| { |
| "best_metric": 3.3066227436065674, |
| "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": 1.1552565097808838, |
| "learning_rate": 0.0003, |
| "loss": 8.6283, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.010763104079216447, |
| "grad_norm": 1.5108792781829834, |
| "learning_rate": 0.0006, |
| "loss": 6.9523, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.01614465611882467, |
| "grad_norm": 2.4516749382019043, |
| "learning_rate": 0.0005996767589699385, |
| "loss": 6.514, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.021526208158432893, |
| "grad_norm": 1.227023959159851, |
| "learning_rate": 0.0005993535179398771, |
| "loss": 6.2311, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.026907760198041114, |
| "grad_norm": 1.2729941606521606, |
| "learning_rate": 0.0005990302769098158, |
| "loss": 6.0782, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.03228931223764934, |
| "grad_norm": 2.2429041862487793, |
| "learning_rate": 0.0005987070358797543, |
| "loss": 5.9519, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.03767086427725756, |
| "grad_norm": 1.0578374862670898, |
| "learning_rate": 0.0005983837948496929, |
| "loss": 5.8611, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.04305241631686579, |
| "grad_norm": 1.2515922784805298, |
| "learning_rate": 0.0005980605538196314, |
| "loss": 5.7841, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.048433968356474004, |
| "grad_norm": 1.5836460590362549, |
| "learning_rate": 0.0005977373127895701, |
| "loss": 5.7439, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.05381552039608223, |
| "grad_norm": 1.2947415113449097, |
| "learning_rate": 0.0005974140717595086, |
| "loss": 5.6449, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.05919707243569045, |
| "grad_norm": 1.0450425148010254, |
| "learning_rate": 0.0005970908307294472, |
| "loss": 5.5658, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.06457862447529868, |
| "grad_norm": 1.5436805486679077, |
| "learning_rate": 0.0005967675896993858, |
| "loss": 5.5045, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.0699601765149069, |
| "grad_norm": 1.5273220539093018, |
| "learning_rate": 0.0005964443486693243, |
| "loss": 5.4187, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.07534172855451512, |
| "grad_norm": 0.809162974357605, |
| "learning_rate": 0.000596121107639263, |
| "loss": 5.3605, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.08072328059412334, |
| "grad_norm": 1.2222342491149902, |
| "learning_rate": 0.0005957978666092015, |
| "loss": 5.3092, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.08610483263373157, |
| "grad_norm": 1.3174521923065186, |
| "learning_rate": 0.0005954746255791401, |
| "loss": 5.273, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.09148638467333979, |
| "grad_norm": 1.2672854661941528, |
| "learning_rate": 0.0005951513845490787, |
| "loss": 5.2322, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.09686793671294801, |
| "grad_norm": 1.2101420164108276, |
| "learning_rate": 0.0005948281435190174, |
| "loss": 5.1579, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.10224948875255624, |
| "grad_norm": 1.4577302932739258, |
| "learning_rate": 0.0005945049024889559, |
| "loss": 5.1254, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.10763104079216446, |
| "grad_norm": 1.1115610599517822, |
| "learning_rate": 0.0005941816614588944, |
| "loss": 5.0941, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.10763104079216446, |
| "eval_accuracy": 0.22676085857946562, |
| "eval_loss": 5.024355888366699, |
| "eval_runtime": 186.3243, |
| "eval_samples_per_second": 96.665, |
| "eval_steps_per_second": 6.043, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.11301259283177269, |
| "grad_norm": 1.2621811628341675, |
| "learning_rate": 0.000593858420428833, |
| "loss": 5.0471, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.1183941448713809, |
| "grad_norm": 1.290496587753296, |
| "learning_rate": 0.0005935351793987716, |
| "loss": 5.0281, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.12377569691098914, |
| "grad_norm": 1.2553092241287231, |
| "learning_rate": 0.0005932119383687103, |
| "loss": 4.9935, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.12915724895059735, |
| "grad_norm": 1.1223114728927612, |
| "learning_rate": 0.0005928886973386488, |
| "loss": 4.9514, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.13453880099020557, |
| "grad_norm": 0.9260576963424683, |
| "learning_rate": 0.0005925654563085874, |
| "loss": 4.9153, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.1399203530298138, |
| "grad_norm": 1.0368200540542603, |
| "learning_rate": 0.000592242215278526, |
| "loss": 4.8948, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.14530190506942203, |
| "grad_norm": 1.1109979152679443, |
| "learning_rate": 0.0005919189742484645, |
| "loss": 4.8731, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.15068345710903025, |
| "grad_norm": 1.4146970510482788, |
| "learning_rate": 0.0005915957332184032, |
| "loss": 4.8403, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.15606500914863847, |
| "grad_norm": 1.0073134899139404, |
| "learning_rate": 0.0005912724921883417, |
| "loss": 4.8369, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.16144656118824668, |
| "grad_norm": 0.950721025466919, |
| "learning_rate": 0.0005909492511582803, |
| "loss": 4.7835, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.1668281132278549, |
| "grad_norm": 0.9828820824623108, |
| "learning_rate": 0.0005906260101282189, |
| "loss": 4.7715, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.17220966526746315, |
| "grad_norm": 0.9663535356521606, |
| "learning_rate": 0.0005903027690981575, |
| "loss": 4.751, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.17759121730707136, |
| "grad_norm": 0.8463447093963623, |
| "learning_rate": 0.000589979528068096, |
| "loss": 4.7339, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.18297276934667958, |
| "grad_norm": 0.9118891358375549, |
| "learning_rate": 0.0005896562870380347, |
| "loss": 4.6909, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.1883543213862878, |
| "grad_norm": 0.8051866888999939, |
| "learning_rate": 0.0005893330460079732, |
| "loss": 4.7067, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.19373587342589602, |
| "grad_norm": 0.9270610809326172, |
| "learning_rate": 0.0005890098049779118, |
| "loss": 4.6868, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.19911742546550426, |
| "grad_norm": 0.7820844650268555, |
| "learning_rate": 0.0005886865639478504, |
| "loss": 4.6451, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.20449897750511248, |
| "grad_norm": 0.9605551362037659, |
| "learning_rate": 0.0005883633229177889, |
| "loss": 4.6169, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.2098805295447207, |
| "grad_norm": 0.8132458925247192, |
| "learning_rate": 0.0005880400818877276, |
| "loss": 4.611, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.2152620815843289, |
| "grad_norm": 0.952850341796875, |
| "learning_rate": 0.0005877168408576662, |
| "loss": 4.5795, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.2152620815843289, |
| "eval_accuracy": 0.27085241776035757, |
| "eval_loss": 4.507682800292969, |
| "eval_runtime": 184.2381, |
| "eval_samples_per_second": 97.759, |
| "eval_steps_per_second": 6.112, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.22064363362393713, |
| "grad_norm": 0.9736102223396301, |
| "learning_rate": 0.0005873935998276048, |
| "loss": 4.5552, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.22602518566354537, |
| "grad_norm": 0.8927708268165588, |
| "learning_rate": 0.0005870703587975433, |
| "loss": 4.5378, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.2314067377031536, |
| "grad_norm": 1.0510213375091553, |
| "learning_rate": 0.0005867471177674818, |
| "loss": 4.539, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.2367882897427618, |
| "grad_norm": 0.9041006565093994, |
| "learning_rate": 0.0005864238767374205, |
| "loss": 4.5068, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.24216984178237003, |
| "grad_norm": 0.9170295596122742, |
| "learning_rate": 0.0005861006357073591, |
| "loss": 4.4987, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.24755139382197827, |
| "grad_norm": 1.4727824926376343, |
| "learning_rate": 0.0005857773946772977, |
| "loss": 4.4592, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.2529329458615865, |
| "grad_norm": 1.0264145135879517, |
| "learning_rate": 0.0005854541536472362, |
| "loss": 4.4577, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.2583144979011947, |
| "grad_norm": 0.8357675671577454, |
| "learning_rate": 0.0005851309126171749, |
| "loss": 4.4338, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.2636960499408029, |
| "grad_norm": 0.8532307744026184, |
| "learning_rate": 0.0005848076715871134, |
| "loss": 4.4297, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.26907760198041114, |
| "grad_norm": 0.7271891236305237, |
| "learning_rate": 0.000584484430557052, |
| "loss": 4.4147, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.27445915402001936, |
| "grad_norm": 0.7774198651313782, |
| "learning_rate": 0.0005841611895269906, |
| "loss": 4.4133, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.2798407060596276, |
| "grad_norm": 0.8901899456977844, |
| "learning_rate": 0.0005838379484969291, |
| "loss": 4.4158, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.2852222580992358, |
| "grad_norm": 1.0149247646331787, |
| "learning_rate": 0.0005835147074668678, |
| "loss": 4.3799, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.29060381013884407, |
| "grad_norm": 0.7539450526237488, |
| "learning_rate": 0.0005831914664368063, |
| "loss": 4.3692, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.2959853621784523, |
| "grad_norm": 0.8020486831665039, |
| "learning_rate": 0.0005828682254067449, |
| "loss": 4.3373, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.3013669142180605, |
| "grad_norm": 0.9698400497436523, |
| "learning_rate": 0.0005825449843766835, |
| "loss": 4.3398, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.3067484662576687, |
| "grad_norm": 0.9938604831695557, |
| "learning_rate": 0.0005822217433466221, |
| "loss": 4.3472, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.31213001829727693, |
| "grad_norm": 0.8308948278427124, |
| "learning_rate": 0.0005818985023165607, |
| "loss": 4.3363, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.31751157033688515, |
| "grad_norm": 0.8094744086265564, |
| "learning_rate": 0.0005815752612864992, |
| "loss": 4.3147, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.32289312237649337, |
| "grad_norm": 0.8576586246490479, |
| "learning_rate": 0.0005812520202564378, |
| "loss": 4.3173, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.32289312237649337, |
| "eval_accuracy": 0.2982191465728543, |
| "eval_loss": 4.243260860443115, |
| "eval_runtime": 184.3833, |
| "eval_samples_per_second": 97.682, |
| "eval_steps_per_second": 6.107, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.3282746744161016, |
| "grad_norm": 0.8846203088760376, |
| "learning_rate": 0.0005809287792263764, |
| "loss": 4.301, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.3336562264557098, |
| "grad_norm": 0.8314961791038513, |
| "learning_rate": 0.0005806055381963151, |
| "loss": 4.3089, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.3390377784953181, |
| "grad_norm": 0.8715083003044128, |
| "learning_rate": 0.0005802822971662536, |
| "loss": 4.2974, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.3444193305349263, |
| "grad_norm": 0.9279531836509705, |
| "learning_rate": 0.0005799590561361922, |
| "loss": 4.3041, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.3498008825745345, |
| "grad_norm": 0.8645572662353516, |
| "learning_rate": 0.0005796358151061307, |
| "loss": 4.287, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.35518243461414273, |
| "grad_norm": 0.9392287731170654, |
| "learning_rate": 0.0005793125740760694, |
| "loss": 4.2519, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.36056398665375095, |
| "grad_norm": 0.7573714852333069, |
| "learning_rate": 0.0005789893330460079, |
| "loss": 4.2413, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.36594553869335916, |
| "grad_norm": 0.6203613877296448, |
| "learning_rate": 0.0005786660920159465, |
| "loss": 4.2458, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.3713270907329674, |
| "grad_norm": 0.8517956137657166, |
| "learning_rate": 0.0005783428509858851, |
| "loss": 4.2433, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.3767086427725756, |
| "grad_norm": 0.832983136177063, |
| "learning_rate": 0.0005780196099558237, |
| "loss": 4.2253, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.3820901948121838, |
| "grad_norm": 0.5542912483215332, |
| "learning_rate": 0.0005776963689257623, |
| "loss": 4.2293, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.38747174685179203, |
| "grad_norm": 0.6788548827171326, |
| "learning_rate": 0.0005773731278957008, |
| "loss": 4.236, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.3928532988914003, |
| "grad_norm": 0.6434756517410278, |
| "learning_rate": 0.0005770498868656394, |
| "loss": 4.2193, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.3982348509310085, |
| "grad_norm": 0.7149718403816223, |
| "learning_rate": 0.000576726645835578, |
| "loss": 4.223, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.40361640297061674, |
| "grad_norm": 0.6300427913665771, |
| "learning_rate": 0.0005764034048055167, |
| "loss": 4.2004, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.40899795501022496, |
| "grad_norm": 0.8285446763038635, |
| "learning_rate": 0.0005760801637754552, |
| "loss": 4.1971, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.4143795070498332, |
| "grad_norm": 0.6956706643104553, |
| "learning_rate": 0.0005757569227453937, |
| "loss": 4.1895, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.4197610590894414, |
| "grad_norm": 0.770892858505249, |
| "learning_rate": 0.0005754336817153324, |
| "loss": 4.182, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.4251426111290496, |
| "grad_norm": 0.5891619920730591, |
| "learning_rate": 0.0005751104406852709, |
| "loss": 4.1666, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.4305241631686578, |
| "grad_norm": 0.6791778802871704, |
| "learning_rate": 0.0005747871996552096, |
| "loss": 4.1689, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.4305241631686578, |
| "eval_accuracy": 0.31222939319209253, |
| "eval_loss": 4.0884881019592285, |
| "eval_runtime": 184.3094, |
| "eval_samples_per_second": 97.722, |
| "eval_steps_per_second": 6.109, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.43590571520826604, |
| "grad_norm": 0.6507901549339294, |
| "learning_rate": 0.0005744639586251481, |
| "loss": 4.1549, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.44128726724787426, |
| "grad_norm": 0.7162640690803528, |
| "learning_rate": 0.0005741407175950867, |
| "loss": 4.1497, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.44666881928748253, |
| "grad_norm": 0.7263709902763367, |
| "learning_rate": 0.0005738174765650253, |
| "loss": 4.1271, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.45205037132709075, |
| "grad_norm": 0.6067292094230652, |
| "learning_rate": 0.0005734942355349638, |
| "loss": 4.1378, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.45743192336669897, |
| "grad_norm": 0.7140578031539917, |
| "learning_rate": 0.0005731709945049025, |
| "loss": 4.1394, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.4628134754063072, |
| "grad_norm": 0.8744107484817505, |
| "learning_rate": 0.000572847753474841, |
| "loss": 4.1203, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.4681950274459154, |
| "grad_norm": 0.6609825491905212, |
| "learning_rate": 0.0005725245124447796, |
| "loss": 4.1102, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.4735765794855236, |
| "grad_norm": 0.6945375204086304, |
| "learning_rate": 0.0005722012714147182, |
| "loss": 4.1125, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.47895813152513184, |
| "grad_norm": 0.7473881840705872, |
| "learning_rate": 0.0005718780303846568, |
| "loss": 4.1133, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.48433968356474005, |
| "grad_norm": 0.6308166980743408, |
| "learning_rate": 0.0005715547893545953, |
| "loss": 4.1187, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.48972123560434827, |
| "grad_norm": 0.8003607988357544, |
| "learning_rate": 0.000571231548324534, |
| "loss": 4.1168, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.49510278764395654, |
| "grad_norm": 0.6896440386772156, |
| "learning_rate": 0.0005709083072944725, |
| "loss": 4.1019, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.5004843396835648, |
| "grad_norm": 0.5829487442970276, |
| "learning_rate": 0.0005705850662644111, |
| "loss": 4.0929, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.505865891723173, |
| "grad_norm": 0.6313600540161133, |
| "learning_rate": 0.0005702618252343497, |
| "loss": 4.0804, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.5112474437627812, |
| "grad_norm": 0.6282418966293335, |
| "learning_rate": 0.0005699385842042882, |
| "loss": 4.1074, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.5166289958023894, |
| "grad_norm": 0.5928139090538025, |
| "learning_rate": 0.0005696153431742269, |
| "loss": 4.0956, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.5220105478419976, |
| "grad_norm": 0.6114786267280579, |
| "learning_rate": 0.0005692921021441655, |
| "loss": 4.0867, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.5273920998816058, |
| "grad_norm": 0.6260164380073547, |
| "learning_rate": 0.0005689688611141041, |
| "loss": 4.07, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.5327736519212141, |
