| "epoch": 9.885911096760307, |
| "grad_norm": 0.8716703057289124, |
| "learning_rate": 7.085443378946234e-06, |
| "loss": 3.121, |
| "step": 91850 |
| }, |
| { |
| "epoch": 9.891292648799913, |
| "grad_norm": 0.880731463432312, |
| "learning_rate": 6.762202348884817e-06, |
| "loss": 3.1255, |
| "step": 91900 |
| }, |
| { |
| "epoch": 9.896674200839522, |
| "grad_norm": 0.8716903328895569, |
| "learning_rate": 6.438961318823403e-06, |
| "loss": 3.1238, |
| "step": 91950 |
| }, |
| { |
| "epoch": 9.90205575287913, |
| "grad_norm": 0.8360830545425415, |
| "learning_rate": 6.115720288761986e-06, |
| "loss": 3.1415, |
| "step": 92000 |
| }, |
| { |
| "epoch": 9.90205575287913, |
| "eval_accuracy": 0.3942502630214782, |
| "eval_loss": 3.3022382259368896, |
| "eval_runtime": 179.9303, |
| "eval_samples_per_second": 100.1, |
| "eval_steps_per_second": 6.258, |
| "step": 92000 |
| }, |
| { |
| "epoch": 9.907437304918739, |
| "grad_norm": 0.8879337906837463, |
| "learning_rate": 5.792479258700571e-06, |
| "loss": 3.1039, |
| "step": 92050 |
| }, |
| { |
| "epoch": 9.912818856958348, |
| "grad_norm": 0.8420632481575012, |
| "learning_rate": 5.469238228639154e-06, |
| "loss": 3.1355, |
| "step": 92100 |
| }, |
| { |
| "epoch": 9.918200408997954, |
| "grad_norm": 0.8572197556495667, |
| "learning_rate": 5.145997198577739e-06, |
| "loss": 3.1128, |
| "step": 92150 |
| }, |
| { |
| "epoch": 9.923581961037563, |
| "grad_norm": 0.9185923337936401, |
| "learning_rate": 4.829220989117551e-06, |
| "loss": 3.1528, |
| "step": 92200 |
| }, |
| { |
| "epoch": 9.928963513077171, |
| "grad_norm": 0.8786590695381165, |
| "learning_rate": 4.505979959056136e-06, |
| "loss": 3.123, |
| "step": 92250 |
| }, |
| { |
| "epoch": 9.93434506511678, |
| "grad_norm": 0.8364728689193726, |
| "learning_rate": 4.18273892899472e-06, |
| "loss": 3.1189, |
| "step": 92300 |
| }, |
| { |
| "epoch": 9.939726617156388, |
| "grad_norm": 0.9233676791191101, |
| "learning_rate": 3.859497898933305e-06, |
| "loss": 3.1193, |
| "step": 92350 |
| }, |
| { |
| "epoch": 9.945108169195997, |
| "grad_norm": 0.874893844127655, |
| "learning_rate": 3.5362568688718884e-06, |
| "loss": 3.1365, |
| "step": 92400 |
| }, |
| { |
| "epoch": 9.950489721235604, |
| "grad_norm": 0.8843012452125549, |
| "learning_rate": 3.2130158388104725e-06, |
| "loss": 3.1061, |
| "step": 92450 |
| }, |
| { |
| "epoch": 9.955871273275212, |
| "grad_norm": 0.9005907773971558, |
| "learning_rate": 2.889774808749057e-06, |
| "loss": 3.1091, |
| "step": 92500 |
| }, |
| { |
| "epoch": 9.961252825314821, |
| "grad_norm": 0.881787896156311, |
| "learning_rate": 2.566533778687641e-06, |
| "loss": 3.1411, |
| "step": 92550 |
| }, |
| { |
| "epoch": 9.96663437735443, |
| "grad_norm": 0.8765003681182861, |
| "learning_rate": 2.243292748626225e-06, |
| "loss": 3.1421, |
| "step": 92600 |
| }, |
| { |
| "epoch": 9.972015929394038, |
| "grad_norm": 0.8894038200378418, |
| "learning_rate": 1.9200517185648096e-06, |
| "loss": 3.1143, |
| "step": 92650 |
| }, |
| { |
| "epoch": 9.977397481433645, |
| "grad_norm": 0.9114412069320679, |
| "learning_rate": 1.596810688503394e-06, |
| "loss": 3.1172, |
| "step": 92700 |
| }, |
| { |
| "epoch": 9.982779033473253, |
| "grad_norm": 0.8864458203315735, |
| "learning_rate": 1.2735696584419782e-06, |
| "loss": 3.1232, |
| "step": 92750 |
| }, |
| { |
| "epoch": 9.988160585512862, |
| "grad_norm": 0.8886809349060059, |
| "learning_rate": 9.503286283805623e-07, |
| "loss": 3.1238, |
| "step": 92800 |
| }, |
| { |
| "epoch": 9.99354213755247, |
| "grad_norm": 0.8665906190872192, |
| "learning_rate": 6.270875983191466e-07, |
| "loss": 3.1535, |
| "step": 92850 |
| }, |
| { |
| "epoch": 9.998923689592079, |
| "grad_norm": 0.8708318471908569, |
| "learning_rate": 3.038465682577308e-07, |
| "loss": 3.1364, |
| "step": 92900 |
| }, |
| { |
| "epoch": 10.0, |
| "step": 92910, |
| "total_flos": 7.7681598529536e+17, |
| "train_loss": 3.4556805386993372, |
| "train_runtime": 79114.6094, |
| "train_samples_per_second": 37.578, |
| "train_steps_per_second": 1.174 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 92910, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "save_steps": 10000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 7.7681598529536e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