| "grad_norm": 0.7232828140258789, |
| "learning_rate": 0.0005686456200840426, |
| "loss": 4.0782, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.5381552039608223, |
| "grad_norm": 0.6681451201438904, |
| "learning_rate": 0.0005683223790539811, |
| "loss": 4.0569, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5381552039608223, |
| "eval_accuracy": 0.3208234020066667, |
| "eval_loss": 3.996339797973633, |
| "eval_runtime": 184.1238, |
| "eval_samples_per_second": 97.82, |
| "eval_steps_per_second": 6.115, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5435367560004305, |
| "grad_norm": 0.6551739573478699, |
| "learning_rate": 0.0005679991380239198, |
| "loss": 4.0632, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.5489183080400387, |
| "grad_norm": 0.6114643812179565, |
| "learning_rate": 0.0005676758969938584, |
| "loss": 4.0459, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.5542998600796469, |
| "grad_norm": 0.6146873831748962, |
| "learning_rate": 0.000567352655963797, |
| "loss": 4.0601, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.5596814121192552, |
| "grad_norm": 0.6343026757240295, |
| "learning_rate": 0.0005670294149337355, |
| "loss": 4.044, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.5650629641588634, |
| "grad_norm": 0.6887774467468262, |
| "learning_rate": 0.0005667061739036742, |
| "loss": 4.0369, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.5704445161984716, |
| "grad_norm": 0.775775671005249, |
| "learning_rate": 0.0005663829328736127, |
| "loss": 4.041, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.5758260682380799, |
| "grad_norm": 0.646882176399231, |
| "learning_rate": 0.0005660596918435512, |
| "loss": 4.0477, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.5812076202776881, |
| "grad_norm": 0.645609438419342, |
| "learning_rate": 0.0005657364508134899, |
| "loss": 4.0269, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.5865891723172963, |
| "grad_norm": 0.637544572353363, |
| "learning_rate": 0.0005654132097834284, |
| "loss": 4.0299, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.5919707243569046, |
| "grad_norm": 0.5601037740707397, |
| "learning_rate": 0.0005650899687533671, |
| "loss": 4.0364, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.5973522763965128, |
| "grad_norm": 0.6552086472511292, |
| "learning_rate": 0.0005647667277233056, |
| "loss": 4.0319, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.602733828436121, |
| "grad_norm": 0.6226367950439453, |
| "learning_rate": 0.0005644434866932442, |
| "loss": 4.0239, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.6081153804757292, |
| "grad_norm": 0.6150733828544617, |
| "learning_rate": 0.0005641202456631828, |
| "loss": 4.0096, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.6134969325153374, |
| "grad_norm": 0.6194325685501099, |
| "learning_rate": 0.0005637970046331214, |
| "loss": 4.0067, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.6188784845549457, |
| "grad_norm": 0.6258173584938049, |
| "learning_rate": 0.00056347376360306, |
| "loss": 4.0021, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.6242600365945539, |
| "grad_norm": 0.5563567876815796, |
| "learning_rate": 0.0005631505225729985, |
| "loss": 3.9966, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.6296415886341621, |
| "grad_norm": 0.7610234618186951, |
| "learning_rate": 0.0005628272815429371, |
| "loss": 3.9795, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.6350231406737703, |
| "grad_norm": 0.6821996569633484, |
| "learning_rate": 0.0005625040405128757, |
| "loss": 3.9997, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.6404046927133785, |
| "grad_norm": 0.6400163173675537, |
| "learning_rate": 0.0005621807994828143, |
| "loss": 4.0132, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.6457862447529867, |
| "grad_norm": 0.7269625067710876, |
| "learning_rate": 0.0005618575584527529, |
| "loss": 4.0125, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.6457862447529867, |
| "eval_accuracy": 0.3269717429694248, |
| "eval_loss": 3.925267457962036, |
| "eval_runtime": 184.1043, |
| "eval_samples_per_second": 97.83, |
| "eval_steps_per_second": 6.116, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.651167796792595, |
| "grad_norm": 0.6363526582717896, |
| "learning_rate": 0.0005615343174226915, |
| "loss": 3.9904, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.6565493488322032, |
| "grad_norm": 0.6771544218063354, |
| "learning_rate": 0.00056121107639263, |
| "loss": 3.9856, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.6619309008718114, |
| "grad_norm": 0.6212124824523926, |
| "learning_rate": 0.0005608878353625687, |
| "loss": 3.9796, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.6673124529114196, |
| "grad_norm": 0.7166991233825684, |
| "learning_rate": 0.0005605645943325072, |
| "loss": 3.9659, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.6726940049510278, |
| "grad_norm": 0.6104118824005127, |
| "learning_rate": 0.0005602413533024458, |
| "loss": 3.9671, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.6780755569906362, |
| "grad_norm": 0.6991214156150818, |
| "learning_rate": 0.0005599181122723844, |
| "loss": 3.9859, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.6834571090302444, |
| "grad_norm": 0.6045164465904236, |
| "learning_rate": 0.000559594871242323, |
| "loss": 3.9565, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.6888386610698526, |
| "grad_norm": 0.6517415046691895, |
| "learning_rate": 0.0005592716302122616, |
| "loss": 3.9662, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.6942202131094608, |
| "grad_norm": 0.6841729283332825, |
| "learning_rate": 0.0005589483891822001, |
| "loss": 3.9579, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.699601765149069, |
| "grad_norm": 0.591302216053009, |
| "learning_rate": 0.0005586251481521387, |
| "loss": 3.9575, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.7049833171886772, |
| "grad_norm": 0.6354752779006958, |
| "learning_rate": 0.0005583083719426786, |
| "loss": 3.9505, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.7103648692282855, |
| "grad_norm": 0.6361892223358154, |
| "learning_rate": 0.0005579851309126171, |
| "loss": 3.9571, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.7157464212678937, |
| "grad_norm": 0.5964643955230713, |
| "learning_rate": 0.0005576618898825558, |
| "loss": 3.9424, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.7211279733075019, |
| "grad_norm": 0.718136727809906, |
| "learning_rate": 0.0005573386488524943, |
| "loss": 3.9508, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.7265095253471101, |
| "grad_norm": 0.6186783909797668, |
| "learning_rate": 0.0005570154078224328, |
| "loss": 3.9547, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.7318910773867183, |
| "grad_norm": 0.5839592814445496, |
| "learning_rate": 0.0005566921667923715, |
| "loss": 3.9387, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.7372726294263265, |
| "grad_norm": 0.5628191232681274, |
| "learning_rate": 0.00055636892576231, |
| "loss": 3.9408, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.7426541814659348, |
| "grad_norm": 0.7687479853630066, |
| "learning_rate": 0.0005560456847322487, |
| "loss": 3.9491, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.748035733505543, |
| "grad_norm": 0.5638557076454163, |
| "learning_rate": 0.0005557224437021872, |
| "loss": 3.9488, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.7534172855451512, |
| "grad_norm": 0.6913768649101257, |
| "learning_rate": 0.0005553992026721258, |
| "loss": 3.9257, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7534172855451512, |
| "eval_accuracy": 0.3327180682472692, |
| "eval_loss": 3.864708662033081, |
| "eval_runtime": 184.286, |
| "eval_samples_per_second": 97.734, |
| "eval_steps_per_second": 6.11, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7587988375847594, |
| "grad_norm": 0.5453241467475891, |
| "learning_rate": 0.0005550759616420644, |
| "loss": 3.9302, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.7641803896243676, |
| "grad_norm": 0.6204462051391602, |
| "learning_rate": 0.000554752720612003, |
| "loss": 3.9324, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.7695619416639758, |
| "grad_norm": 0.5680400133132935, |
| "learning_rate": 0.0005544294795819415, |
| "loss": 3.9254, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.7749434937035841, |
| "grad_norm": 0.6182464957237244, |
| "learning_rate": 0.0005541062385518801, |
| "loss": 3.9359, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.7803250457431924, |
| "grad_norm": 0.6028207540512085, |
| "learning_rate": 0.0005537829975218188, |
| "loss": 3.9335, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.7857065977828006, |
| "grad_norm": 0.5548726320266724, |
| "learning_rate": 0.0005534597564917573, |
| "loss": 3.913, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.7910881498224088, |
| "grad_norm": 0.5448580384254456, |
| "learning_rate": 0.0005531365154616959, |
| "loss": 3.9274, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.796469701862017, |
| "grad_norm": 0.686208963394165, |
| "learning_rate": 0.0005528132744316344, |
| "loss": 3.9146, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.8018512539016253, |
| "grad_norm": 0.6147581934928894, |
| "learning_rate": 0.0005524900334015731, |
| "loss": 3.9209, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.8072328059412335, |
| "grad_norm": 0.6350402235984802, |
| "learning_rate": 0.0005521667923715117, |
| "loss": 3.9114, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.8126143579808417, |
| "grad_norm": 0.598574161529541, |
| "learning_rate": 0.0005518435513414502, |
| "loss": 3.9369, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.8179959100204499, |
| "grad_norm": 0.6298902630805969, |
| "learning_rate": 0.0005515203103113888, |
| "loss": 3.9029, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.8233774620600581, |
| "grad_norm": 0.5852333307266235, |
| "learning_rate": 0.0005511970692813274, |
| "loss": 3.8858, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.8287590140996663, |
| "grad_norm": 0.6372036337852478, |
| "learning_rate": 0.000550873828251266, |
| "loss": 3.9097, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.8341405661392746, |
| "grad_norm": 0.6588336825370789, |
| "learning_rate": 0.0005505505872212045, |
| "loss": 3.9069, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.8395221181788828, |
| "grad_norm": 0.6206655502319336, |
| "learning_rate": 0.0005502273461911432, |
| "loss": 3.8922, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.844903670218491, |
| "grad_norm": 0.6114910244941711, |
| "learning_rate": 0.0005499041051610817, |
| "loss": 3.8982, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.8502852222580992, |
| "grad_norm": 0.6116674542427063, |
| "learning_rate": 0.0005495808641310204, |
| "loss": 3.8957, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.8556667742977074, |
| "grad_norm": 0.6183781623840332, |
| "learning_rate": 0.0005492576231009589, |
| "loss": 3.8894, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.8610483263373157, |
| "grad_norm": 0.6823594570159912, |
| "learning_rate": 0.0005489343820708974, |
| "loss": 3.8987, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8610483263373157, |
| "eval_accuracy": 0.3365249394776252, |
| "eval_loss": 3.825028419494629, |
| "eval_runtime": 184.1616, |
| "eval_samples_per_second": 97.8, |
| "eval_steps_per_second": 6.114, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8664298783769239, |
| "grad_norm": 0.5773827433586121, |
| "learning_rate": 0.0005486111410408361, |
| "loss": 3.8826, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.8718114304165321, |
| "grad_norm": 0.6269072890281677, |
| "learning_rate": 0.0005482879000107746, |
| "loss": 3.8726, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.8771929824561403, |
| "grad_norm": 0.5973522663116455, |
| "learning_rate": 0.0005479646589807133, |
| "loss": 3.87, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.8825745344957485, |
| "grad_norm": 0.6495493650436401, |
| "learning_rate": 0.0005476414179506518, |
| "loss": 3.8642, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.8879560865353568, |
| "grad_norm": 0.5791492462158203, |
| "learning_rate": 0.0005473181769205904, |
| "loss": 3.8504, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.8933376385749651, |
| "grad_norm": 0.5900397300720215, |
| "learning_rate": 0.000546994935890529, |
| "loss": 3.8612, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.8987191906145733, |
| "grad_norm": 0.5566937327384949, |
| "learning_rate": 0.0005466716948604677, |
| "loss": 3.8794, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.9041007426541815, |
| "grad_norm": 0.6604649424552917, |
| "learning_rate": 0.0005463484538304062, |
| "loss": 3.8659, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.9094822946937897, |
| "grad_norm": 0.5881569385528564, |
| "learning_rate": 0.0005460252128003447, |
| "loss": 3.8646, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.9148638467333979, |
| "grad_norm": 0.586227297782898, |
| "learning_rate": 0.0005457019717702833, |
| "loss": 3.8677, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.9202453987730062, |
| "grad_norm": 0.5569677352905273, |
| "learning_rate": 0.0005453787307402219, |
| "loss": 3.8548, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.9256269508126144, |
| "grad_norm": 0.5891553163528442, |
| "learning_rate": 0.0005450554897101605, |
| "loss": 3.8741, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.9310085028522226, |
| "grad_norm": 0.5274822115898132, |
| "learning_rate": 0.0005447322486800991, |
| "loss": 3.8376, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.9363900548918308, |
| "grad_norm": 0.6486146450042725, |
| "learning_rate": 0.0005444090076500377, |
| "loss": 3.8607, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.941771606931439, |
| "grad_norm": 0.5777039527893066, |
| "learning_rate": 0.0005440922314405775, |
| "loss": 3.8713, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.9471531589710472, |
| "grad_norm": 0.5529994964599609, |
| "learning_rate": 0.0005437689904105161, |
| "loss": 3.8491, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.9525347110106555, |
| "grad_norm": 0.5525511503219604, |
| "learning_rate": 0.0005434457493804546, |
| "loss": 3.8521, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.9579162630502637, |
| "grad_norm": 0.6336971521377563, |
| "learning_rate": 0.0005431225083503932, |
| "loss": 3.8456, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.9632978150898719, |
| "grad_norm": 0.6398906707763672, |
| "learning_rate": 0.0005427992673203318, |
| "loss": 3.8494, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.9686793671294801, |
| "grad_norm": 0.5508920550346375, |
| "learning_rate": 0.0005424760262902704, |
| "loss": 3.837, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9686793671294801, |
| "eval_accuracy": 0.34044165877756155, |
| "eval_loss": 3.782215118408203, |
| "eval_runtime": 183.9619, |
| "eval_samples_per_second": 97.906, |
| "eval_steps_per_second": 6.121, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9740609191690883, |
| "grad_norm": 0.5972535014152527, |
| "learning_rate": 0.000542152785260209, |
| "loss": 3.8537, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.9794424712086965, |
| "grad_norm": 0.6076335310935974, |
| "learning_rate": 0.0005418295442301476, |
| "loss": 3.8602, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.9848240232483048, |
| "grad_norm": 0.6718324422836304, |
| "learning_rate": 0.0005415063032000861, |
| "loss": 3.8268, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.9902055752879131, |
| "grad_norm": 0.5986764430999756, |
| "learning_rate": 0.0005411830621700248, |
| "loss": 3.8535, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.9955871273275213, |
| "grad_norm": 0.5580224990844727, |
| "learning_rate": 0.0005408598211399633, |
| "loss": 3.8519, |
| "step": 9250 |
| }, |
| { |
| "epoch": 1.0009686793671295, |
| "grad_norm": 0.5752728581428528, |
| "learning_rate": 0.0005405365801099019, |
| "loss": 3.8272, |
| "step": 9300 |
| }, |
| { |
| "epoch": 1.0063502314067376, |
| "grad_norm": 0.6517691016197205, |
| "learning_rate": 0.0005402133390798405, |
| "loss": 3.7748, |
| "step": 9350 |
| }, |
| { |
| "epoch": 1.011731783446346, |
| "grad_norm": 0.5505602955818176, |
| "learning_rate": 0.000539890098049779, |
| "loss": 3.773, |
| "step": 9400 |
| }, |
| { |
| "epoch": 1.017113335485954, |
| "grad_norm": 0.5624109506607056, |
| "learning_rate": 0.0005395668570197177, |
| "loss": 3.784, |
| "step": 9450 |
| }, |
| { |
| "epoch": 1.0224948875255624, |
| "grad_norm": 0.5721527338027954, |
| "learning_rate": 0.0005392436159896562, |
| "loss": 3.7722, |
| "step": 9500 |
| }, |
| { |
| "epoch": 1.0278764395651705, |
| "grad_norm": 0.5518069863319397, |
| "learning_rate": 0.0005389203749595948, |
| "loss": 3.7686, |
| "step": 9550 |
| }, |
| { |
| "epoch": 1.0332579916047788, |
| "grad_norm": 0.5892334580421448, |
| "learning_rate": 0.0005385971339295334, |
| "loss": 3.768, |
| "step": 9600 |
| }, |
| { |
| "epoch": 1.0386395436443872, |
| "grad_norm": 0.5937067866325378, |
| "learning_rate": 0.000538273892899472, |
| "loss": 3.7844, |
| "step": 9650 |
| }, |
| { |
| "epoch": 1.0440210956839953, |
| "grad_norm": 0.6311096549034119, |
| "learning_rate": 0.0005379506518694106, |
| "loss": 3.7654, |
| "step": 9700 |
| }, |
| { |
| "epoch": 1.0494026477236036, |
| "grad_norm": 0.5864768624305725, |
| "learning_rate": 0.0005376274108393491, |
| "loss": 3.7586, |
| "step": 9750 |
| }, |
| { |
| "epoch": 1.0547841997632117, |
| "grad_norm": 0.5333934426307678, |
| "learning_rate": 0.0005373041698092877, |
| "loss": 3.7562, |
| "step": 9800 |
| }, |
| { |
| "epoch": 1.06016575180282, |
| "grad_norm": 0.6055988073348999, |
| "learning_rate": 0.0005369809287792263, |
| "loss": 3.7772, |
| "step": 9850 |
| }, |
| { |
| "epoch": 1.0655473038424281, |
| "grad_norm": 0.534672737121582, |
| "learning_rate": 0.000536657687749165, |
| "loss": 3.7656, |
| "step": 9900 |
| }, |
| { |
| "epoch": 1.0709288558820365, |
| "grad_norm": 0.644813597202301, |
| "learning_rate": 0.0005363344467191035, |
| "loss": 3.7714, |
| "step": 9950 |
| }, |
| { |
| "epoch": 1.0763104079216446, |
| "grad_norm": 0.6052290797233582, |
| "learning_rate": 0.000536011205689042, |
| "loss": 3.7725, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0763104079216446, |
| "eval_accuracy": 0.34394756150867145, |
| "eval_loss": 3.754002809524536, |
| "eval_runtime": 184.3547, |
| "eval_samples_per_second": 97.698, |
| "eval_steps_per_second": 6.108, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.081691959961253, |
| "grad_norm": 0.5823466777801514, |
| "learning_rate": 0.0005356879646589807, |
| "loss": 3.7756, |
| "step": 10050 |
| }, |
| { |
| "epoch": 1.087073512000861, |
| "grad_norm": 0.5655180811882019, |
| "learning_rate": 0.0005353647236289192, |
| "loss": 3.7547, |
| "step": 10100 |
| }, |
| { |
| "epoch": 1.0924550640404693, |
| "grad_norm": 0.5540060997009277, |
| "learning_rate": 0.0005350414825988579, |
| "loss": 3.7692, |
| "step": 10150 |
| }, |
| { |
| "epoch": 1.0978366160800774, |
| "grad_norm": 0.5735357999801636, |
| "learning_rate": 0.0005347182415687964, |
| "loss": 3.7782, |
| "step": 10200 |
| }, |
| { |
| "epoch": 1.1032181681196858, |
| "grad_norm": 0.6658727526664734, |
| "learning_rate": 0.000534395000538735, |
| "loss": 3.7972, |
| "step": 10250 |
| }, |
| { |
| "epoch": 1.1085997201592939, |
| "grad_norm": 0.6762692332267761, |
| "learning_rate": 0.0005340717595086736, |
| "loss": 3.7446, |
| "step": 10300 |
| }, |
| { |
| "epoch": 1.1139812721989022, |
| "grad_norm": 0.5580800175666809, |
| "learning_rate": 0.0005337485184786122, |
| "loss": 3.7644, |
| "step": 10350 |
| }, |
| { |
| "epoch": 1.1193628242385103, |
| "grad_norm": 0.6203590631484985, |
| "learning_rate": 0.0005334252774485507, |
| "loss": 3.77, |
| "step": 10400 |
| }, |
| { |
| "epoch": 1.1247443762781186, |
| "grad_norm": 0.5936480760574341, |
| "learning_rate": 0.0005331020364184894, |
| "loss": 3.7849, |
| "step": 10450 |
| }, |
| { |
| "epoch": 1.1301259283177267, |
| "grad_norm": 0.5875517129898071, |
| "learning_rate": 0.0005327787953884279, |
| "loss": 3.7623, |
| "step": 10500 |
| }, |
| { |
| "epoch": 1.135507480357335, |
| "grad_norm": 0.5761053562164307, |
| "learning_rate": 0.0005324555543583665, |
| "loss": 3.7619, |
| "step": 10550 |
| }, |
| { |
| "epoch": 1.1408890323969434, |
| "grad_norm": 0.6478322744369507, |
| "learning_rate": 0.0005321323133283051, |
| "loss": 3.749, |
| "step": 10600 |
| }, |
| { |
| "epoch": 1.1462705844365515, |
| "grad_norm": 0.6216648817062378, |
| "learning_rate": 0.0005318090722982436, |
| "loss": 3.7505, |
| "step": 10650 |
| }, |
| { |
| "epoch": 1.1516521364761596, |
| "grad_norm": 0.5962103009223938, |
| "learning_rate": 0.0005314858312681823, |
| "loss": 3.7563, |
| "step": 10700 |
| }, |
| { |
| "epoch": 1.157033688515768, |
| "grad_norm": 0.6087831854820251, |
| "learning_rate": 0.0005311625902381209, |
| "loss": 3.7601, |
| "step": 10750 |
| }, |
| { |
| "epoch": 1.1624152405553763, |
| "grad_norm": 0.6549679636955261, |
| "learning_rate": 0.0005308393492080595, |
| "loss": 3.7554, |
| "step": 10800 |
| }, |
| { |
| "epoch": 1.1677967925949844, |
| "grad_norm": 0.5932186841964722, |
| "learning_rate": 0.000530516108177998, |
| "loss": 3.7631, |
| "step": 10850 |
| }, |
| { |
| "epoch": 1.1731783446345927, |
| "grad_norm": 0.5829725861549377, |
| "learning_rate": 0.0005301928671479365, |
| "loss": 3.7619, |
| "step": 10900 |
| }, |
| { |
| "epoch": 1.1785598966742008, |
| "grad_norm": 0.6248167753219604, |
| "learning_rate": 0.0005298696261178752, |
| "loss": 3.7551, |
| "step": 10950 |
| }, |
| { |
| "epoch": 1.1839414487138091, |
| "grad_norm": 0.5710045695304871, |
| "learning_rate": 0.0005295463850878138, |
| "loss": 3.7577, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1839414487138091, |
| "eval_accuracy": 0.3465084014215709, |
| "eval_loss": 3.7283456325531006, |
| "eval_runtime": 184.2095, |
| "eval_samples_per_second": 97.775, |
| "eval_steps_per_second": 6.113, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1893230007534172, |
| "grad_norm": 0.5649476051330566, |
| "learning_rate": 0.0005292296088783535, |
| "loss": 3.7498, |
| "step": 11050 |
| }, |
| { |
| "epoch": 1.1947045527930256, |
| "grad_norm": 0.5573504567146301, |
| "learning_rate": 0.0005289063678482922, |
| "loss": 3.7595, |
| "step": 11100 |
| }, |
| { |
| "epoch": 1.2000861048326337, |
| "grad_norm": 0.5430361032485962, |
| "learning_rate": 0.0005285831268182307, |
| "loss": 3.7529, |
| "step": 11150 |
| }, |
| { |
| "epoch": 1.205467656872242, |
| "grad_norm": 0.5485645532608032, |
| "learning_rate": 0.0005282598857881694, |
| "loss": 3.7594, |
| "step": 11200 |
| }, |
| { |
| "epoch": 1.21084920891185, |
| "grad_norm": 0.5611818432807922, |
| "learning_rate": 0.0005279366447581079, |
| "loss": 3.7436, |
| "step": 11250 |
| }, |
| { |
| "epoch": 1.2162307609514584, |
| "grad_norm": 0.5610353350639343, |
| "learning_rate": 0.0005276134037280465, |
| "loss": 3.7635, |
| "step": 11300 |
| }, |
| { |
| "epoch": 1.2216123129910665, |
| "grad_norm": 0.5841692686080933, |
| "learning_rate": 0.0005272901626979851, |
| "loss": 3.7718, |
| "step": 11350 |
| }, |
| { |
| "epoch": 1.2269938650306749, |
| "grad_norm": 0.5393407940864563, |
| "learning_rate": 0.0005269669216679236, |
| "loss": 3.7475, |
| "step": 11400 |
| }, |
| { |
| "epoch": 1.232375417070283, |
| "grad_norm": 0.561508059501648, |
| "learning_rate": 0.0005266436806378623, |
| "loss": 3.7624, |
| "step": 11450 |
| }, |
| { |
| "epoch": 1.2377569691098913, |
| "grad_norm": 0.6286839842796326, |
| "learning_rate": 0.0005263204396078008, |
| "loss": 3.7587, |
| "step": 11500 |
| }, |
| { |
| "epoch": 1.2431385211494996, |
| "grad_norm": 0.6179966926574707, |
| "learning_rate": 0.0005259971985777394, |
| "loss": 3.7298, |
| "step": 11550 |
| }, |
| { |
| "epoch": 1.2485200731891077, |
| "grad_norm": 0.5545424222946167, |
| "learning_rate": 0.000525673957547678, |
| "loss": 3.7309, |
| "step": 11600 |
| }, |
| { |
| "epoch": 1.2539016252287158, |
| "grad_norm": 0.5041254162788391, |
| "learning_rate": 0.0005253507165176167, |
| "loss": 3.7386, |
| "step": 11650 |
| }, |
| { |
| "epoch": 1.2592831772683242, |
| "grad_norm": 0.5765021443367004, |
| "learning_rate": 0.0005250274754875552, |
| "loss": 3.7517, |
| "step": 11700 |
| }, |
| { |
| "epoch": 1.2646647293079325, |
| "grad_norm": 0.5996841788291931, |
| "learning_rate": 0.0005247042344574938, |
| "loss": 3.7354, |
| "step": 11750 |
| }, |
| { |
| "epoch": 1.2700462813475406, |
| "grad_norm": 0.5650532841682434, |
| "learning_rate": 0.0005243809934274323, |
| "loss": 3.7446, |
| "step": 11800 |
| }, |
| { |
| "epoch": 1.275427833387149, |
| "grad_norm": 0.6425164937973022, |
| "learning_rate": 0.0005240577523973709, |
| "loss": 3.7483, |
| "step": 11850 |
| }, |
| { |
| "epoch": 1.280809385426757, |
| "grad_norm": 0.6343678832054138, |
| "learning_rate": 0.0005237345113673095, |
| "loss": 3.7166, |
| "step": 11900 |
| }, |
| { |
| "epoch": 1.2861909374663654, |
| "grad_norm": 0.6143035888671875, |
| "learning_rate": 0.0005234112703372481, |
| "loss": 3.7269, |
| "step": 11950 |
| }, |
| { |
| "epoch": 1.2915724895059735, |
| "grad_norm": 0.5740583539009094, |
| "learning_rate": 0.0005230880293071867, |
| "loss": 3.7502, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2915724895059735, |
| "eval_accuracy": 0.34853836332460886, |
| "eval_loss": 3.703268051147461, |
| "eval_runtime": 184.8299, |
| "eval_samples_per_second": 97.446, |
| "eval_steps_per_second": 6.092, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2969540415455818, |
| "grad_norm": 0.5388001799583435, |
| "learning_rate": 0.0005227647882771253, |
| "loss": 3.7413, |
| "step": 12050 |
| }, |
| { |
| "epoch": 1.30233559358519, |
| "grad_norm": 0.5561287999153137, |
| "learning_rate": 0.0005224415472470639, |
| "loss": 3.7277, |
| "step": 12100 |
| }, |
| { |
| "epoch": 1.3077171456247982, |
| "grad_norm": 0.5779213309288025, |
| "learning_rate": 0.0005221183062170024, |
| "loss": 3.7232, |
| "step": 12150 |
| }, |
| { |
| "epoch": 1.3130986976644063, |
| "grad_norm": 0.5532439351081848, |
| "learning_rate": 0.0005217950651869409, |
| "loss": 3.742, |
| "step": 12200 |
| }, |
| { |
| "epoch": 1.3184802497040147, |
| "grad_norm": 0.619836151599884, |
| "learning_rate": 0.0005214718241568796, |
| "loss": 3.7342, |
| "step": 12250 |
| }, |
| { |
| "epoch": 1.3238618017436228, |
| "grad_norm": 0.5282240509986877, |
| "learning_rate": 0.0005211485831268182, |
| "loss": 3.7343, |
| "step": 12300 |
| }, |
| { |
| "epoch": 1.329243353783231, |
| "grad_norm": 0.5425903797149658, |
| "learning_rate": 0.0005208253420967568, |
| "loss": 3.724, |
| "step": 12350 |
| }, |
| { |
| "epoch": 1.3346249058228392, |
| "grad_norm": 0.5502631664276123, |
| "learning_rate": 0.0005205021010666953, |
| "loss": 3.7427, |
| "step": 12400 |
| }, |
| { |
| "epoch": 1.3400064578624475, |
| "grad_norm": 0.5649226307868958, |
| "learning_rate": 0.0005201788600366339, |
| "loss": 3.7308, |
| "step": 12450 |
| }, |
| { |
| "epoch": 1.3453880099020559, |
| "grad_norm": 0.6057882905006409, |
| "learning_rate": 0.0005198556190065725, |
| "loss": 3.7329, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.350769561941664, |
| "grad_norm": 0.5616216659545898, |
| "learning_rate": 0.0005195323779765112, |
| "loss": 3.7356, |
| "step": 12550 |
| }, |
| { |
| "epoch": 1.356151113981272, |
| "grad_norm": 0.5632114410400391, |
| "learning_rate": 0.0005192091369464497, |
| "loss": 3.7176, |
| "step": 12600 |
| }, |
| { |
| "epoch": 1.3615326660208804, |
| "grad_norm": 0.5245296359062195, |
| "learning_rate": 0.0005188858959163882, |
| "loss": 3.7434, |
| "step": 12650 |
| }, |
| { |
| "epoch": 1.3669142180604887, |
| "grad_norm": 0.566827654838562, |
| "learning_rate": 0.0005185626548863269, |
| "loss": 3.7283, |
| "step": 12700 |
| }, |
| { |
| "epoch": 1.3722957701000968, |
| "grad_norm": 0.5973430871963501, |
| "learning_rate": 0.0005182394138562654, |
| "loss": 3.7161, |
| "step": 12750 |
| }, |
| { |
| "epoch": 1.3776773221397052, |
| "grad_norm": 0.6542244553565979, |
| "learning_rate": 0.0005179161728262041, |
| "loss": 3.7091, |
| "step": 12800 |
| }, |
| { |
| "epoch": 1.3830588741793133, |
| "grad_norm": 0.5695278644561768, |
| "learning_rate": 0.0005175929317961426, |
| "loss": 3.7389, |
| "step": 12850 |
| }, |
| { |
| "epoch": 1.3884404262189216, |
| "grad_norm": 0.5513305068016052, |
| "learning_rate": 0.0005172696907660812, |
| "loss": 3.7244, |
| "step": 12900 |
| }, |
| { |
| "epoch": 1.3938219782585297, |
| "grad_norm": 0.6421329379081726, |
| "learning_rate": 0.0005169464497360198, |
| "loss": 3.721, |
| "step": 12950 |
| }, |
| { |
| "epoch": 1.399203530298138, |
| "grad_norm": 0.5805631279945374, |
| "learning_rate": 0.0005166232087059583, |
| "loss": 3.7256, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.399203530298138, |
| "eval_accuracy": 0.3509969608700749, |
| "eval_loss": 3.6784543991088867, |
| "eval_runtime": 184.6324, |
| "eval_samples_per_second": 97.551, |
| "eval_steps_per_second": 6.099, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.4045850823377461, |
| "grad_norm": 0.5761812925338745, |
| "learning_rate": 0.0005162999676758969, |
| "loss": 3.716, |
| "step": 13050 |
| }, |
| { |
| "epoch": 1.4099666343773545, |
| "grad_norm": 0.6273037791252136, |
| "learning_rate": 0.0005159767266458355, |
| "loss": 3.7195, |
| "step": 13100 |
| }, |
| { |
| "epoch": 1.4153481864169626, |
| "grad_norm": 0.5659797787666321, |
| "learning_rate": 0.0005156534856157741, |
| "loss": 3.7202, |
| "step": 13150 |
| }, |
| { |
| "epoch": 1.420729738456571, |
| "grad_norm": 0.5564277768135071, |
| "learning_rate": 0.0005153302445857127, |
| "loss": 3.7025, |
| "step": 13200 |
| }, |
| { |
| "epoch": 1.426111290496179, |
| "grad_norm": 0.5497375130653381, |
| "learning_rate": 0.0005150070035556513, |
| "loss": 3.7169, |
| "step": 13250 |
| }, |
| { |
| "epoch": 1.4314928425357873, |
| "grad_norm": 0.5665987730026245, |
| "learning_rate": 0.0005146902273461911, |
| "loss": 3.7145, |
| "step": 13300 |
| }, |
| { |
| "epoch": 1.4368743945753955, |
| "grad_norm": 0.5927955508232117, |
| "learning_rate": 0.0005143669863161297, |
| "loss": 3.7128, |
| "step": 13350 |
| }, |
| { |
| "epoch": 1.4422559466150038, |
| "grad_norm": 0.553808867931366, |
| "learning_rate": 0.0005140437452860683, |
| "loss": 3.7183, |
| "step": 13400 |
| }, |
| { |
| "epoch": 1.447637498654612, |
| "grad_norm": 0.5321818590164185, |
| "learning_rate": 0.0005137205042560069, |
| "loss": 3.7137, |
| "step": 13450 |
| }, |
| { |
| "epoch": 1.4530190506942202, |
| "grad_norm": 0.6058953404426575, |
| "learning_rate": 0.0005133972632259455, |
| "loss": 3.7225, |
| "step": 13500 |
| }, |
| { |
| "epoch": 1.4584006027338283, |
| "grad_norm": 0.5817305445671082, |
| "learning_rate": 0.000513074022195884, |
| "loss": 3.7164, |
| "step": 13550 |
| }, |
| { |
| "epoch": 1.4637821547734367, |
| "grad_norm": 0.565757155418396, |
| "learning_rate": 0.0005127507811658226, |
| "loss": 3.7019, |
| "step": 13600 |
| }, |
| { |
| "epoch": 1.469163706813045, |
| "grad_norm": 0.5681242942810059, |
| "learning_rate": 0.0005124275401357612, |
| "loss": 3.7118, |
| "step": 13650 |
| }, |
| { |
| "epoch": 1.474545258852653, |
| "grad_norm": 0.5758519768714905, |
| "learning_rate": 0.0005121042991056997, |
| "loss": 3.7026, |
| "step": 13700 |
| }, |
| { |
| "epoch": 1.4799268108922612, |
| "grad_norm": 0.6025715470314026, |
| "learning_rate": 0.0005117810580756384, |
| "loss": 3.6834, |
| "step": 13750 |
| }, |
| { |
| "epoch": 1.4853083629318695, |
| "grad_norm": 0.5520561337471008, |
| "learning_rate": 0.0005114578170455769, |
| "loss": 3.6964, |
| "step": 13800 |
| }, |
| { |
| "epoch": 1.4906899149714778, |
| "grad_norm": 0.604680061340332, |
| "learning_rate": 0.0005111345760155156, |
| "loss": 3.7189, |
| "step": 13850 |
| }, |
| { |
| "epoch": 1.496071467011086, |
| "grad_norm": 0.7172315716743469, |
| "learning_rate": 0.0005108113349854541, |
| "loss": 3.7077, |
| "step": 13900 |
| }, |
| { |
| "epoch": 1.501453019050694, |
| "grad_norm": 0.5513233542442322, |
| "learning_rate": 0.0005104880939553926, |
| "loss": 3.6821, |
| "step": 13950 |
| }, |
| { |
| "epoch": 1.5068345710903024, |
| "grad_norm": 0.6223646402359009, |
| "learning_rate": 0.0005101648529253313, |
| "loss": 3.7026, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5068345710903024, |
| "eval_accuracy": 0.3527702846520951, |
| "eval_loss": 3.66027569770813, |
| "eval_runtime": 184.9365, |
| "eval_samples_per_second": 97.39, |
| "eval_steps_per_second": 6.089, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5122161231299107, |
| "grad_norm": 0.5914530754089355, |
| "learning_rate": 0.0005098416118952699, |
| "loss": 3.6923, |
| "step": 14050 |
| }, |
| { |
| "epoch": 1.5175976751695188, |
| "grad_norm": 0.5675052404403687, |
| "learning_rate": 0.0005095183708652085, |
| "loss": 3.7059, |
| "step": 14100 |
| }, |
| { |
| "epoch": 1.5229792272091272, |
| "grad_norm": 0.5429747700691223, |
| "learning_rate": 0.000509195129835147, |
| "loss": 3.6907, |
| "step": 14150 |
| }, |
| { |
| "epoch": 1.5283607792487355, |
| "grad_norm": 0.5908409953117371, |
| "learning_rate": 0.0005088718888050856, |
| "loss": 3.6958, |
| "step": 14200 |
| }, |
| { |
| "epoch": 1.5337423312883436, |
| "grad_norm": 0.600124180316925, |
| "learning_rate": 0.0005085486477750242, |
| "loss": 3.702, |
| "step": 14250 |
| }, |
| { |
| "epoch": 1.5391238833279517, |
| "grad_norm": 0.5889090895652771, |
| "learning_rate": 0.0005082254067449629, |
| "loss": 3.704, |
| "step": 14300 |
| }, |
| { |
| "epoch": 1.54450543536756, |
| "grad_norm": 0.5812370777130127, |
| "learning_rate": 0.0005079021657149014, |
| "loss": 3.7089, |
| "step": 14350 |
| }, |
| { |
| "epoch": 1.5498869874071683, |
| "grad_norm": 0.640620231628418, |
| "learning_rate": 0.0005075789246848399, |
| "loss": 3.676, |
| "step": 14400 |
| }, |
| { |
| "epoch": 1.5552685394467765, |
| "grad_norm": 0.5642813444137573, |
| "learning_rate": 0.0005072556836547785, |
| "loss": 3.7029, |
| "step": 14450 |
| }, |
| { |
| "epoch": 1.5606500914863846, |
| "grad_norm": 0.5681257247924805, |
| "learning_rate": 0.0005069324426247171, |
| "loss": 3.6864, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.566031643525993, |
| "grad_norm": 0.5891586542129517, |
| "learning_rate": 0.0005066092015946557, |
| "loss": 3.6999, |
| "step": 14550 |
| }, |
| { |
| "epoch": 1.5714131955656012, |
| "grad_norm": 0.5758991837501526, |
| "learning_rate": 0.0005062859605645943, |
| "loss": 3.6825, |
| "step": 14600 |
| }, |
| { |
| "epoch": 1.5767947476052093, |
| "grad_norm": 0.5694742202758789, |
| "learning_rate": 0.0005059627195345329, |
| "loss": 3.6859, |
| "step": 14650 |
| }, |
| { |
| "epoch": 1.5821762996448174, |
| "grad_norm": 0.5646113157272339, |
| "learning_rate": 0.0005056394785044715, |
| "loss": 3.6862, |
| "step": 14700 |
| }, |
| { |
| "epoch": 1.5875578516844258, |
| "grad_norm": 0.5336816310882568, |
| "learning_rate": 0.00050531623747441, |
| "loss": 3.6966, |
| "step": 14750 |
| }, |
| { |
| "epoch": 1.592939403724034, |
| "grad_norm": 0.546383798122406, |
| "learning_rate": 0.0005049929964443486, |
| "loss": 3.6899, |
| "step": 14800 |
| }, |
| { |
| "epoch": 1.5983209557636422, |
| "grad_norm": 0.6023898124694824, |
| "learning_rate": 0.0005046697554142871, |
| "loss": 3.6788, |
| "step": 14850 |
| }, |
| { |
| "epoch": 1.6037025078032503, |
| "grad_norm": 0.5576519966125488, |
| "learning_rate": 0.0005043465143842258, |
| "loss": 3.708, |
| "step": 14900 |
| }, |
| { |
| "epoch": 1.6090840598428586, |
| "grad_norm": 0.5140047073364258, |
| "learning_rate": 0.0005040232733541644, |
| "loss": 3.6735, |
| "step": 14950 |
| }, |
| { |
| "epoch": 1.614465611882467, |
| "grad_norm": 0.5628222823143005, |
| "learning_rate": 0.000503700032324103, |
| "loss": 3.6954, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.614465611882467, |
| "eval_accuracy": 0.35463226919057184, |
| "eval_loss": 3.6429481506347656, |
| "eval_runtime": 184.8414, |
| "eval_samples_per_second": 97.44, |
| "eval_steps_per_second": 6.092, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.619847163922075, |
| "grad_norm": 0.5722090601921082, |
| "learning_rate": 0.0005033767912940415, |
| "loss": 3.6689, |
| "step": 15050 |
| }, |
| { |
| "epoch": 1.6252287159616834, |
| "grad_norm": 0.6029437184333801, |
| "learning_rate": 0.0005030535502639802, |
| "loss": 3.682, |
| "step": 15100 |
| }, |
| { |
| "epoch": 1.6306102680012917, |
| "grad_norm": 0.5833781957626343, |
| "learning_rate": 0.0005027303092339187, |
| "loss": 3.6812, |
| "step": 15150 |
| }, |
| { |
| "epoch": 1.6359918200408998, |
| "grad_norm": 0.5960364937782288, |
| "learning_rate": 0.0005024135330244585, |
| "loss": 3.6792, |
| "step": 15200 |
| }, |
| { |
| "epoch": 1.641373372080508, |
| "grad_norm": 0.5771504044532776, |
| "learning_rate": 0.0005020902919943972, |
| "loss": 3.6862, |
| "step": 15250 |
| }, |
| { |
| "epoch": 1.6467549241201163, |
| "grad_norm": 0.6127772331237793, |
| "learning_rate": 0.0005017670509643357, |
| "loss": 3.6975, |
| "step": 15300 |
| }, |
| { |
| "epoch": 1.6521364761597246, |
| "grad_norm": 0.5885007381439209, |
| "learning_rate": 0.0005014438099342743, |
| "loss": 3.68, |
| "step": 15350 |
| }, |
| { |
| "epoch": 1.6575180281993327, |
| "grad_norm": 0.6041487455368042, |
| "learning_rate": 0.0005011205689042129, |
| "loss": 3.6602, |
| "step": 15400 |
| }, |
| { |
| "epoch": 1.6628995802389408, |
| "grad_norm": 0.5758805871009827, |
| "learning_rate": 0.0005007973278741514, |
| "loss": 3.6751, |
| "step": 15450 |
| }, |
| { |
| "epoch": 1.6682811322785491, |
| "grad_norm": 0.5186255574226379, |
| "learning_rate": 0.00050047408684409, |
| "loss": 3.6795, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.6736626843181575, |
| "grad_norm": 0.5514072179794312, |
| "learning_rate": 0.0005001508458140286, |
| "loss": 3.6774, |
| "step": 15550 |
| }, |
| { |
| "epoch": 1.6790442363577656, |
| "grad_norm": 0.5339574813842773, |
| "learning_rate": 0.0004998276047839673, |
| "loss": 3.6791, |
| "step": 15600 |
| }, |
| { |
| "epoch": 1.6844257883973737, |
| "grad_norm": 0.5713083148002625, |
| "learning_rate": 0.0004995043637539058, |
| "loss": 3.685, |
| "step": 15650 |
| }, |
| { |
| "epoch": 1.689807340436982, |
| "grad_norm": 0.543732762336731, |
| "learning_rate": 0.0004991811227238443, |
| "loss": 3.6788, |
| "step": 15700 |
| }, |
| { |
| "epoch": 1.6951888924765903, |
| "grad_norm": 0.5611950755119324, |
| "learning_rate": 0.0004988578816937829, |
| "loss": 3.6788, |
| "step": 15750 |
| }, |
| { |
| "epoch": 1.7005704445161984, |
| "grad_norm": 1.1558012962341309, |
| "learning_rate": 0.0004985346406637215, |
| "loss": 3.6853, |
| "step": 15800 |
| }, |
| { |
| "epoch": 1.7059519965558065, |
| "grad_norm": 0.5830097794532776, |
| "learning_rate": 0.0004982113996336602, |
| "loss": 3.6819, |
| "step": 15850 |
| }, |
| { |
| "epoch": 1.7113335485954149, |
| "grad_norm": 0.5888424515724182, |
| "learning_rate": 0.0004978881586035987, |
| "loss": 3.6499, |
| "step": 15900 |
| }, |
| { |
| "epoch": 1.7167151006350232, |
| "grad_norm": 0.5985537171363831, |
| "learning_rate": 0.0004975649175735373, |
| "loss": 3.6656, |
| "step": 15950 |
| }, |
| { |
| "epoch": 1.7220966526746313, |
| "grad_norm": 0.5927515625953674, |
| "learning_rate": 0.0004972416765434759, |
| "loss": 3.6735, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7220966526746313, |
| "eval_accuracy": 0.3565279361242711, |
| "eval_loss": 3.6250205039978027, |
| "eval_runtime": 184.8324, |
| "eval_samples_per_second": 97.445, |
| "eval_steps_per_second": 6.092, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7274782047142396, |
| "grad_norm": 0.5919635891914368, |
| "learning_rate": 0.0004969184355134145, |
| "loss": 3.6675, |
| "step": 16050 |
| }, |
| { |
| "epoch": 1.732859756753848, |
| "grad_norm": 0.541195809841156, |
| "learning_rate": 0.0004965951944833531, |
| "loss": 3.6654, |
| "step": 16100 |
| }, |
| { |
| "epoch": 1.738241308793456, |
| "grad_norm": 0.5535009503364563, |
| "learning_rate": 0.0004962719534532916, |
| "loss": 3.6582, |
| "step": 16150 |
| }, |
| { |
| "epoch": 1.7436228608330642, |
| "grad_norm": 0.5698593854904175, |
| "learning_rate": 0.0004959487124232302, |
| "loss": 3.6677, |
| "step": 16200 |
| }, |
| { |
| "epoch": 1.7490044128726725, |
| "grad_norm": 0.5787349343299866, |
| "learning_rate": 0.0004956254713931688, |
| "loss": 3.665, |
| "step": 16250 |
| }, |
| { |
| "epoch": 1.7543859649122808, |
| "grad_norm": 0.61586993932724, |
| "learning_rate": 0.0004953022303631074, |
| "loss": 3.6734, |
| "step": 16300 |
| }, |
| { |
| "epoch": 1.759767516951889, |
| "grad_norm": 0.5861958861351013, |
| "learning_rate": 0.0004949789893330459, |
| "loss": 3.6602, |
| "step": 16350 |
| }, |
| { |
| "epoch": 1.765149068991497, |
| "grad_norm": 0.5748381614685059, |
| "learning_rate": 0.0004946557483029846, |
| "loss": 3.6709, |
| "step": 16400 |
| }, |
| { |
| "epoch": 1.7705306210311054, |
| "grad_norm": 0.5469369888305664, |
| "learning_rate": 0.0004943325072729231, |
| "loss": 3.6531, |
| "step": 16450 |
| }, |
| { |
| "epoch": 1.7759121730707137, |
| "grad_norm": 0.5762391686439514, |
| "learning_rate": 0.0004940092662428617, |
| "loss": 3.664, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.7812937251103218, |
| "grad_norm": 0.5528908967971802, |
| "learning_rate": 0.0004936860252128003, |
| "loss": 3.6412, |
| "step": 16550 |
| }, |
| { |
| "epoch": 1.78667527714993, |
| "grad_norm": 0.5524837970733643, |
| "learning_rate": 0.0004933627841827388, |
| "loss": 3.6725, |
| "step": 16600 |
| }, |
| { |
| "epoch": 1.7920568291895382, |
| "grad_norm": 0.5726085305213928, |
| "learning_rate": 0.0004930395431526775, |
| "loss": 3.6769, |
| "step": 16650 |
| }, |
| { |
| "epoch": 1.7974383812291466, |
| "grad_norm": 0.5498278737068176, |
| "learning_rate": 0.0004927163021226161, |
| "loss": 3.6422, |
| "step": 16700 |
| }, |
| { |
| "epoch": 1.8028199332687547, |
| "grad_norm": 0.5690969824790955, |
| "learning_rate": 0.0004923930610925547, |
| "loss": 3.653, |
| "step": 16750 |
| }, |
| { |
| "epoch": 1.8082014853083628, |
| "grad_norm": 0.5563623905181885, |
| "learning_rate": 0.0004920698200624932, |
| "loss": 3.6843, |
| "step": 16800 |
| }, |
| { |
| "epoch": 1.813583037347971, |
| "grad_norm": 0.559473991394043, |
| "learning_rate": 0.0004917465790324317, |
| "loss": 3.656, |
| "step": 16850 |
| }, |
| { |
| "epoch": 1.8189645893875794, |
| "grad_norm": 1.0415881872177124, |
| "learning_rate": 0.0004914233380023704, |
| "loss": 3.6526, |
| "step": 16900 |
| }, |
| { |
| "epoch": 1.8243461414271875, |
| "grad_norm": 0.5609972476959229, |
| "learning_rate": 0.0004911000969723089, |
| "loss": 3.6581, |
| "step": 16950 |
| }, |
| { |
| "epoch": 1.8297276934667959, |
| "grad_norm": 0.5753895044326782, |
| "learning_rate": 0.0004907768559422476, |
| "loss": 3.6405, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8297276934667959, |
| "eval_accuracy": 0.3581517535326585, |
| "eval_loss": 3.6095125675201416, |
| "eval_runtime": 185.0185, |
| "eval_samples_per_second": 97.347, |
| "eval_steps_per_second": 6.086, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8351092455064042, |
| "grad_norm": 0.5633401870727539, |
| "learning_rate": 0.0004904536149121861, |
| "loss": 3.6531, |
| "step": 17050 |
| }, |
| { |
| "epoch": 1.8404907975460123, |
| "grad_norm": 0.5552918910980225, |
| "learning_rate": 0.0004901303738821248, |
| "loss": 3.647, |
| "step": 17100 |
| }, |
| { |
| "epoch": 1.8458723495856204, |
| "grad_norm": 0.5768826603889465, |
| "learning_rate": 0.0004898071328520633, |
| "loss": 3.6472, |
| "step": 17150 |
| }, |
| { |
| "epoch": 1.8512539016252287, |
| "grad_norm": 0.5807743072509766, |
| "learning_rate": 0.0004894838918220019, |
| "loss": 3.6456, |
| "step": 17200 |
| }, |
| { |
| "epoch": 1.856635453664837, |
| "grad_norm": 0.6019062399864197, |
| "learning_rate": 0.0004891606507919405, |
| "loss": 3.6529, |
| "step": 17250 |
| }, |
| { |
| "epoch": 1.8620170057044452, |
| "grad_norm": 0.5893161296844482, |
| "learning_rate": 0.000488837409761879, |
| "loss": 3.6711, |
| "step": 17300 |
| }, |
| { |
| "epoch": 1.8673985577440533, |
| "grad_norm": 0.5603693723678589, |
| "learning_rate": 0.0004885141687318177, |
| "loss": 3.6614, |
| "step": 17350 |
| }, |
| { |
| "epoch": 1.8727801097836616, |
| "grad_norm": 0.631818413734436, |
| "learning_rate": 0.00048819092770175623, |
| "loss": 3.6642, |
| "step": 17400 |
| }, |
| { |
| "epoch": 1.87816166182327, |
| "grad_norm": 0.5838807821273804, |
| "learning_rate": 0.0004878676866716948, |
| "loss": 3.6581, |
| "step": 17450 |
| }, |
| { |
| "epoch": 1.883543213862878, |
| "grad_norm": 0.5909974575042725, |
| "learning_rate": 0.00048754444564163337, |
| "loss": 3.6499, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.8889247659024861, |
| "grad_norm": 0.5818292498588562, |
| "learning_rate": 0.000487221204611572, |
| "loss": 3.6563, |
| "step": 17550 |
| }, |
| { |
| "epoch": 1.8943063179420945, |
| "grad_norm": 0.6259658336639404, |
| "learning_rate": 0.00048689796358151056, |
| "loss": 3.6557, |
| "step": 17600 |
| }, |
| { |
| "epoch": 1.8996878699817028, |
| "grad_norm": 0.5432948470115662, |
| "learning_rate": 0.00048657472255144915, |
| "loss": 3.6522, |
| "step": 17650 |
| }, |
| { |
| "epoch": 1.905069422021311, |
| "grad_norm": 0.5726639032363892, |
| "learning_rate": 0.00048625148152138775, |
| "loss": 3.6559, |
| "step": 17700 |
| }, |
| { |
| "epoch": 1.910450974060919, |
| "grad_norm": 0.6220903396606445, |
| "learning_rate": 0.0004859282404913263, |
| "loss": 3.6441, |
| "step": 17750 |
| }, |
| { |
| "epoch": 1.9158325261005273, |
| "grad_norm": 0.5544300079345703, |
| "learning_rate": 0.0004856049994612649, |
| "loss": 3.643, |
| "step": 17800 |
| }, |
| { |
| "epoch": 1.9212140781401357, |
| "grad_norm": 0.6041421294212341, |
| "learning_rate": 0.00048528175843120353, |
| "loss": 3.6461, |
| "step": 17850 |
| }, |
| { |
| "epoch": 1.9265956301797438, |
| "grad_norm": 0.5323398113250732, |
| "learning_rate": 0.0004849585174011421, |
| "loss": 3.6522, |
| "step": 17900 |
| }, |
| { |
| "epoch": 1.931977182219352, |
| "grad_norm": 0.5484367609024048, |
| "learning_rate": 0.00048463527637108067, |
| "loss": 3.6477, |
| "step": 17950 |
| }, |
| { |
| "epoch": 1.9373587342589604, |
| "grad_norm": 0.6333380937576294, |
| "learning_rate": 0.0004843120353410192, |
| "loss": 3.6439, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9373587342589604, |
| "eval_accuracy": 0.3589290562920833, |
| "eval_loss": 3.5955045223236084, |
| "eval_runtime": 184.9578, |
| "eval_samples_per_second": 97.379, |
| "eval_steps_per_second": 6.088, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9427402862985685, |
| "grad_norm": 0.5808707475662231, |
| "learning_rate": 0.0004839887943109578, |
| "loss": 3.64, |
| "step": 18050 |
| }, |
| { |
| "epoch": 1.9481218383381766, |
| "grad_norm": 0.5633965134620667, |
| "learning_rate": 0.00048366555328089645, |
| "loss": 3.629, |
| "step": 18100 |
| }, |
| { |
| "epoch": 1.953503390377785, |
| "grad_norm": 0.5531556606292725, |
| "learning_rate": 0.000483342312250835, |
| "loss": 3.6406, |
| "step": 18150 |
| }, |
| { |
| "epoch": 1.9588849424173933, |
| "grad_norm": 0.635948896408081, |
| "learning_rate": 0.0004830190712207736, |
| "loss": 3.6499, |
| "step": 18200 |
| }, |
| { |
| "epoch": 1.9642664944570014, |
| "grad_norm": 0.5666787624359131, |
| "learning_rate": 0.0004826958301907122, |
| "loss": 3.6559, |
| "step": 18250 |
| }, |
| { |
| "epoch": 1.9696480464966095, |
| "grad_norm": 0.5713225603103638, |
| "learning_rate": 0.0004823725891606507, |
| "loss": 3.6499, |
| "step": 18300 |
| }, |
| { |
| "epoch": 1.9750295985362178, |
| "grad_norm": 0.6086699962615967, |
| "learning_rate": 0.0004820493481305893, |
| "loss": 3.6176, |
| "step": 18350 |
| }, |
| { |
| "epoch": 1.9804111505758262, |
| "grad_norm": 0.577656626701355, |
| "learning_rate": 0.00048172610710052797, |
| "loss": 3.6473, |
| "step": 18400 |
| }, |
| { |
| "epoch": 1.9857927026154343, |
| "grad_norm": 0.5730327367782593, |
| "learning_rate": 0.0004814028660704665, |
| "loss": 3.6484, |
| "step": 18450 |
| }, |
| { |
| "epoch": 1.9911742546550424, |
| "grad_norm": 0.542858362197876, |
| "learning_rate": 0.0004810796250404051, |
| "loss": 3.6421, |
| "step": 18500 |
| }, |
| { |
| "epoch": 1.9965558066946507, |
| "grad_norm": 0.5895008444786072, |
| "learning_rate": 0.00048075638401034364, |
| "loss": 3.6335, |
| "step": 18550 |
| }, |
| { |
| "epoch": 2.001937358734259, |
| "grad_norm": 0.5415585041046143, |
| "learning_rate": 0.00048043314298028224, |
| "loss": 3.6123, |
| "step": 18600 |
| }, |
| { |
| "epoch": 2.007318910773867, |
| "grad_norm": 0.5612041354179382, |
| "learning_rate": 0.00048010990195022083, |
| "loss": 3.5606, |
| "step": 18650 |
| }, |
| { |
| "epoch": 2.0127004628134753, |
| "grad_norm": 0.5955508351325989, |
| "learning_rate": 0.0004797866609201594, |
| "loss": 3.5534, |
| "step": 18700 |
| }, |
| { |
| "epoch": 2.018082014853084, |
| "grad_norm": 0.5306816101074219, |
| "learning_rate": 0.000479463419890098, |
| "loss": 3.5606, |
| "step": 18750 |
| }, |
| { |
| "epoch": 2.023463566892692, |
| "grad_norm": 0.6179097890853882, |
| "learning_rate": 0.0004791401788600366, |
| "loss": 3.5398, |
| "step": 18800 |
| }, |
| { |
| "epoch": 2.0288451189323, |
| "grad_norm": 0.5494289994239807, |
| "learning_rate": 0.00047881693782997515, |
| "loss": 3.5486, |
| "step": 18850 |
| }, |
| { |
| "epoch": 2.034226670971908, |
| "grad_norm": 0.6472309827804565, |
| "learning_rate": 0.00047849369679991375, |
| "loss": 3.5584, |
| "step": 18900 |
| }, |
| { |
| "epoch": 2.0396082230115167, |
| "grad_norm": 0.5931106209754944, |
| "learning_rate": 0.0004781704557698523, |
| "loss": 3.5716, |
| "step": 18950 |
| }, |
| { |
| "epoch": 2.044989775051125, |
| "grad_norm": 0.555692732334137, |
| "learning_rate": 0.00047784721473979094, |
| "loss": 3.5585, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.044989775051125, |
| "eval_accuracy": 0.3609611912528775, |
| "eval_loss": 3.585639476776123, |
| "eval_runtime": 184.5705, |
| "eval_samples_per_second": 97.583, |
| "eval_steps_per_second": 6.101, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.050371327090733, |
| "grad_norm": 0.5866180062294006, |
| "learning_rate": 0.00047752397370972953, |
| "loss": 3.5712, |
| "step": 19050 |
| }, |
| { |
| "epoch": 2.055752879130341, |
| "grad_norm": 0.6226833462715149, |
| "learning_rate": 0.0004772007326796681, |
| "loss": 3.5541, |
| "step": 19100 |
| }, |
| { |
| "epoch": 2.0611344311699495, |
| "grad_norm": 0.622482180595398, |
| "learning_rate": 0.00047687749164960667, |
| "loss": 3.5497, |
| "step": 19150 |
| }, |
| { |
| "epoch": 2.0665159832095576, |
| "grad_norm": 0.5930569171905518, |
| "learning_rate": 0.00047655425061954526, |
| "loss": 3.543, |
| "step": 19200 |
| }, |
| { |
| "epoch": 2.0718975352491658, |
| "grad_norm": 0.6076319217681885, |
| "learning_rate": 0.00047623100958948386, |
| "loss": 3.5663, |
| "step": 19250 |
| }, |
| { |
| "epoch": 2.0772790872887743, |
| "grad_norm": 0.5735373497009277, |
| "learning_rate": 0.00047590776855942245, |
| "loss": 3.5541, |
| "step": 19300 |
| }, |
| { |
| "epoch": 2.0826606393283824, |
| "grad_norm": 0.6126927137374878, |
| "learning_rate": 0.00047559099234996226, |
| "loss": 3.559, |
| "step": 19350 |
| }, |
| { |
| "epoch": 2.0880421913679905, |
| "grad_norm": 0.5821768641471863, |
| "learning_rate": 0.00047526775131990085, |
| "loss": 3.5555, |
| "step": 19400 |
| }, |
| { |
| "epoch": 2.0934237434075986, |
| "grad_norm": 0.5864294171333313, |
| "learning_rate": 0.0004749445102898394, |
| "loss": 3.5562, |
| "step": 19450 |
| }, |
| { |
| "epoch": 2.098805295447207, |
| "grad_norm": 0.6452410817146301, |
| "learning_rate": 0.000474621269259778, |
| "loss": 3.5593, |
| "step": 19500 |
| }, |
| { |
| "epoch": 2.1041868474868153, |
| "grad_norm": 0.5725786089897156, |
| "learning_rate": 0.0004742980282297166, |
| "loss": 3.5832, |
| "step": 19550 |
| }, |
| { |
| "epoch": 2.1095683995264234, |
| "grad_norm": 0.6051315665245056, |
| "learning_rate": 0.0004739747871996551, |
| "loss": 3.5634, |
| "step": 19600 |
| }, |
| { |
| "epoch": 2.1149499515660315, |
| "grad_norm": 0.5349266529083252, |
| "learning_rate": 0.00047365154616959377, |
| "loss": 3.5558, |
| "step": 19650 |
| }, |
| { |
| "epoch": 2.12033150360564, |
| "grad_norm": 0.5834625363349915, |
| "learning_rate": 0.00047332830513953237, |
| "loss": 3.5382, |
| "step": 19700 |
| }, |
| { |
| "epoch": 2.125713055645248, |
| "grad_norm": 0.5570130944252014, |
| "learning_rate": 0.0004730050641094709, |
| "loss": 3.5871, |
| "step": 19750 |
| }, |
| { |
| "epoch": 2.1310946076848563, |
| "grad_norm": 0.6045011281967163, |
| "learning_rate": 0.0004726818230794095, |
| "loss": 3.5578, |
| "step": 19800 |
| }, |
| { |
| "epoch": 2.1364761597244644, |
| "grad_norm": 0.5597689747810364, |
| "learning_rate": 0.00047235858204934804, |
| "loss": 3.5797, |
| "step": 19850 |
| }, |
| { |
| "epoch": 2.141857711764073, |
| "grad_norm": 0.582958996295929, |
| "learning_rate": 0.0004720353410192867, |
| "loss": 3.5748, |
| "step": 19900 |
| }, |
| { |
| "epoch": 2.147239263803681, |
| "grad_norm": 0.5808094143867493, |
| "learning_rate": 0.0004717120999892253, |
| "loss": 3.5758, |
| "step": 19950 |
| }, |
| { |
| "epoch": 2.152620815843289, |
| "grad_norm": 0.6471470594406128, |
| "learning_rate": 0.0004713888589591638, |
| "loss": 3.5596, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.152620815843289, |
| "eval_accuracy": 0.36173936323540484, |
| "eval_loss": 3.5779969692230225, |
| "eval_runtime": 184.816, |
| "eval_samples_per_second": 97.454, |
| "eval_steps_per_second": 6.093, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.1580023678828972, |
| "grad_norm": 0.5732181668281555, |
| "learning_rate": 0.0004710720827497037, |
| "loss": 3.5622, |
| "step": 20050 |
| }, |
| { |
| "epoch": 2.163383919922506, |
| "grad_norm": 0.582992672920227, |
| "learning_rate": 0.00047074884171964223, |
| "loss": 3.5469, |
| "step": 20100 |
| }, |
| { |
| "epoch": 2.168765471962114, |
| "grad_norm": 0.5529500842094421, |
| "learning_rate": 0.0004704256006895808, |
| "loss": 3.5583, |
| "step": 20150 |
| }, |
| { |
| "epoch": 2.174147024001722, |
| "grad_norm": 0.587487518787384, |
| "learning_rate": 0.00047010235965951936, |
| "loss": 3.5612, |
| "step": 20200 |
| }, |
| { |
| "epoch": 2.1795285760413305, |
| "grad_norm": 0.5818561911582947, |
| "learning_rate": 0.00046977911862945796, |
| "loss": 3.5628, |
| "step": 20250 |
| }, |
| { |
| "epoch": 2.1849101280809387, |
| "grad_norm": 0.5599380731582642, |
| "learning_rate": 0.0004694558775993966, |
| "loss": 3.5771, |
| "step": 20300 |
| }, |
| { |
| "epoch": 2.1902916801205468, |
| "grad_norm": 0.5862258076667786, |
| "learning_rate": 0.00046913263656933515, |
| "loss": 3.5635, |
| "step": 20350 |
| }, |
| { |
| "epoch": 2.195673232160155, |
| "grad_norm": 0.5630679726600647, |
| "learning_rate": 0.00046880939553927374, |
| "loss": 3.5593, |
| "step": 20400 |
| }, |
| { |
| "epoch": 2.2010547841997634, |
| "grad_norm": 0.6166080832481384, |
| "learning_rate": 0.00046848615450921234, |
| "loss": 3.5634, |
| "step": 20450 |
| }, |
| { |
| "epoch": 2.2064363362393715, |
| "grad_norm": 0.5633708834648132, |
| "learning_rate": 0.0004681629134791509, |
| "loss": 3.5678, |
| "step": 20500 |
| }, |
| { |
| "epoch": 2.2118178882789796, |
| "grad_norm": 0.5323472023010254, |
| "learning_rate": 0.00046783967244908947, |
| "loss": 3.577, |
| "step": 20550 |
| }, |
| { |
| "epoch": 2.2171994403185877, |
| "grad_norm": 0.5587198138237, |
| "learning_rate": 0.0004675164314190281, |
| "loss": 3.5656, |
| "step": 20600 |
| }, |
| { |
| "epoch": 2.2225809923581963, |
| "grad_norm": 0.6203996539115906, |
| "learning_rate": 0.00046719319038896666, |
| "loss": 3.5725, |
| "step": 20650 |
| }, |
| { |
| "epoch": 2.2279625443978044, |
| "grad_norm": 0.6120006442070007, |
| "learning_rate": 0.00046686994935890526, |
| "loss": 3.5511, |
| "step": 20700 |
| }, |
| { |
| "epoch": 2.2333440964374125, |
| "grad_norm": 0.5724984407424927, |
| "learning_rate": 0.0004665467083288438, |
| "loss": 3.5857, |
| "step": 20750 |
| }, |
| { |
| "epoch": 2.2387256484770206, |
| "grad_norm": 0.5859040021896362, |
| "learning_rate": 0.0004662234672987824, |
| "loss": 3.5626, |
| "step": 20800 |
| }, |
| { |
| "epoch": 2.244107200516629, |
| "grad_norm": 0.6283140778541565, |
| "learning_rate": 0.00046590022626872104, |
| "loss": 3.554, |
| "step": 20850 |
| }, |
| { |
| "epoch": 2.2494887525562373, |
| "grad_norm": 0.6019113659858704, |
| "learning_rate": 0.0004655769852386596, |
| "loss": 3.5601, |
| "step": 20900 |
| }, |
| { |
| "epoch": 2.2548703045958454, |
| "grad_norm": 0.5885460376739502, |
| "learning_rate": 0.0004652537442085982, |
| "loss": 3.5968, |
| "step": 20950 |
| }, |
| { |
| "epoch": 2.2602518566354535, |
| "grad_norm": 0.5559438467025757, |
| "learning_rate": 0.00046493050317853677, |
| "loss": 3.5619, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.2602518566354535, |
| "eval_accuracy": 0.362556867563321, |
| "eval_loss": 3.565824508666992, |
| "eval_runtime": 184.5191, |
| "eval_samples_per_second": 97.61, |
| "eval_steps_per_second": 6.102, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.265633408675062, |
| "grad_norm": 0.5775765776634216, |
| "learning_rate": 0.0004646072621484753, |
| "loss": 3.5615, |
| "step": 21050 |
| }, |
| { |
| "epoch": 2.27101496071467, |
| "grad_norm": 0.669364869594574, |
| "learning_rate": 0.0004642840211184139, |
| "loss": 3.5574, |
| "step": 21100 |
| }, |
| { |
| "epoch": 2.2763965127542782, |
| "grad_norm": 0.5841721892356873, |
| "learning_rate": 0.00046396078008835255, |
| "loss": 3.5482, |
| "step": 21150 |
| }, |
| { |
| "epoch": 2.281778064793887, |
| "grad_norm": 0.6061359643936157, |
| "learning_rate": 0.0004636375390582911, |
| "loss": 3.5743, |
| "step": 21200 |
| }, |
| { |
| "epoch": 2.287159616833495, |
| "grad_norm": 0.629613995552063, |
| "learning_rate": 0.0004633142980282297, |
| "loss": 3.5631, |
| "step": 21250 |
| }, |
| { |
| "epoch": 2.292541168873103, |
| "grad_norm": 0.5746256113052368, |
| "learning_rate": 0.00046299105699816823, |
| "loss": 3.5674, |
| "step": 21300 |
| }, |
| { |
| "epoch": 2.297922720912711, |
| "grad_norm": 0.5825964212417603, |
| "learning_rate": 0.0004626678159681068, |
| "loss": 3.5444, |
| "step": 21350 |
| }, |
| { |
| "epoch": 2.303304272952319, |
| "grad_norm": 0.6356550455093384, |
| "learning_rate": 0.0004623445749380454, |
| "loss": 3.5491, |
| "step": 21400 |
| }, |
| { |
| "epoch": 2.3086858249919278, |
| "grad_norm": 0.5417643189430237, |
| "learning_rate": 0.000462021333907984, |
| "loss": 3.5494, |
| "step": 21450 |
| }, |
| { |
| "epoch": 2.314067377031536, |
| "grad_norm": 0.543899416923523, |
| "learning_rate": 0.0004616980928779226, |
| "loss": 3.5541, |
| "step": 21500 |
| }, |
| { |
| "epoch": 2.319448929071144, |
| "grad_norm": 0.6250973343849182, |
| "learning_rate": 0.0004613748518478612, |
| "loss": 3.5749, |
| "step": 21550 |
| }, |
| { |
| "epoch": 2.3248304811107525, |
| "grad_norm": 0.6272781491279602, |
| "learning_rate": 0.00046105161081779974, |
| "loss": 3.574, |
| "step": 21600 |
| }, |
| { |
| "epoch": 2.3302120331503606, |
| "grad_norm": 0.6016408205032349, |
| "learning_rate": 0.00046072836978773834, |
| "loss": 3.5511, |
| "step": 21650 |
| }, |
| { |
| "epoch": 2.3355935851899687, |
| "grad_norm": 0.5845016837120056, |
| "learning_rate": 0.000460405128757677, |
| "loss": 3.5522, |
| "step": 21700 |
| }, |
| { |
| "epoch": 2.340975137229577, |
| "grad_norm": 0.5752265453338623, |
| "learning_rate": 0.0004600818877276155, |
| "loss": 3.5494, |
| "step": 21750 |
| }, |
| { |
| "epoch": 2.3463566892691854, |
| "grad_norm": 0.6028287410736084, |
| "learning_rate": 0.0004597586466975541, |
| "loss": 3.5626, |
| "step": 21800 |
| }, |
| { |
| "epoch": 2.3517382413087935, |
| "grad_norm": 0.5635079145431519, |
| "learning_rate": 0.00045943540566749266, |
| "loss": 3.5678, |
| "step": 21850 |
| }, |
| { |
| "epoch": 2.3571197933484016, |
| "grad_norm": 0.6141257286071777, |
| "learning_rate": 0.00045911216463743126, |
| "loss": 3.5413, |
| "step": 21900 |
| }, |
| { |
| "epoch": 2.3625013453880097, |
| "grad_norm": 0.6080167293548584, |
| "learning_rate": 0.00045878892360736985, |
| "loss": 3.5746, |
| "step": 21950 |
| }, |
| { |
| "epoch": 2.3678828974276183, |
| "grad_norm": 0.6267144083976746, |
| "learning_rate": 0.00045846568257730845, |
| "loss": 3.5593, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3678828974276183, |
| "eval_accuracy": 0.3641260325691377, |
| "eval_loss": 3.552825689315796, |
| "eval_runtime": 185.2813, |
| "eval_samples_per_second": 97.209, |
| "eval_steps_per_second": 6.077, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3732644494672264, |
| "grad_norm": 0.591131865978241, |
| "learning_rate": 0.00045814244154724704, |
| "loss": 3.5471, |
| "step": 22050 |
| }, |
| { |
| "epoch": 2.3786460015068345, |
| "grad_norm": 0.5588999390602112, |
| "learning_rate": 0.00045781920051718563, |
| "loss": 3.5592, |
| "step": 22100 |
| }, |
| { |
| "epoch": 2.384027553546443, |
| "grad_norm": 0.5897203087806702, |
| "learning_rate": 0.0004574959594871242, |
| "loss": 3.5639, |
| "step": 22150 |
| }, |
| { |
| "epoch": 2.389409105586051, |
| "grad_norm": 0.5783429741859436, |
| "learning_rate": 0.00045717271845706277, |
| "loss": 3.5625, |
| "step": 22200 |
| }, |
| { |
| "epoch": 2.3947906576256592, |
| "grad_norm": 0.595344603061676, |
| "learning_rate": 0.0004568494774270013, |
| "loss": 3.5561, |
| "step": 22250 |
| }, |
| { |
| "epoch": 2.4001722096652673, |
| "grad_norm": 0.6287717819213867, |
| "learning_rate": 0.00045652623639693996, |
| "loss": 3.5577, |
| "step": 22300 |
| }, |
| { |
| "epoch": 2.4055537617048754, |
| "grad_norm": 0.7863216996192932, |
| "learning_rate": 0.00045620299536687855, |
| "loss": 3.5778, |
| "step": 22350 |
| }, |
| { |
| "epoch": 2.410935313744484, |
| "grad_norm": 0.5969358086585999, |
| "learning_rate": 0.0004558797543368171, |
| "loss": 3.5605, |
| "step": 22400 |
| }, |
| { |
| "epoch": 2.416316865784092, |
| "grad_norm": 0.6180084347724915, |
| "learning_rate": 0.0004555565133067557, |
| "loss": 3.5457, |
| "step": 22450 |
| }, |
| { |
| "epoch": 2.4216984178237, |
| "grad_norm": 0.6318864226341248, |
| "learning_rate": 0.0004552332722766943, |
| "loss": 3.5685, |
| "step": 22500 |
| }, |
| { |
| "epoch": 2.4270799698633088, |
| "grad_norm": 0.6047439575195312, |
| "learning_rate": 0.0004549100312466328, |
| "loss": 3.5711, |
| "step": 22550 |
| }, |
| { |
| "epoch": 2.432461521902917, |
| "grad_norm": 0.5744851231575012, |
| "learning_rate": 0.00045459325503717263, |
| "loss": 3.5702, |
| "step": 22600 |
| }, |
| { |
| "epoch": 2.437843073942525, |
| "grad_norm": 0.5994943976402283, |
| "learning_rate": 0.0004542700140071113, |
| "loss": 3.569, |
| "step": 22650 |
| }, |
| { |
| "epoch": 2.443224625982133, |
| "grad_norm": 0.5537005662918091, |
| "learning_rate": 0.0004539467729770499, |
| "loss": 3.5438, |
| "step": 22700 |
| }, |
| { |
| "epoch": 2.4486061780217416, |
| "grad_norm": 0.576801061630249, |
| "learning_rate": 0.0004536235319469884, |
| "loss": 3.5601, |
| "step": 22750 |
| }, |
| { |
| "epoch": 2.4539877300613497, |
| "grad_norm": 0.7851975560188293, |
| "learning_rate": 0.000453300290916927, |
| "loss": 3.5645, |
| "step": 22800 |
| }, |
| { |
| "epoch": 2.459369282100958, |
| "grad_norm": 0.5843862891197205, |
| "learning_rate": 0.0004529770498868656, |
| "loss": 3.5571, |
| "step": 22850 |
| }, |
| { |
| "epoch": 2.464750834140566, |
| "grad_norm": 0.5837774872779846, |
| "learning_rate": 0.00045265380885680414, |
| "loss": 3.5662, |
| "step": 22900 |
| }, |
| { |
| "epoch": 2.4701323861801745, |
| "grad_norm": 0.5811747312545776, |
| "learning_rate": 0.0004523305678267428, |
| "loss": 3.57, |
| "step": 22950 |
| }, |
| { |
| "epoch": 2.4755139382197826, |
| "grad_norm": 0.6275676488876343, |
| "learning_rate": 0.0004520073267966814, |
| "loss": 3.5547, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4755139382197826, |
| "eval_accuracy": 0.3650710953873481, |
| "eval_loss": 3.5427539348602295, |
| "eval_runtime": 184.4438, |
| "eval_samples_per_second": 97.65, |
| "eval_steps_per_second": 6.105, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4808954902593907, |
| "grad_norm": 0.6013798117637634, |
| "learning_rate": 0.00045168408576661993, |
| "loss": 3.5469, |
| "step": 23050 |
| }, |
| { |
| "epoch": 2.4862770422989993, |
| "grad_norm": 0.6493346095085144, |
| "learning_rate": 0.0004513608447365585, |
| "loss": 3.561, |
| "step": 23100 |
| }, |
| { |
| "epoch": 2.4916585943386074, |
| "grad_norm": 0.5985797643661499, |
| "learning_rate": 0.00045103760370649706, |
| "loss": 3.5711, |
| "step": 23150 |
| }, |
| { |
| "epoch": 2.4970401463782155, |
| "grad_norm": 0.6186266541481018, |
| "learning_rate": 0.00045071436267643566, |
| "loss": 3.5492, |
| "step": 23200 |
| }, |
| { |
| "epoch": 2.5024216984178236, |
| "grad_norm": 0.560845136642456, |
| "learning_rate": 0.0004503911216463743, |
| "loss": 3.5565, |
| "step": 23250 |
| }, |
| { |
| "epoch": 2.5078032504574317, |
| "grad_norm": 0.5958519577980042, |
| "learning_rate": 0.00045006788061631285, |
| "loss": 3.568, |
| "step": 23300 |
| }, |
| { |
| "epoch": 2.5131848024970402, |
| "grad_norm": 0.5718041062355042, |
| "learning_rate": 0.00044974463958625144, |
| "loss": 3.5392, |
| "step": 23350 |
| }, |
| { |
| "epoch": 2.5185663545366483, |
| "grad_norm": 0.6036562323570251, |
| "learning_rate": 0.00044942139855619004, |
| "loss": 3.567, |
| "step": 23400 |
| }, |
| { |
| "epoch": 2.5239479065762565, |
| "grad_norm": 0.5902794003486633, |
| "learning_rate": 0.0004490981575261286, |
| "loss": 3.551, |
| "step": 23450 |
| }, |
| { |
| "epoch": 2.529329458615865, |
| "grad_norm": 0.5769206285476685, |
| "learning_rate": 0.0004487749164960672, |
| "loss": 3.557, |
| "step": 23500 |
| }, |
| { |
| "epoch": 2.534711010655473, |
| "grad_norm": 0.5881948471069336, |
| "learning_rate": 0.0004484516754660058, |
| "loss": 3.5472, |
| "step": 23550 |
| }, |
| { |
| "epoch": 2.540092562695081, |
| "grad_norm": 0.5914930105209351, |
| "learning_rate": 0.00044812843443594436, |
| "loss": 3.5434, |
| "step": 23600 |
| }, |
| { |
| "epoch": 2.5454741147346893, |
| "grad_norm": 0.5718581080436707, |
| "learning_rate": 0.00044780519340588296, |
| "loss": 3.559, |
| "step": 23650 |
| }, |
| { |
| "epoch": 2.550855666774298, |
| "grad_norm": 0.5887609720230103, |
| "learning_rate": 0.0004474819523758215, |
| "loss": 3.5548, |
| "step": 23700 |
| }, |
| { |
| "epoch": 2.556237218813906, |
| "grad_norm": 0.57635498046875, |
| "learning_rate": 0.0004471587113457601, |
| "loss": 3.5431, |
| "step": 23750 |
| }, |
| { |
| "epoch": 2.561618770853514, |
| "grad_norm": 0.6003997325897217, |
| "learning_rate": 0.00044683547031569874, |
| "loss": 3.5628, |
| "step": 23800 |
| }, |
| { |
| "epoch": 2.567000322893122, |
| "grad_norm": 0.6113025546073914, |
| "learning_rate": 0.0004465122292856373, |
| "loss": 3.5497, |
| "step": 23850 |
| }, |
| { |
| "epoch": 2.5723818749327307, |
| "grad_norm": 0.5799350738525391, |
| "learning_rate": 0.0004461889882555759, |
| "loss": 3.5676, |
| "step": 23900 |
| }, |
| { |
| "epoch": 2.577763426972339, |
| "grad_norm": 0.5788649916648865, |
| "learning_rate": 0.00044586574722551447, |
| "loss": 3.5609, |
| "step": 23950 |
| }, |
| { |
| "epoch": 2.583144979011947, |
| "grad_norm": 0.5850538611412048, |
| "learning_rate": 0.000445542506195453, |
| "loss": 3.5571, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.583144979011947, |
| "eval_accuracy": 0.3657125820370048, |
| "eval_loss": 3.534817695617676, |
| "eval_runtime": 185.0176, |
| "eval_samples_per_second": 97.347, |
| "eval_steps_per_second": 6.086, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.5885265310515555, |
| "grad_norm": 0.6198786497116089, |
| "learning_rate": 0.0004452192651653916, |
| "loss": 3.5545, |
| "step": 24050 |
| }, |
| { |
| "epoch": 2.5939080830911636, |
| "grad_norm": 0.5845891833305359, |
| "learning_rate": 0.00044489602413533025, |
| "loss": 3.54, |
| "step": 24100 |
| }, |
| { |
| "epoch": 2.5992896351307717, |
| "grad_norm": 0.5993114113807678, |
| "learning_rate": 0.0004445727831052688, |
| "loss": 3.5377, |
| "step": 24150 |
| }, |
| { |
| "epoch": 2.60467118717038, |
| "grad_norm": 0.5982822775840759, |
| "learning_rate": 0.0004442495420752074, |
| "loss": 3.5527, |
| "step": 24200 |
| }, |
| { |
| "epoch": 2.610052739209988, |
| "grad_norm": 0.6251617670059204, |
| "learning_rate": 0.00044392630104514593, |
| "loss": 3.5481, |
| "step": 24250 |
| }, |
| { |
| "epoch": 2.6154342912495965, |
| "grad_norm": 0.617463231086731, |
| "learning_rate": 0.0004436030600150845, |
| "loss": 3.5421, |
| "step": 24300 |
| }, |
| { |
| "epoch": 2.6208158432892046, |
| "grad_norm": 0.5983661413192749, |
| "learning_rate": 0.0004432798189850231, |
| "loss": 3.5651, |
| "step": 24350 |
| }, |
| { |
| "epoch": 2.6261973953288127, |
| "grad_norm": 0.6203315258026123, |
| "learning_rate": 0.0004429565779549617, |
| "loss": 3.5403, |
| "step": 24400 |
| }, |
| { |
| "epoch": 2.6315789473684212, |
| "grad_norm": 0.5848509073257446, |
| "learning_rate": 0.0004426333369249003, |
| "loss": 3.5581, |
| "step": 24450 |
| }, |
| { |
| "epoch": 2.6369604994080293, |
| "grad_norm": 0.5791871547698975, |
| "learning_rate": 0.0004423100958948389, |
| "loss": 3.5359, |
| "step": 24500 |
| }, |
| { |
| "epoch": 2.6423420514476375, |
| "grad_norm": 0.634689211845398, |
| "learning_rate": 0.00044198685486477744, |
| "loss": 3.5436, |
| "step": 24550 |
| }, |
| { |
| "epoch": 2.6477236034872456, |
| "grad_norm": 0.5859078764915466, |
| "learning_rate": 0.00044166361383471604, |
| "loss": 3.5412, |
| "step": 24600 |
| }, |
| { |
| "epoch": 2.653105155526854, |
| "grad_norm": 0.6053293943405151, |
| "learning_rate": 0.0004413403728046547, |
| "loss": 3.5539, |
| "step": 24650 |
| }, |
| { |
| "epoch": 2.658486707566462, |
| "grad_norm": 0.6215761303901672, |
| "learning_rate": 0.0004410171317745932, |
| "loss": 3.5523, |
| "step": 24700 |
| }, |
| { |
| "epoch": 2.6638682596060703, |
| "grad_norm": 0.5783427357673645, |
| "learning_rate": 0.0004406938907445318, |
| "loss": 3.5632, |
| "step": 24750 |
| }, |
| { |
| "epoch": 2.6692498116456784, |
| "grad_norm": 0.6102443933486938, |
| "learning_rate": 0.00044037064971447036, |
| "loss": 3.5519, |
| "step": 24800 |
| }, |
| { |
| "epoch": 2.674631363685287, |
| "grad_norm": 0.61601322889328, |
| "learning_rate": 0.00044004740868440896, |
| "loss": 3.5425, |
| "step": 24850 |
| }, |
| { |
| "epoch": 2.680012915724895, |
| "grad_norm": 0.5947862863540649, |
| "learning_rate": 0.00043972416765434755, |
| "loss": 3.5461, |
| "step": 24900 |
| }, |
| { |
| "epoch": 2.685394467764503, |
| "grad_norm": 0.6124972105026245, |
| "learning_rate": 0.00043940092662428615, |
| "loss": 3.547, |
| "step": 24950 |
| }, |
| { |
| "epoch": 2.6907760198041117, |
| "grad_norm": 0.5721360445022583, |
| "learning_rate": 0.00043907768559422474, |
| "loss": 3.5501, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.6907760198041117, |
| "eval_accuracy": 0.3670556403832796, |
| "eval_loss": 3.5255908966064453, |
| "eval_runtime": 184.7747, |
| "eval_samples_per_second": 97.475, |
| "eval_steps_per_second": 6.094, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.69615757184372, |
| "grad_norm": 0.6443188190460205, |
| "learning_rate": 0.00043875444456416334, |
| "loss": 3.5278, |
| "step": 25050 |
| }, |
| { |
| "epoch": 2.701539123883328, |
| "grad_norm": 0.5788137316703796, |
| "learning_rate": 0.0004384312035341019, |
| "loss": 3.5573, |
| "step": 25100 |
| }, |
| { |
| "epoch": 2.706920675922936, |
| "grad_norm": 0.5494847297668457, |
| "learning_rate": 0.00043810796250404047, |
| "loss": 3.5508, |
| "step": 25150 |
| }, |
| { |
| "epoch": 2.712302227962544, |
| "grad_norm": 0.5972323417663574, |
| "learning_rate": 0.000437784721473979, |
| "loss": 3.535, |
| "step": 25200 |
| }, |
| { |
| "epoch": 2.7176837800021527, |
| "grad_norm": 0.6219172477722168, |
| "learning_rate": 0.00043746148044391766, |
| "loss": 3.5357, |
| "step": 25250 |
| }, |
| { |
| "epoch": 2.723065332041761, |
| "grad_norm": 0.5936007499694824, |
| "learning_rate": 0.00043713823941385625, |
| "loss": 3.5371, |
| "step": 25300 |
| }, |
| { |
| "epoch": 2.728446884081369, |
| "grad_norm": 0.5830941200256348, |
| "learning_rate": 0.0004368149983837948, |
| "loss": 3.5308, |
| "step": 25350 |
| }, |
| { |
| "epoch": 2.7338284361209775, |
| "grad_norm": 0.5860649943351746, |
| "learning_rate": 0.0004364917573537334, |
| "loss": 3.5441, |
| "step": 25400 |
| }, |
| { |
| "epoch": 2.7392099881605856, |
| "grad_norm": 0.6084650158882141, |
| "learning_rate": 0.00043616851632367193, |
| "loss": 3.5439, |
| "step": 25450 |
| }, |
| { |
| "epoch": 2.7445915402001937, |
| "grad_norm": 0.6109269857406616, |
| "learning_rate": 0.0004358452752936106, |
| "loss": 3.5458, |
| "step": 25500 |
| }, |
| { |
| "epoch": 2.749973092239802, |
| "grad_norm": 0.5673308372497559, |
| "learning_rate": 0.0004355220342635492, |
| "loss": 3.545, |
| "step": 25550 |
| }, |
| { |
| "epoch": 2.7553546442794103, |
| "grad_norm": 0.6352481842041016, |
| "learning_rate": 0.00043519879323348777, |
| "loss": 3.5535, |
| "step": 25600 |
| }, |
| { |
| "epoch": 2.7607361963190185, |
| "grad_norm": 0.6171466112136841, |
| "learning_rate": 0.0004348755522034263, |
| "loss": 3.549, |
| "step": 25650 |
| }, |
| { |
| "epoch": 2.7661177483586266, |
| "grad_norm": 0.5649158358573914, |
| "learning_rate": 0.0004345523111733649, |
| "loss": 3.55, |
| "step": 25700 |
| }, |
| { |
| "epoch": 2.7714993003982347, |
| "grad_norm": 0.6140876412391663, |
| "learning_rate": 0.00043422907014330344, |
| "loss": 3.5487, |
| "step": 25750 |
| }, |
| { |
| "epoch": 2.776880852437843, |
| "grad_norm": 0.5735211968421936, |
| "learning_rate": 0.0004339058291132421, |
| "loss": 3.5289, |
| "step": 25800 |
| }, |
| { |
| "epoch": 2.7822624044774513, |
| "grad_norm": 0.596508800983429, |
| "learning_rate": 0.0004335825880831807, |
| "loss": 3.522, |
| "step": 25850 |
| }, |
| { |
| "epoch": 2.7876439565170594, |
| "grad_norm": 0.5887395739555359, |
| "learning_rate": 0.00043325934705311923, |
| "loss": 3.5414, |
| "step": 25900 |
| }, |
| { |
| "epoch": 2.793025508556668, |
| "grad_norm": 0.6502258777618408, |
| "learning_rate": 0.0004329361060230578, |
| "loss": 3.5267, |
| "step": 25950 |
| }, |
| { |
| "epoch": 2.798407060596276, |
| "grad_norm": 0.5818353295326233, |
| "learning_rate": 0.00043261286499299636, |
| "loss": 3.5323, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.798407060596276, |
| "eval_accuracy": 0.3677229864202361, |
| "eval_loss": 3.5174596309661865, |
| "eval_runtime": 185.1441, |
| "eval_samples_per_second": 97.281, |
| "eval_steps_per_second": 6.082, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.803788612635884, |
| "grad_norm": 0.641747236251831, |
| "learning_rate": 0.00043228962396293496, |
| "loss": 3.5431, |
| "step": 26050 |
| }, |
| { |
| "epoch": 2.8091701646754923, |
| "grad_norm": 0.6401011943817139, |
| "learning_rate": 0.0004319663829328736, |
| "loss": 3.5387, |
| "step": 26100 |
| }, |
| { |
| "epoch": 2.8145517167151004, |
| "grad_norm": 0.6039129495620728, |
| "learning_rate": 0.00043164314190281215, |
| "loss": 3.5148, |
| "step": 26150 |
| }, |
| { |
| "epoch": 2.819933268754709, |
| "grad_norm": 0.5612286925315857, |
| "learning_rate": 0.00043131990087275074, |
| "loss": 3.5341, |
| "step": 26200 |
| }, |
| { |
| "epoch": 2.825314820794317, |
| "grad_norm": 0.5970532894134521, |
| "learning_rate": 0.00043099665984268934, |
| "loss": 3.5381, |
| "step": 26250 |
| }, |
| { |
| "epoch": 2.830696372833925, |
| "grad_norm": 0.5767912864685059, |
| "learning_rate": 0.0004306734188126279, |
| "loss": 3.5239, |
| "step": 26300 |
| }, |
| { |
| "epoch": 2.8360779248735337, |
| "grad_norm": 0.5821993350982666, |
| "learning_rate": 0.00043035017778256647, |
| "loss": 3.5205, |
| "step": 26350 |
| }, |
| { |
| "epoch": 2.841459476913142, |
| "grad_norm": 0.6017364263534546, |
| "learning_rate": 0.0004300269367525051, |
| "loss": 3.538, |
| "step": 26400 |
| }, |
| { |
| "epoch": 2.84684102895275, |
| "grad_norm": 0.5902091860771179, |
| "learning_rate": 0.00042970369572244366, |
| "loss": 3.5355, |
| "step": 26450 |
| }, |
| { |
| "epoch": 2.852222580992358, |
| "grad_norm": 0.6088672280311584, |
| "learning_rate": 0.00042938045469238226, |
| "loss": 3.5265, |
| "step": 26500 |
| }, |
| { |
| "epoch": 2.857604133031966, |
| "grad_norm": 0.5895884037017822, |
| "learning_rate": 0.00042906367848292206, |
| "loss": 3.5467, |
| "step": 26550 |
| }, |
| { |
| "epoch": 2.8629856850715747, |
| "grad_norm": 0.5715399980545044, |
| "learning_rate": 0.00042874043745286066, |
| "loss": 3.5379, |
| "step": 26600 |
| }, |
| { |
| "epoch": 2.868367237111183, |
| "grad_norm": 0.6140602231025696, |
| "learning_rate": 0.0004284171964227992, |
| "loss": 3.539, |
| "step": 26650 |
| }, |
| { |
| "epoch": 2.873748789150791, |
| "grad_norm": 0.5857102274894714, |
| "learning_rate": 0.0004280939553927378, |
| "loss": 3.5439, |
| "step": 26700 |
| }, |
| { |
| "epoch": 2.8791303411903995, |
| "grad_norm": 0.6274014711380005, |
| "learning_rate": 0.00042777071436267644, |
| "loss": 3.5369, |
| "step": 26750 |
| }, |
| { |
| "epoch": 2.8845118932300076, |
| "grad_norm": 0.5715806484222412, |
| "learning_rate": 0.000427447473332615, |
| "loss": 3.5394, |
| "step": 26800 |
| }, |
| { |
| "epoch": 2.8898934452696157, |
| "grad_norm": 0.5876508951187134, |
| "learning_rate": 0.0004271242323025536, |
| "loss": 3.5329, |
| "step": 26850 |
| }, |
| { |
| "epoch": 2.895274997309224, |
| "grad_norm": 0.5981003642082214, |
| "learning_rate": 0.00042680099127249217, |
| "loss": 3.5231, |
| "step": 26900 |
| }, |
| { |
| "epoch": 2.9006565493488323, |
| "grad_norm": 0.5943262577056885, |
| "learning_rate": 0.0004264777502424307, |
| "loss": 3.538, |
| "step": 26950 |
| }, |
| { |
| "epoch": 2.9060381013884404, |
| "grad_norm": 0.8639909029006958, |
| "learning_rate": 0.0004261545092123693, |
| "loss": 3.536, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9060381013884404, |
| "eval_accuracy": 0.3691632891011049, |
| "eval_loss": 3.5098328590393066, |
| "eval_runtime": 184.886, |
| "eval_samples_per_second": 97.417, |
| "eval_steps_per_second": 6.09, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9114196534280485, |
| "grad_norm": 0.5836945176124573, |
| "learning_rate": 0.00042583126818230795, |
| "loss": 3.5395, |
| "step": 27050 |
| }, |
| { |
| "epoch": 2.9168012054676566, |
| "grad_norm": 0.6182237267494202, |
| "learning_rate": 0.0004255080271522465, |
| "loss": 3.5359, |
| "step": 27100 |
| }, |
| { |
| "epoch": 2.922182757507265, |
| "grad_norm": 0.6002602577209473, |
| "learning_rate": 0.0004251847861221851, |
| "loss": 3.5197, |
| "step": 27150 |
| }, |
| { |
| "epoch": 2.9275643095468733, |
| "grad_norm": 0.5895917415618896, |
| "learning_rate": 0.00042486154509212363, |
| "loss": 3.5076, |
| "step": 27200 |
| }, |
| { |
| "epoch": 2.9329458615864814, |
| "grad_norm": 0.5710264444351196, |
| "learning_rate": 0.0004245383040620622, |
| "loss": 3.5327, |
| "step": 27250 |
| }, |
| { |
| "epoch": 2.93832741362609, |
| "grad_norm": 0.5726727247238159, |
| "learning_rate": 0.0004242150630320009, |
| "loss": 3.5243, |
| "step": 27300 |
| }, |
| { |
| "epoch": 2.943708965665698, |
| "grad_norm": 0.5930039286613464, |
| "learning_rate": 0.0004238918220019394, |
| "loss": 3.5362, |
| "step": 27350 |
| }, |
| { |
| "epoch": 2.949090517705306, |
| "grad_norm": 0.54901522397995, |
| "learning_rate": 0.000423568580971878, |
| "loss": 3.5241, |
| "step": 27400 |
| }, |
| { |
| "epoch": 2.9544720697449143, |
| "grad_norm": 0.6401501893997192, |
| "learning_rate": 0.00042324533994181655, |
| "loss": 3.5304, |
| "step": 27450 |
| }, |
| { |
| "epoch": 2.9598536217845224, |
| "grad_norm": 0.5905794501304626, |
| "learning_rate": 0.00042292209891175514, |
| "loss": 3.5298, |
| "step": 27500 |
| }, |
| { |
| "epoch": 2.965235173824131, |
| "grad_norm": 0.5865811109542847, |
| "learning_rate": 0.00042259885788169374, |
| "loss": 3.5374, |
| "step": 27550 |
| }, |
| { |
| "epoch": 2.970616725863739, |
| "grad_norm": 0.5635622143745422, |
| "learning_rate": 0.00042227561685163233, |
| "loss": 3.5246, |
| "step": 27600 |
| }, |
| { |
| "epoch": 2.975998277903347, |
| "grad_norm": 0.5960383415222168, |
| "learning_rate": 0.00042195237582157093, |
| "loss": 3.5413, |
| "step": 27650 |
| }, |
| { |
| "epoch": 2.9813798299429557, |
| "grad_norm": 0.6016942262649536, |
| "learning_rate": 0.0004216291347915095, |
| "loss": 3.5191, |
| "step": 27700 |
| }, |
| { |
| "epoch": 2.986761381982564, |
| "grad_norm": 0.56882244348526, |
| "learning_rate": 0.00042130589376144806, |
| "loss": 3.5143, |
| "step": 27750 |
| }, |
| { |
| "epoch": 2.992142934022172, |
| "grad_norm": 0.589320182800293, |
| "learning_rate": 0.00042098265273138666, |
| "loss": 3.5266, |
| "step": 27800 |
| }, |
| { |
| "epoch": 2.9975244860617805, |
| "grad_norm": 0.6223992109298706, |
| "learning_rate": 0.0004206594117013252, |
| "loss": 3.5296, |
| "step": 27850 |
| }, |
| { |
| "epoch": 3.0029060381013886, |
| "grad_norm": 0.5933626294136047, |
| "learning_rate": 0.00042033617067126385, |
| "loss": 3.4722, |
| "step": 27900 |
| }, |
| { |
| "epoch": 3.0082875901409967, |
| "grad_norm": 0.643946647644043, |
| "learning_rate": 0.00042001292964120244, |
| "loss": 3.4474, |
| "step": 27950 |
| }, |
| { |
| "epoch": 3.0136691421806048, |
| "grad_norm": 0.6281617283821106, |
| "learning_rate": 0.000419689688611141, |
| "loss": 3.4438, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.0136691421806048, |
| "eval_accuracy": 0.36952868876282496, |
| "eval_loss": 3.505746841430664, |
| "eval_runtime": 185.2261, |
| "eval_samples_per_second": 97.238, |
| "eval_steps_per_second": 6.079, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.0190506942202133, |
| "grad_norm": 0.5724912881851196, |
| "learning_rate": 0.0004193664475810796, |
| "loss": 3.4348, |
| "step": 28050 |
| }, |
| { |
| "epoch": 3.0244322462598214, |
| "grad_norm": 0.5964477062225342, |
| "learning_rate": 0.00041904320655101817, |
| "loss": 3.4444, |
| "step": 28100 |
| }, |
| { |
| "epoch": 3.0298137982994295, |
| "grad_norm": 0.6437315344810486, |
| "learning_rate": 0.0004187199655209567, |
| "loss": 3.4424, |
| "step": 28150 |
| }, |
| { |
| "epoch": 3.0351953503390376, |
| "grad_norm": 0.7275903820991516, |
| "learning_rate": 0.00041839672449089536, |
| "loss": 3.4293, |
| "step": 28200 |
| }, |
| { |
| "epoch": 3.040576902378646, |
| "grad_norm": 0.5967044830322266, |
| "learning_rate": 0.00041807348346083395, |
| "loss": 3.4381, |
| "step": 28250 |
| }, |
| { |
| "epoch": 3.0459584544182543, |
| "grad_norm": 0.6012422442436218, |
| "learning_rate": 0.0004177502424307725, |
| "loss": 3.4267, |
| "step": 28300 |
| }, |
| { |
| "epoch": 3.0513400064578624, |
| "grad_norm": 0.6100385785102844, |
| "learning_rate": 0.0004174270014007111, |
| "loss": 3.4447, |
| "step": 28350 |
| }, |
| { |
| "epoch": 3.0567215584974705, |
| "grad_norm": 0.6521284580230713, |
| "learning_rate": 0.00041710376037064963, |
| "loss": 3.4316, |
| "step": 28400 |
| }, |
| { |
| "epoch": 3.062103110537079, |
| "grad_norm": 0.5814330577850342, |
| "learning_rate": 0.0004167805193405883, |
| "loss": 3.4391, |
| "step": 28450 |
| }, |
| { |
| "epoch": 3.067484662576687, |
| "grad_norm": 0.6042176485061646, |
| "learning_rate": 0.0004164572783105269, |
| "loss": 3.4447, |
| "step": 28500 |
| }, |
| { |
| "epoch": 3.0728662146162953, |
| "grad_norm": 0.6039173603057861, |
| "learning_rate": 0.0004161340372804654, |
| "loss": 3.4407, |
| "step": 28550 |
| }, |
| { |
| "epoch": 3.0782477666559034, |
| "grad_norm": 0.624396800994873, |
| "learning_rate": 0.000415810796250404, |
| "loss": 3.4477, |
| "step": 28600 |
| }, |
| { |
| "epoch": 3.083629318695512, |
| "grad_norm": 0.6126594543457031, |
| "learning_rate": 0.0004154875552203426, |
| "loss": 3.4456, |
| "step": 28650 |
| }, |
| { |
| "epoch": 3.08901087073512, |
| "grad_norm": 0.6176913380622864, |
| "learning_rate": 0.00041516431419028114, |
| "loss": 3.4701, |
| "step": 28700 |
| }, |
| { |
| "epoch": 3.094392422774728, |
| "grad_norm": 0.624182403087616, |
| "learning_rate": 0.0004148410731602198, |
| "loss": 3.4544, |
| "step": 28750 |
| }, |
| { |
| "epoch": 3.0997739748143363, |
| "grad_norm": 0.5999160408973694, |
| "learning_rate": 0.0004145178321301584, |
| "loss": 3.4638, |
| "step": 28800 |
| }, |
| { |
| "epoch": 3.105155526853945, |
| "grad_norm": 0.5691066980361938, |
| "learning_rate": 0.00041419459110009693, |
| "loss": 3.4382, |
| "step": 28850 |
| }, |
| { |
| "epoch": 3.110537078893553, |
| "grad_norm": 0.5719393491744995, |
| "learning_rate": 0.0004138713500700355, |
| "loss": 3.4553, |
| "step": 28900 |
| }, |
| { |
| "epoch": 3.115918630933161, |
| "grad_norm": 0.585585355758667, |
| "learning_rate": 0.00041354810903997406, |
| "loss": 3.4603, |
| "step": 28950 |
| }, |
| { |
| "epoch": 3.121300182972769, |
| "grad_norm": 0.5965518355369568, |
| "learning_rate": 0.00041322486800991266, |
| "loss": 3.4492, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.121300182972769, |
| "eval_accuracy": 0.3698451946250286, |
| "eval_loss": 3.5004117488861084, |
| "eval_runtime": 184.5771, |
| "eval_samples_per_second": 97.58, |
| "eval_steps_per_second": 6.1, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.1266817350123777, |
| "grad_norm": 0.559124231338501, |
| "learning_rate": 0.0004129016269798513, |
| "loss": 3.4561, |
| "step": 29050 |
| }, |
| { |
| "epoch": 3.132063287051986, |
| "grad_norm": 0.6254011392593384, |
| "learning_rate": 0.00041257838594978985, |
| "loss": 3.4717, |
| "step": 29100 |
| }, |
| { |
| "epoch": 3.137444839091594, |
| "grad_norm": 0.5621007084846497, |
| "learning_rate": 0.00041225514491972844, |
| "loss": 3.458, |
| "step": 29150 |
| }, |
| { |
| "epoch": 3.1428263911312024, |
| "grad_norm": 0.60045325756073, |
| "learning_rate": 0.00041193190388966704, |
| "loss": 3.4385, |
| "step": 29200 |
| }, |
| { |
| "epoch": 3.1482079431708105, |
| "grad_norm": 0.5860231518745422, |
| "learning_rate": 0.0004116086628596056, |
| "loss": 3.459, |
| "step": 29250 |
| }, |
| { |
| "epoch": 3.1535894952104186, |
| "grad_norm": 0.5977693796157837, |
| "learning_rate": 0.0004112854218295442, |
| "loss": 3.4527, |
| "step": 29300 |
| }, |
| { |
| "epoch": 3.1589710472500268, |
| "grad_norm": 0.6494083404541016, |
| "learning_rate": 0.0004109621807994828, |
| "loss": 3.4596, |
| "step": 29350 |
| }, |
| { |
| "epoch": 3.1643525992896353, |
| "grad_norm": 0.6031472086906433, |
| "learning_rate": 0.00041063893976942136, |
| "loss": 3.4657, |
| "step": 29400 |
| }, |
| { |
| "epoch": 3.1697341513292434, |
| "grad_norm": 0.6102636456489563, |
| "learning_rate": 0.00041031569873935996, |
| "loss": 3.4436, |
| "step": 29450 |
| }, |
| { |
| "epoch": 3.1751157033688515, |
| "grad_norm": 0.6388329863548279, |
| "learning_rate": 0.0004099924577092985, |
| "loss": 3.4628, |
| "step": 29500 |
| }, |
| { |
| "epoch": 3.1804972554084596, |
| "grad_norm": 0.5955821871757507, |
| "learning_rate": 0.0004096692166792371, |
| "loss": 3.4558, |
| "step": 29550 |
| }, |
| { |
| "epoch": 3.185878807448068, |
| "grad_norm": 0.5899354815483093, |
| "learning_rate": 0.00040934597564917574, |
| "loss": 3.4744, |
| "step": 29600 |
| }, |
| { |
| "epoch": 3.1912603594876763, |
| "grad_norm": 0.6263838410377502, |
| "learning_rate": 0.0004090227346191143, |
| "loss": 3.4617, |
| "step": 29650 |
| }, |
| { |
| "epoch": 3.1966419115272844, |
| "grad_norm": 0.6136727333068848, |
| "learning_rate": 0.0004086994935890529, |
| "loss": 3.462, |
| "step": 29700 |
| }, |
| { |
| "epoch": 3.2020234635668925, |
| "grad_norm": 0.6539897918701172, |
| "learning_rate": 0.00040837625255899147, |
| "loss": 3.4374, |
| "step": 29750 |
| }, |
| { |
| "epoch": 3.207405015606501, |
| "grad_norm": 0.643663763999939, |
| "learning_rate": 0.00040805301152893, |
| "loss": 3.4465, |
| "step": 29800 |
| }, |
| { |
| "epoch": 3.212786567646109, |
| "grad_norm": 0.6039734482765198, |
| "learning_rate": 0.0004077297704988686, |
| "loss": 3.4573, |
| "step": 29850 |
| }, |
| { |
| "epoch": 3.2181681196857173, |
| "grad_norm": 0.6215190291404724, |
| "learning_rate": 0.00040740652946880725, |
| "loss": 3.4581, |
| "step": 29900 |
| }, |
| { |
| "epoch": 3.2235496717253254, |
| "grad_norm": 0.6273991465568542, |
| "learning_rate": 0.0004070832884387458, |
| "loss": 3.4525, |
| "step": 29950 |
| }, |
| { |
| "epoch": 3.228931223764934, |
| "grad_norm": 0.6081337928771973, |
| "learning_rate": 0.0004067600474086844, |
| "loss": 3.4644, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.228931223764934, |
| "eval_accuracy": 0.3708994535954925, |
| "eval_loss": 3.497182607650757, |
| "eval_runtime": 185.0908, |
| "eval_samples_per_second": 97.309, |
| "eval_steps_per_second": 6.084, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.234312775804542, |
| "grad_norm": 0.7007036805152893, |
| "learning_rate": 0.00040643680637862293, |
| "loss": 3.4624, |
| "step": 30050 |
| }, |
| { |
| "epoch": 3.23969432784415, |
| "grad_norm": 0.5929351449012756, |
| "learning_rate": 0.0004061135653485615, |
| "loss": 3.4567, |
| "step": 30100 |
| }, |
| { |
| "epoch": 3.2450758798837587, |
| "grad_norm": 0.6256581544876099, |
| "learning_rate": 0.0004057903243185001, |
| "loss": 3.4358, |
| "step": 30150 |
| }, |
| { |
| "epoch": 3.250457431923367, |
| "grad_norm": 0.5989969968795776, |
| "learning_rate": 0.0004054670832884387, |
| "loss": 3.436, |
| "step": 30200 |
| }, |
| { |
| "epoch": 3.255838983962975, |
| "grad_norm": 0.623378574848175, |
| "learning_rate": 0.0004051438422583773, |
| "loss": 3.4733, |
| "step": 30250 |
| }, |
| { |
| "epoch": 3.261220536002583, |
| "grad_norm": 0.6182704567909241, |
| "learning_rate": 0.0004048206012283159, |
| "loss": 3.4613, |
| "step": 30300 |
| }, |
| { |
| "epoch": 3.2666020880421915, |
| "grad_norm": 0.6270396113395691, |
| "learning_rate": 0.00040449736019825444, |
| "loss": 3.4558, |
| "step": 30350 |
| }, |
| { |
| "epoch": 3.2719836400817996, |
| "grad_norm": 0.6733869314193726, |
| "learning_rate": 0.00040417411916819304, |
| "loss": 3.4665, |
| "step": 30400 |
| }, |
| { |
| "epoch": 3.2773651921214078, |
| "grad_norm": 0.6124371886253357, |
| "learning_rate": 0.0004038508781381317, |
| "loss": 3.4659, |
| "step": 30450 |
| }, |
| { |
| "epoch": 3.282746744161016, |
| "grad_norm": 0.6001761555671692, |
| "learning_rate": 0.0004035276371080702, |
| "loss": 3.4618, |
| "step": 30500 |
| }, |
| { |
| "epoch": 3.2881282962006244, |
| "grad_norm": 0.6276858448982239, |
| "learning_rate": 0.00040321086089861003, |
| "loss": 3.4604, |
| "step": 30550 |
| }, |
| { |
| "epoch": 3.2935098482402325, |
| "grad_norm": 0.5796586275100708, |
| "learning_rate": 0.00040288761986854863, |
| "loss": 3.4587, |
| "step": 30600 |
| }, |
| { |
| "epoch": 3.2988914002798406, |
| "grad_norm": 0.6277230978012085, |
| "learning_rate": 0.0004025643788384872, |
| "loss": 3.4438, |
| "step": 30650 |
| }, |
| { |
| "epoch": 3.304272952319449, |
| "grad_norm": 0.6220398545265198, |
| "learning_rate": 0.00040224113780842576, |
| "loss": 3.4767, |
| "step": 30700 |
| }, |
| { |
| "epoch": 3.3096545043590573, |
| "grad_norm": 0.6023754477500916, |
| "learning_rate": 0.00040191789677836436, |
| "loss": 3.4706, |
| "step": 30750 |
| }, |
| { |
| "epoch": 3.3150360563986654, |
| "grad_norm": 0.6297191977500916, |
| "learning_rate": 0.0004015946557483029, |
| "loss": 3.447, |
| "step": 30800 |
| }, |
| { |
| "epoch": 3.3204176084382735, |
| "grad_norm": 0.597350537776947, |
| "learning_rate": 0.00040127141471824155, |
| "loss": 3.4861, |
| "step": 30850 |
| }, |
| { |
| "epoch": 3.3257991604778816, |
| "grad_norm": 0.6226769685745239, |
| "learning_rate": 0.00040094817368818014, |
| "loss": 3.4636, |
| "step": 30900 |
| }, |
| { |
| "epoch": 3.33118071251749, |
| "grad_norm": 0.5647996664047241, |
| "learning_rate": 0.0004006249326581187, |
| "loss": 3.4799, |
| "step": 30950 |
| }, |
| { |
| "epoch": 3.3365622645570983, |
| "grad_norm": 0.6258048415184021, |
| "learning_rate": 0.0004003016916280573, |
| "loss": 3.4772, |
| "step": 31000 |
| }, |
| { |
| "epoch": 3.3365622645570983, |
| "eval_accuracy": 0.3716034156556425, |
| "eval_loss": 3.4863648414611816, |
| "eval_runtime": 185.2474, |
| "eval_samples_per_second": 97.227, |
| "eval_steps_per_second": 6.078, |
| "step": 31000 |
| }, |
| { |
| "epoch": 3.3419438165967064, |
| "grad_norm": 0.617977499961853, |
| "learning_rate": 0.00039997845059799587, |
| "loss": 3.4577, |
| "step": 31050 |
| }, |
| { |
| "epoch": 3.347325368636315, |
| "grad_norm": 0.6409074664115906, |
| "learning_rate": 0.00039965520956793447, |
| "loss": 3.4636, |
| "step": 31100 |
| }, |
| { |
| "epoch": 3.352706920675923, |
| "grad_norm": 0.5829642415046692, |
| "learning_rate": 0.00039933196853787306, |
| "loss": 3.4648, |
| "step": 31150 |
| }, |
| { |
| "epoch": 3.358088472715531, |
| "grad_norm": 0.5979641675949097, |
| "learning_rate": 0.00039900872750781166, |
| "loss": 3.4759, |
| "step": 31200 |
| }, |
| { |
| "epoch": 3.3634700247551392, |
| "grad_norm": 0.6275162100791931, |
| "learning_rate": 0.0003986854864777502, |
| "loss": 3.4603, |
| "step": 31250 |
| }, |
| { |
| "epoch": 3.368851576794748, |
| "grad_norm": 0.675043523311615, |
| "learning_rate": 0.0003983622454476888, |
| "loss": 3.4723, |
| "step": 31300 |
| }, |
| { |
| "epoch": 3.374233128834356, |
| "grad_norm": 0.6001876592636108, |
| "learning_rate": 0.00039803900441762733, |
| "loss": 3.4641, |
| "step": 31350 |
| }, |
| { |
| "epoch": 3.379614680873964, |
| "grad_norm": 0.7512485384941101, |
| "learning_rate": 0.000397715763387566, |
| "loss": 3.4683, |
| "step": 31400 |
| }, |
| { |
| "epoch": 3.384996232913572, |
| "grad_norm": 0.6349093914031982, |
| "learning_rate": 0.0003973925223575046, |
| "loss": 3.441, |
| "step": 31450 |
| }, |
| { |
| "epoch": 3.3903777849531807, |
| "grad_norm": 0.6452442407608032, |
| "learning_rate": 0.0003970692813274431, |
| "loss": 3.4758, |
| "step": 31500 |
| }, |
| { |
| "epoch": 3.3957593369927888, |
| "grad_norm": 0.8671071529388428, |
| "learning_rate": 0.0003967460402973817, |
| "loss": 3.4781, |
| "step": 31550 |
| }, |
| { |
| "epoch": 3.401140889032397, |
| "grad_norm": 0.6597549319267273, |
| "learning_rate": 0.0003964227992673203, |
| "loss": 3.4631, |
| "step": 31600 |
| }, |
| { |
| "epoch": 3.4065224410720054, |
| "grad_norm": 0.6139020919799805, |
| "learning_rate": 0.00039609955823725884, |
| "loss": 3.4709, |
| "step": 31650 |
| }, |
| { |
| "epoch": 3.4119039931116135, |
| "grad_norm": 0.6076966524124146, |
| "learning_rate": 0.0003957763172071975, |
| "loss": 3.4493, |
| "step": 31700 |
| }, |
| { |
| "epoch": 3.4172855451512216, |
| "grad_norm": 0.6119977831840515, |
| "learning_rate": 0.0003954530761771361, |
| "loss": 3.4535, |
| "step": 31750 |
| }, |
| { |
| "epoch": 3.4226670971908297, |
| "grad_norm": 0.6142314672470093, |
| "learning_rate": 0.00039512983514707463, |
| "loss": 3.4606, |
| "step": 31800 |
| }, |
| { |
| "epoch": 3.428048649230438, |
| "grad_norm": 0.6000369191169739, |
| "learning_rate": 0.0003948065941170132, |
| "loss": 3.4516, |
| "step": 31850 |
| }, |
| { |
| "epoch": 3.4334302012700464, |
| "grad_norm": 0.5796836018562317, |
| "learning_rate": 0.00039448335308695176, |
| "loss": 3.4648, |
| "step": 31900 |
| }, |
| { |
| "epoch": 3.4388117533096545, |
| "grad_norm": 0.5938079357147217, |
| "learning_rate": 0.00039416011205689036, |
| "loss": 3.4688, |
| "step": 31950 |
| }, |
| { |
| "epoch": 3.4441933053492626, |
| "grad_norm": 0.6108924150466919, |
| "learning_rate": 0.000393836871026829, |
| "loss": 3.4697, |
| "step": 32000 |
| }, |
| { |
| "epoch": 3.4441933053492626, |
| "eval_accuracy": 0.3729599469600063, |
| "eval_loss": 3.4790189266204834, |
| "eval_runtime": 184.7239, |
| "eval_samples_per_second": 97.502, |
| "eval_steps_per_second": 6.096, |
| "step": 32000 |
| }, |
| { |
| "epoch": 3.449574857388871, |
| "grad_norm": 0.6282352209091187, |
| "learning_rate": 0.00039351362999676755, |
| "loss": 3.4755, |
| "step": 32050 |
| }, |
| { |
| "epoch": 3.4549564094284793, |
| "grad_norm": 0.5978196859359741, |
| "learning_rate": 0.00039319038896670614, |
| "loss": 3.4675, |
| "step": 32100 |
| }, |
| { |
| "epoch": 3.4603379614680874, |
| "grad_norm": 0.603289008140564, |
| "learning_rate": 0.00039286714793664474, |
| "loss": 3.4656, |
| "step": 32150 |
| }, |
| { |
| "epoch": 3.4657195135076955, |
| "grad_norm": 0.6244550943374634, |
| "learning_rate": 0.0003925439069065833, |
| "loss": 3.439, |
| "step": 32200 |
| }, |
| { |
| "epoch": 3.471101065547304, |
| "grad_norm": 0.6003792881965637, |
| "learning_rate": 0.0003922206658765219, |
| "loss": 3.4448, |
| "step": 32250 |
| }, |
| { |
| "epoch": 3.476482617586912, |
| "grad_norm": 0.5921474099159241, |
| "learning_rate": 0.0003918974248464605, |
| "loss": 3.464, |
| "step": 32300 |
| }, |
| { |
| "epoch": 3.4818641696265202, |
| "grad_norm": 0.5765640735626221, |
| "learning_rate": 0.00039157418381639906, |
| "loss": 3.4772, |
| "step": 32350 |
| }, |
| { |
| "epoch": 3.4872457216661283, |
| "grad_norm": 0.6701759099960327, |
| "learning_rate": 0.00039125094278633766, |
| "loss": 3.4741, |
| "step": 32400 |
| }, |
| { |
| "epoch": 3.492627273705737, |
| "grad_norm": 0.6599100828170776, |
| "learning_rate": 0.0003909277017562762, |
| "loss": 3.4774, |
| "step": 32450 |
| }, |
| { |
| "epoch": 3.498008825745345, |
| "grad_norm": 0.5839077830314636, |
| "learning_rate": 0.0003906044607262148, |
| "loss": 3.4744, |
| "step": 32500 |
| }, |
| { |
| "epoch": 3.503390377784953, |
| "grad_norm": 0.6445351839065552, |
| "learning_rate": 0.0003902876845167546, |
| "loss": 3.467, |
| "step": 32550 |
| }, |
| { |
| "epoch": 3.5087719298245617, |
| "grad_norm": 0.6389232277870178, |
| "learning_rate": 0.0003899644434866932, |
| "loss": 3.4601, |
| "step": 32600 |
| }, |
| { |
| "epoch": 3.5141534818641698, |
| "grad_norm": 0.6152642965316772, |
| "learning_rate": 0.00038964120245663184, |
| "loss": 3.46, |
| "step": 32650 |
| }, |
| { |
| "epoch": 3.519535033903778, |
| "grad_norm": 0.6026628613471985, |
| "learning_rate": 0.0003893179614265704, |
| "loss": 3.4665, |
| "step": 32700 |
| }, |
| { |
| "epoch": 3.524916585943386, |
| "grad_norm": 0.658417820930481, |
| "learning_rate": 0.000388994720396509, |
| "loss": 3.4735, |
| "step": 32750 |
| }, |
| { |
| "epoch": 3.530298137982994, |
| "grad_norm": 0.6930371522903442, |
| "learning_rate": 0.0003886714793664475, |
| "loss": 3.4673, |
| "step": 32800 |
| }, |
| { |
| "epoch": 3.5356796900226026, |
| "grad_norm": 0.6179644465446472, |
| "learning_rate": 0.0003883482383363861, |
| "loss": 3.4923, |
| "step": 32850 |
| }, |
| { |
| "epoch": 3.5410612420622107, |
| "grad_norm": 0.6348962783813477, |
| "learning_rate": 0.00038802499730632476, |
| "loss": 3.4828, |
| "step": 32900 |
| }, |
| { |
| "epoch": 3.546442794101819, |
| "grad_norm": 0.6657530665397644, |
| "learning_rate": 0.0003877082210968645, |
| "loss": 3.4599, |
| "step": 32950 |
| }, |
| { |
| "epoch": 3.5518243461414274, |
| "grad_norm": 0.6153150200843811, |
| "learning_rate": 0.00038738498006680316, |
| "loss": 3.4765, |
| "step": 33000 |
| }, |
| { |
| "epoch": 3.5518243461414274, |
| "eval_accuracy": 0.37310097840838946, |
| "eval_loss": 3.472996473312378, |
| "eval_runtime": 184.7663, |
| "eval_samples_per_second": 97.48, |
| "eval_steps_per_second": 6.094, |
| "step": 33000 |
| }, |
| { |
| "epoch": 3.5572058981810355, |
| "grad_norm": 0.621093213558197, |
| "learning_rate": 0.0003870617390367417, |
| "loss": 3.4507, |
| "step": 33050 |
| }, |
| { |
| "epoch": 3.5625874502206436, |
| "grad_norm": 0.6191295981407166, |
| "learning_rate": 0.0003867384980066803, |
| "loss": 3.4579, |
| "step": 33100 |
| }, |
| { |
| "epoch": 3.5679690022602517, |
| "grad_norm": 0.6550074815750122, |
| "learning_rate": 0.00038641525697661884, |
| "loss": 3.4754, |
| "step": 33150 |
| }, |
| { |
| "epoch": 3.57335055429986, |
| "grad_norm": 0.6438351273536682, |
| "learning_rate": 0.00038609201594655743, |
| "loss": 3.4556, |
| "step": 33200 |
| }, |
| { |
| "epoch": 3.5787321063394684, |
| "grad_norm": 0.6011756062507629, |
| "learning_rate": 0.000385768774916496, |
| "loss": 3.4769, |
| "step": 33250 |
| }, |
| { |
| "epoch": 3.5841136583790765, |
| "grad_norm": 0.5651047825813293, |
| "learning_rate": 0.0003854455338864346, |
| "loss": 3.4487, |
| "step": 33300 |
| }, |
| { |
| "epoch": 3.5894952104186846, |
| "grad_norm": 0.6038775444030762, |
| "learning_rate": 0.0003851222928563732, |
| "loss": 3.4777, |
| "step": 33350 |
| }, |
| { |
| "epoch": 3.594876762458293, |
| "grad_norm": 0.6581705212593079, |
| "learning_rate": 0.0003847990518263118, |
| "loss": 3.4427, |
| "step": 33400 |
| }, |
| { |
| "epoch": 3.6002583144979012, |
| "grad_norm": 0.61561119556427, |
| "learning_rate": 0.00038447581079625035, |
| "loss": 3.476, |
| "step": 33450 |
| }, |
| { |
| "epoch": 3.6056398665375093, |
| "grad_norm": 0.6354854702949524, |
| "learning_rate": 0.00038415256976618895, |
| "loss": 3.4703, |
| "step": 33500 |
| }, |
| { |
| "epoch": 3.611021418577118, |
| "grad_norm": 0.6150094866752625, |
| "learning_rate": 0.0003838293287361275, |
| "loss": 3.4711, |
| "step": 33550 |
| }, |
| { |
| "epoch": 3.616402970616726, |
| "grad_norm": 0.611519455909729, |
| "learning_rate": 0.00038350608770606613, |
| "loss": 3.4708, |
| "step": 33600 |
| }, |
| { |
| "epoch": 3.621784522656334, |
| "grad_norm": 0.6256572604179382, |
| "learning_rate": 0.00038318284667600473, |
| "loss": 3.4477, |
| "step": 33650 |
| }, |
| { |
| "epoch": 3.627166074695942, |
| "grad_norm": 0.6752513647079468, |
| "learning_rate": 0.00038285960564594327, |
| "loss": 3.4523, |
| "step": 33700 |
| }, |
| { |
| "epoch": 3.6325476267355503, |
| "grad_norm": 0.6451570391654968, |
| "learning_rate": 0.00038253636461588186, |
| "loss": 3.4629, |
| "step": 33750 |
| }, |
| { |
| "epoch": 3.637929178775159, |
| "grad_norm": 0.6020840406417847, |
| "learning_rate": 0.00038221312358582046, |
| "loss": 3.4861, |
| "step": 33800 |
| }, |
| { |
| "epoch": 3.643310730814767, |
| "grad_norm": 0.5956219434738159, |
| "learning_rate": 0.00038188988255575905, |
| "loss": 3.4752, |
| "step": 33850 |
| }, |
| { |
| "epoch": 3.648692282854375, |
| "grad_norm": 0.6432216763496399, |
| "learning_rate": 0.00038156664152569765, |
| "loss": 3.4618, |
| "step": 33900 |
| }, |
| { |
| "epoch": 3.6540738348939836, |
| "grad_norm": 0.6001318097114563, |
| "learning_rate": 0.00038124340049563624, |
| "loss": 3.4456, |
| "step": 33950 |
| }, |
| { |
| "epoch": 3.6594553869335917, |
| "grad_norm": 0.5723239183425903, |
| "learning_rate": 0.0003809201594655748, |
| "loss": 3.458, |
| "step": 34000 |
| }, |
| { |
| "epoch": 3.6594553869335917, |
| "eval_accuracy": 0.37387969365535584, |
| "eval_loss": 3.4675986766815186, |
| "eval_runtime": 185.1386, |
| "eval_samples_per_second": 97.284, |
| "eval_steps_per_second": 6.082, |
| "step": 34000 |
| }, |
| { |
| "epoch": 3.6648369389732, |
| "grad_norm": 0.6167240142822266, |
| "learning_rate": 0.0003805969184355134, |
| "loss": 3.4585, |
| "step": 34050 |
| }, |
| { |
| "epoch": 3.670218491012808, |
| "grad_norm": 0.6382555961608887, |
| "learning_rate": 0.0003802736774054519, |
| "loss": 3.4491, |
| "step": 34100 |
| }, |
| { |
| "epoch": 3.675600043052416, |
| "grad_norm": 0.5925253033638, |
| "learning_rate": 0.00037995043637539057, |
| "loss": 3.4558, |
| "step": 34150 |
| }, |
| { |
| "epoch": 3.6809815950920246, |
| "grad_norm": 0.5835258960723877, |
| "learning_rate": 0.00037962719534532916, |
| "loss": 3.4596, |
| "step": 34200 |
| }, |
| { |
| "epoch": 3.6863631471316327, |
| "grad_norm": 0.8861356377601624, |
| "learning_rate": 0.0003793039543152677, |
| "loss": 3.4475, |
| "step": 34250 |
| }, |
| { |
| "epoch": 3.691744699171241, |
| "grad_norm": 0.6378123760223389, |
| "learning_rate": 0.0003789807132852063, |
| "loss": 3.4609, |
| "step": 34300 |
| }, |
| { |
| "epoch": 3.6971262512108494, |
| "grad_norm": 0.6498599648475647, |
| "learning_rate": 0.0003786574722551449, |
| "loss": 3.4649, |
| "step": 34350 |
| }, |
| { |
| "epoch": 3.7025078032504575, |
| "grad_norm": 0.6097577810287476, |
| "learning_rate": 0.00037833423122508343, |
| "loss": 3.4591, |
| "step": 34400 |
| }, |
| { |
| "epoch": 3.7078893552900656, |
| "grad_norm": 0.6025764346122742, |
| "learning_rate": 0.0003780109901950221, |
| "loss": 3.4636, |
| "step": 34450 |
| }, |
| { |
| "epoch": 3.713270907329674, |
| "grad_norm": 0.6137334108352661, |
| "learning_rate": 0.0003776877491649607, |
| "loss": 3.4743, |
| "step": 34500 |
| }, |
| { |
| "epoch": 3.7186524593692822, |
| "grad_norm": 0.6193282604217529, |
| "learning_rate": 0.0003773645081348992, |
| "loss": 3.4563, |
| "step": 34550 |
| }, |
| { |
| "epoch": 3.7240340114088903, |
| "grad_norm": 0.6524245738983154, |
| "learning_rate": 0.0003770412671048378, |
| "loss": 3.4657, |
| "step": 34600 |
| }, |
| { |
| "epoch": 3.7294155634484984, |
| "grad_norm": 0.6087914705276489, |
| "learning_rate": 0.00037671802607477635, |
| "loss": 3.4607, |
| "step": 34650 |
| }, |
| { |
| "epoch": 3.7347971154881066, |
| "grad_norm": 0.6221836805343628, |
| "learning_rate": 0.000376394785044715, |
| "loss": 3.4561, |
| "step": 34700 |
| }, |
| { |
| "epoch": 3.740178667527715, |
| "grad_norm": 0.6097055077552795, |
| "learning_rate": 0.0003760715440146536, |
| "loss": 3.452, |
| "step": 34750 |
| }, |
| { |
| "epoch": 3.745560219567323, |
| "grad_norm": 0.655005693435669, |
| "learning_rate": 0.00037574830298459214, |
| "loss": 3.4711, |
| "step": 34800 |
| }, |
| { |
| "epoch": 3.7509417716069313, |
| "grad_norm": 0.6379830837249756, |
| "learning_rate": 0.00037542506195453073, |
| "loss": 3.4451, |
| "step": 34850 |
| }, |
| { |
| "epoch": 3.75632332364654, |
| "grad_norm": 0.6179572939872742, |
| "learning_rate": 0.0003751018209244693, |
| "loss": 3.4654, |
| "step": 34900 |
| }, |
| { |
| "epoch": 3.761704875686148, |
| "grad_norm": 0.612697184085846, |
| "learning_rate": 0.00037477857989440787, |
| "loss": 3.4587, |
| "step": 34950 |
| }, |
| { |
| "epoch": 3.767086427725756, |
| "grad_norm": 0.6860320568084717, |
| "learning_rate": 0.0003744553388643465, |
| "loss": 3.4454, |
| "step": 35000 |
| }, |
| { |
| "epoch": 3.767086427725756, |
| "eval_accuracy": 0.37395770642880666, |
| "eval_loss": 3.4650323390960693, |
| "eval_runtime": 184.8816, |
| "eval_samples_per_second": 97.419, |
| "eval_steps_per_second": 6.09, |
| "step": 35000 |
| }, |
| { |
| "epoch": 3.772467979765364, |
| "grad_norm": 0.6109294295310974, |
| "learning_rate": 0.0003741320978342851, |
| "loss": 3.4285, |
| "step": 35050 |
| }, |
| { |
| "epoch": 3.7778495318049723, |
| "grad_norm": 0.6423705816268921, |
| "learning_rate": 0.00037380885680422365, |
| "loss": 3.4731, |
| "step": 35100 |
| }, |
| { |
| "epoch": 3.783231083844581, |
| "grad_norm": 0.6706321835517883, |
| "learning_rate": 0.00037348561577416224, |
| "loss": 3.4448, |
| "step": 35150 |
| }, |
| { |
| "epoch": 3.788612635884189, |
| "grad_norm": 0.5894853472709656, |
| "learning_rate": 0.0003731623747441008, |
| "loss": 3.4727, |
| "step": 35200 |
| }, |
| { |
| "epoch": 3.793994187923797, |
| "grad_norm": 0.6134861707687378, |
| "learning_rate": 0.0003728391337140394, |
| "loss": 3.4464, |
| "step": 35250 |
| }, |
| { |
| "epoch": 3.7993757399634056, |
| "grad_norm": 0.620707094669342, |
| "learning_rate": 0.00037251589268397803, |
| "loss": 3.4611, |
| "step": 35300 |
| }, |
| { |
| "epoch": 3.8047572920030137, |
| "grad_norm": 0.6619590520858765, |
| "learning_rate": 0.00037219265165391657, |
| "loss": 3.4473, |
| "step": 35350 |
| }, |
| { |
| "epoch": 3.810138844042622, |
| "grad_norm": 0.6385608315467834, |
| "learning_rate": 0.00037186941062385516, |
| "loss": 3.4468, |
| "step": 35400 |
| }, |
| { |
| "epoch": 3.8155203960822304, |
| "grad_norm": 0.5904738903045654, |
| "learning_rate": 0.0003715461695937937, |
| "loss": 3.4418, |
| "step": 35450 |
| }, |
| { |
| "epoch": 3.8209019481218385, |
| "grad_norm": 0.62429279088974, |
| "learning_rate": 0.0003712229285637323, |
| "loss": 3.4486, |
| "step": 35500 |
| }, |
| { |
| "epoch": 3.8262835001614466, |
| "grad_norm": 0.6489128470420837, |
| "learning_rate": 0.0003708996875336709, |
| "loss": 3.4485, |
| "step": 35550 |
| }, |
| { |
| "epoch": 3.8316650522010547, |
| "grad_norm": 0.6631107926368713, |
| "learning_rate": 0.0003705764465036095, |
| "loss": 3.4504, |
| "step": 35600 |
| }, |
| { |
| "epoch": 3.837046604240663, |
| "grad_norm": 0.6365602612495422, |
| "learning_rate": 0.0003702532054735481, |
| "loss": 3.4585, |
| "step": 35650 |
| }, |
| { |
| "epoch": 3.8424281562802713, |
| "grad_norm": 0.6052917838096619, |
| "learning_rate": 0.0003699299644434867, |
| "loss": 3.4588, |
| "step": 35700 |
| }, |
| { |
| "epoch": 3.8478097083198795, |
| "grad_norm": 0.60471510887146, |
| "learning_rate": 0.0003696067234134252, |
| "loss": 3.463, |
| "step": 35750 |
| }, |
| { |
| "epoch": 3.8531912603594876, |
| "grad_norm": 0.6473640203475952, |
| "learning_rate": 0.0003692834823833638, |
| "loss": 3.4513, |
| "step": 35800 |
| }, |
| { |
| "epoch": 3.858572812399096, |
| "grad_norm": 0.6347455978393555, |
| "learning_rate": 0.00036896024135330246, |
| "loss": 3.4592, |
| "step": 35850 |
| }, |
| { |
| "epoch": 3.863954364438704, |
| "grad_norm": 0.646809995174408, |
| "learning_rate": 0.000368637000323241, |
| "loss": 3.4763, |
| "step": 35900 |
| }, |
| { |
| "epoch": 3.8693359164783123, |
| "grad_norm": 0.6334096193313599, |
| "learning_rate": 0.0003683137592931796, |
| "loss": 3.4571, |
| "step": 35950 |
| }, |
| { |
| "epoch": 3.8747174685179204, |
| "grad_norm": 0.6145724654197693, |
| "learning_rate": 0.00036799051826311814, |
| "loss": 3.4562, |
| "step": 36000 |
| }, |
| { |
| "epoch": 3.8747174685179204, |
| "eval_accuracy": 0.3749936030612299, |
| "eval_loss": 3.4543886184692383, |
| "eval_runtime": 185.0342, |
| "eval_samples_per_second": 97.339, |
| "eval_steps_per_second": 6.085, |
| "step": 36000 |
| }, |
| { |
| "epoch": 3.8800990205575285, |
| "grad_norm": 0.640128493309021, |
| "learning_rate": 0.00036766727723305673, |
| "loss": 3.4502, |
| "step": 36050 |
| }, |
| { |
| "epoch": 3.885480572597137, |
| "grad_norm": 0.6229842305183411, |
| "learning_rate": 0.0003673440362029953, |
| "loss": 3.4506, |
| "step": 36100 |
| }, |
| { |
| "epoch": 3.890862124636745, |
| "grad_norm": 0.6296483278274536, |
| "learning_rate": 0.0003670207951729339, |
| "loss": 3.469, |
| "step": 36150 |
| }, |
| { |
| "epoch": 3.8962436766763533, |
| "grad_norm": 0.6755225658416748, |
| "learning_rate": 0.0003666975541428725, |
| "loss": 3.4531, |
| "step": 36200 |
| }, |
| { |
| "epoch": 3.901625228715962, |
| "grad_norm": 0.6031575798988342, |
| "learning_rate": 0.0003663743131128111, |
| "loss": 3.4584, |
| "step": 36250 |
| }, |
| { |
| "epoch": 3.90700678075557, |
| "grad_norm": 0.6367796063423157, |
| "learning_rate": 0.00036605107208274965, |
| "loss": 3.4711, |
| "step": 36300 |
| }, |
| { |
| "epoch": 3.912388332795178, |
| "grad_norm": 0.6142009496688843, |
| "learning_rate": 0.00036572783105268824, |
| "loss": 3.4706, |
| "step": 36350 |
| }, |
| { |
| "epoch": 3.9177698848347866, |
| "grad_norm": 0.5969540476799011, |
| "learning_rate": 0.0003654045900226268, |
| "loss": 3.457, |
| "step": 36400 |
| }, |
| { |
| "epoch": 3.9231514368743947, |
| "grad_norm": 0.6135761737823486, |
| "learning_rate": 0.00036508134899256543, |
| "loss": 3.4594, |
| "step": 36450 |
| }, |
| { |
| "epoch": 3.928532988914003, |
| "grad_norm": 0.6303659677505493, |
| "learning_rate": 0.00036475810796250403, |
| "loss": 3.4445, |
| "step": 36500 |
| }, |
| { |
| "epoch": 3.933914540953611, |
| "grad_norm": 0.6449431777000427, |
| "learning_rate": 0.00036443486693244257, |
| "loss": 3.4505, |
| "step": 36550 |
| }, |
| { |
| "epoch": 3.939296092993219, |
| "grad_norm": 0.6396144032478333, |
| "learning_rate": 0.00036411162590238116, |
| "loss": 3.4268, |
| "step": 36600 |
| }, |
| { |
| "epoch": 3.9446776450328276, |
| "grad_norm": 0.6068968176841736, |
| "learning_rate": 0.00036378838487231976, |
| "loss": 3.4544, |
| "step": 36650 |
| }, |
| { |
| "epoch": 3.9500591970724357, |
| "grad_norm": 0.6107643842697144, |
| "learning_rate": 0.00036346514384225835, |
| "loss": 3.4429, |
| "step": 36700 |
| }, |
| { |
| "epoch": 3.955440749112044, |
| "grad_norm": 0.6284364461898804, |
| "learning_rate": 0.00036314190281219695, |
| "loss": 3.438, |
| "step": 36750 |
| }, |
| { |
| "epoch": 3.9608223011516523, |
| "grad_norm": 0.6059133410453796, |
| "learning_rate": 0.00036281866178213554, |
| "loss": 3.4578, |
| "step": 36800 |
| }, |
| { |
| "epoch": 3.9662038531912605, |
| "grad_norm": 0.5825189352035522, |
| "learning_rate": 0.0003624954207520741, |
| "loss": 3.4431, |
| "step": 36850 |
| }, |
| { |
| "epoch": 3.9715854052308686, |
| "grad_norm": 0.6884462833404541, |
| "learning_rate": 0.0003621721797220127, |
| "loss": 3.4577, |
| "step": 36900 |
| }, |
| { |
| "epoch": 3.9769669572704767, |
| "grad_norm": 0.6246378421783447, |
| "learning_rate": 0.0003618554035125525, |
| "loss": 3.4424, |
| "step": 36950 |
| }, |
| { |
| "epoch": 3.9823485093100848, |
| "grad_norm": 0.6227812767028809, |
| "learning_rate": 0.0003615321624824911, |
| "loss": 3.444, |
| "step": 37000 |
| }, |
| { |
| "epoch": 3.9823485093100848, |
| "eval_accuracy": 0.37545526918155364, |
| "eval_loss": 3.4485981464385986, |
| "eval_runtime": 186.1677, |
| "eval_samples_per_second": 96.746, |
| "eval_steps_per_second": 6.048, |
| "step": 37000 |
| }, |
| { |
| "epoch": 3.9877300613496933, |
| "grad_norm": 0.665170431137085, |
| "learning_rate": 0.0003612089214524296, |
| "loss": 3.4387, |
| "step": 37050 |
| }, |
| { |
| "epoch": 3.9931116133893014, |
| "grad_norm": 0.6156587600708008, |
| "learning_rate": 0.0003608921452429695, |
| "loss": 3.4548, |
| "step": 37100 |
| }, |
| { |
| "epoch": 3.9984931654289095, |
| "grad_norm": 0.6279550194740295, |
| "learning_rate": 0.000360568904212908, |
| "loss": 3.461, |
| "step": 37150 |
| }, |
| { |
| "epoch": 4.003874717468518, |
| "grad_norm": 0.6261448860168457, |
| "learning_rate": 0.00036024566318284667, |
| "loss": 3.3728, |
| "step": 37200 |
| }, |
| { |
| "epoch": 4.009256269508126, |
| "grad_norm": 0.6143143177032471, |
| "learning_rate": 0.00035992242215278526, |
| "loss": 3.3296, |
| "step": 37250 |
| }, |
| { |
| "epoch": 4.014637821547734, |
| "grad_norm": 0.655872642993927, |
| "learning_rate": 0.0003595991811227238, |
| "loss": 3.3614, |
| "step": 37300 |
| }, |
| { |
| "epoch": 4.020019373587343, |
| "grad_norm": 0.6455563306808472, |
| "learning_rate": 0.0003592759400926624, |
| "loss": 3.3377, |
| "step": 37350 |
| }, |
| { |
| "epoch": 4.0254009256269505, |
| "grad_norm": 0.6441978216171265, |
| "learning_rate": 0.00035895269906260094, |
| "loss": 3.3616, |
| "step": 37400 |
| }, |
| { |
| "epoch": 4.030782477666559, |
| "grad_norm": 0.6673616766929626, |
| "learning_rate": 0.0003586294580325396, |
| "loss": 3.3547, |
| "step": 37450 |
| }, |
| { |
| "epoch": 4.036164029706168, |
| "grad_norm": 0.6724904775619507, |
| "learning_rate": 0.0003583062170024782, |
| "loss": 3.3415, |
| "step": 37500 |
| }, |
| { |
| "epoch": 4.041545581745775, |
| "grad_norm": 0.6894497275352478, |
| "learning_rate": 0.0003579829759724167, |
| "loss": 3.3525, |
| "step": 37550 |
| }, |
| { |
| "epoch": 4.046927133785384, |
| "grad_norm": 0.6247220635414124, |
| "learning_rate": 0.0003576597349423553, |
| "loss": 3.3583, |
| "step": 37600 |
| }, |
| { |
| "epoch": 4.0523086858249915, |
| "grad_norm": 0.644463062286377, |
| "learning_rate": 0.0003573364939122939, |
| "loss": 3.363, |
| "step": 37650 |
| }, |
| { |
| "epoch": 4.0576902378646, |
| "grad_norm": 0.6603822112083435, |
| "learning_rate": 0.00035701325288223245, |
| "loss": 3.3778, |
| "step": 37700 |
| }, |
| { |
| "epoch": 4.063071789904209, |
| "grad_norm": 0.652593731880188, |
| "learning_rate": 0.0003566900118521711, |
| "loss": 3.3786, |
| "step": 37750 |
| }, |
| { |
| "epoch": 4.068453341943816, |
| "grad_norm": 0.6232748627662659, |
| "learning_rate": 0.0003563667708221097, |
| "loss": 3.3588, |
| "step": 37800 |
| }, |
| { |
| "epoch": 4.073834893983425, |
| "grad_norm": 0.6294955015182495, |
| "learning_rate": 0.00035604352979204824, |
| "loss": 3.3854, |
| "step": 37850 |
| }, |
| { |
| "epoch": 4.079216446023033, |
| "grad_norm": 0.6011383533477783, |
| "learning_rate": 0.00035572028876198683, |
| "loss": 3.3782, |
| "step": 37900 |
| }, |
| { |
| "epoch": 4.084597998062641, |
| "grad_norm": 0.6583512425422668, |
| "learning_rate": 0.00035539704773192537, |
| "loss": 3.3833, |
| "step": 37950 |
| }, |
| { |
| "epoch": 4.08997955010225, |
| "grad_norm": 0.6446003317832947, |
| "learning_rate": 0.00035507380670186397, |
| "loss": 3.3707, |
| "step": 38000 |
| }, |
| { |
| "epoch": 4.08997955010225, |
| "eval_accuracy": 0.3760752425594231, |
| "eval_loss": 3.4499671459198, |
| "eval_runtime": 184.8025, |
| "eval_samples_per_second": 97.461, |
| "eval_steps_per_second": 6.093, |
| "step": 38000 |
| }, |
| { |
| "epoch": 4.095361102141858, |
| "grad_norm": 0.6237815022468567, |
| "learning_rate": 0.0003547505656718026, |
| "loss": 3.3644, |
| "step": 38050 |
| }, |
| { |
| "epoch": 4.100742654181466, |
| "grad_norm": 0.6614598631858826, |
| "learning_rate": 0.00035442732464174116, |
| "loss": 3.3779, |
| "step": 38100 |
| }, |
| { |
| "epoch": 4.106124206221074, |
| "grad_norm": 0.6301232576370239, |
| "learning_rate": 0.00035410408361167975, |
| "loss": 3.3746, |
| "step": 38150 |
| }, |
| { |
| "epoch": 4.111505758260682, |
| "grad_norm": 0.6670769453048706, |
| "learning_rate": 0.0003537808425816183, |
| "loss": 3.38, |
| "step": 38200 |
| }, |
| { |
| "epoch": 4.1168873103002905, |
| "grad_norm": 0.673061728477478, |
| "learning_rate": 0.0003534576015515569, |
| "loss": 3.3393, |
| "step": 38250 |
| }, |
| { |
| "epoch": 4.122268862339899, |
| "grad_norm": 0.6556143164634705, |
| "learning_rate": 0.00035313436052149553, |
| "loss": 3.3838, |
| "step": 38300 |
| }, |
| { |
| "epoch": 4.127650414379507, |
| "grad_norm": 0.6522071957588196, |
| "learning_rate": 0.0003528111194914341, |
| "loss": 3.3699, |
| "step": 38350 |
| }, |
| { |
| "epoch": 4.133031966419115, |
| "grad_norm": 0.7275823354721069, |
| "learning_rate": 0.00035248787846137267, |
| "loss": 3.3934, |
| "step": 38400 |
| }, |
| { |
| "epoch": 4.138413518458724, |
| "grad_norm": 0.6600149273872375, |
| "learning_rate": 0.00035216463743131126, |
| "loss": 3.399, |
| "step": 38450 |
| }, |
| { |
| "epoch": 4.1437950704983315, |
| "grad_norm": 0.6132925748825073, |
| "learning_rate": 0.0003518413964012498, |
| "loss": 3.3967, |
| "step": 38500 |
| }, |
| { |
| "epoch": 4.14917662253794, |
| "grad_norm": 0.616926908493042, |
| "learning_rate": 0.0003515181553711884, |
| "loss": 3.3699, |
| "step": 38550 |
| }, |
| { |
| "epoch": 4.154558174577549, |
| "grad_norm": 0.6914690732955933, |
| "learning_rate": 0.00035119491434112705, |
| "loss": 3.3801, |
| "step": 38600 |
| }, |
| { |
| "epoch": 4.159939726617156, |
| "grad_norm": 0.631589412689209, |
| "learning_rate": 0.0003508716733110656, |
| "loss": 3.3869, |
| "step": 38650 |
| }, |
| { |
| "epoch": 4.165321278656765, |
| "grad_norm": 0.6257556676864624, |
| "learning_rate": 0.0003505484322810042, |
| "loss": 3.3908, |
| "step": 38700 |
| }, |
| { |
| "epoch": 4.1707028306963725, |
| "grad_norm": 0.6427106261253357, |
| "learning_rate": 0.0003502251912509427, |
| "loss": 3.3834, |
| "step": 38750 |
| }, |
| { |
| "epoch": 4.176084382735981, |
| "grad_norm": 0.6590670347213745, |
| "learning_rate": 0.0003499019502208813, |
| "loss": 3.3694, |
| "step": 38800 |
| }, |
| { |
| "epoch": 4.18146593477559, |
| "grad_norm": 0.6249048113822937, |
| "learning_rate": 0.0003495787091908199, |
| "loss": 3.3961, |
| "step": 38850 |
| }, |
| { |
| "epoch": 4.186847486815197, |
| "grad_norm": 0.6336603760719299, |
| "learning_rate": 0.0003492554681607585, |
| "loss": 3.3878, |
| "step": 38900 |
| }, |
| { |
| "epoch": 4.192229038854806, |
| "grad_norm": 0.6488654613494873, |
| "learning_rate": 0.0003489322271306971, |
| "loss": 3.3747, |
| "step": 38950 |
| }, |
| { |
| "epoch": 4.197610590894414, |
| "grad_norm": 0.6409652829170227, |
| "learning_rate": 0.0003486089861006357, |
| "loss": 3.3754, |
| "step": 39000 |
| }, |
| { |
| "epoch": 4.197610590894414, |
| "eval_accuracy": 0.3763153654414931, |
| "eval_loss": 3.4473249912261963, |
| "eval_runtime": 184.9125, |
| "eval_samples_per_second": 97.403, |
| "eval_steps_per_second": 6.089, |
| "step": 39000 |
| }, |
| { |
| "epoch": 4.202992142934022, |
| "grad_norm": 0.6476558446884155, |
| "learning_rate": 0.00034828574507057424, |
| "loss": 3.3746, |
| "step": 39050 |
| }, |
| { |
| "epoch": 4.208373694973631, |
| "grad_norm": 0.6456419825553894, |
| "learning_rate": 0.00034796250404051283, |
| "loss": 3.4059, |
| "step": 39100 |
| }, |
| { |
| "epoch": 4.213755247013238, |
| "grad_norm": 0.6619618535041809, |
| "learning_rate": 0.00034763926301045137, |
| "loss": 3.3704, |
| "step": 39150 |
| }, |
| { |
| "epoch": 4.219136799052847, |
| "grad_norm": 0.6563810706138611, |
| "learning_rate": 0.00034731602198039, |
| "loss": 3.3754, |
| "step": 39200 |
| }, |
| { |
| "epoch": 4.224518351092455, |
| "grad_norm": 0.6438670754432678, |
| "learning_rate": 0.0003469927809503286, |
| "loss": 3.3833, |
| "step": 39250 |
| }, |
| { |
| "epoch": 4.229899903132063, |
| "grad_norm": 0.6422169208526611, |
| "learning_rate": 0.00034666953992026716, |
| "loss": 3.3859, |
| "step": 39300 |
| }, |
| { |
| "epoch": 4.2352814551716715, |
| "grad_norm": 0.7038455009460449, |
| "learning_rate": 0.00034634629889020575, |
| "loss": 3.3857, |
| "step": 39350 |
| }, |
| { |
| "epoch": 4.24066300721128, |
| "grad_norm": 0.6076750159263611, |
| "learning_rate": 0.00034602305786014435, |
| "loss": 3.3918, |
| "step": 39400 |
| }, |
| { |
| "epoch": 4.246044559250888, |
| "grad_norm": 0.68123859167099, |
| "learning_rate": 0.00034569981683008294, |
| "loss": 3.3651, |
| "step": 39450 |
| }, |
| { |
| "epoch": 4.251426111290496, |
| "grad_norm": 0.6171448826789856, |
| "learning_rate": 0.00034537657580002154, |
| "loss": 3.3783, |
| "step": 39500 |
| }, |
| { |
| "epoch": 4.256807663330104, |
| "grad_norm": 0.710161030292511, |
| "learning_rate": 0.00034505333476996013, |
| "loss": 3.3814, |
| "step": 39550 |
| }, |
| { |
| "epoch": 4.2621892153697125, |
| "grad_norm": 0.636483371257782, |
| "learning_rate": 0.00034473009373989867, |
| "loss": 3.3792, |
| "step": 39600 |
| }, |
| { |
| "epoch": 4.267570767409321, |
| "grad_norm": 0.5944780707359314, |
| "learning_rate": 0.00034440685270983727, |
| "loss": 3.399, |
| "step": 39650 |
| }, |
| { |
| "epoch": 4.272952319448929, |
| "grad_norm": 0.7163312435150146, |
| "learning_rate": 0.0003440836116797758, |
| "loss": 3.3993, |
| "step": 39700 |
| }, |
| { |
| "epoch": 4.278333871488537, |
| "grad_norm": 0.6651536226272583, |
| "learning_rate": 0.00034376037064971445, |
| "loss": 3.3949, |
| "step": 39750 |
| }, |
| { |
| "epoch": 4.283715423528146, |
| "grad_norm": 0.6634711623191833, |
| "learning_rate": 0.00034343712961965305, |
| "loss": 3.3995, |
| "step": 39800 |
| }, |
| { |
| "epoch": 4.2890969755677535, |
| "grad_norm": 0.6421599984169006, |
| "learning_rate": 0.0003431138885895916, |
| "loss": 3.3957, |
| "step": 39850 |
| }, |
| { |
| "epoch": 4.294478527607362, |
| "grad_norm": 0.6643863320350647, |
| "learning_rate": 0.0003427906475595302, |
| "loss": 3.3875, |
| "step": 39900 |
| }, |
| { |
| "epoch": 4.299860079646971, |
| "grad_norm": 0.6470743417739868, |
| "learning_rate": 0.0003424674065294688, |
| "loss": 3.3964, |
| "step": 39950 |
| }, |
| { |
| "epoch": 4.305241631686578, |
| "grad_norm": 0.6629840135574341, |
| "learning_rate": 0.0003421441654994073, |
| "loss": 3.392, |
| "step": 40000 |
| }, |
| { |
| "epoch": 4.305241631686578, |
| "eval_accuracy": 0.37669543324306815, |
| "eval_loss": 3.4432873725891113, |
| "eval_runtime": 184.8006, |
| "eval_samples_per_second": 97.462, |
| "eval_steps_per_second": 6.093, |
| "step": 40000 |
| }, |
| { |
| "epoch": 4.310623183726187, |
| "grad_norm": 0.6477255821228027, |
| "learning_rate": 0.00034182092446934597, |
| "loss": 3.4029, |
| "step": 40050 |
| }, |
| { |
| "epoch": 4.3160047357657945, |
| "grad_norm": 0.6865015029907227, |
| "learning_rate": 0.00034149768343928456, |
| "loss": 3.3759, |
| "step": 40100 |
| }, |
| { |
| "epoch": 4.321386287805403, |
| "grad_norm": 0.8546903729438782, |
| "learning_rate": 0.0003411744424092231, |
| "loss": 3.4042, |
| "step": 40150 |
| }, |
| { |
| "epoch": 4.326767839845012, |
| "grad_norm": 0.6472812294960022, |
| "learning_rate": 0.0003408512013791617, |
| "loss": 3.3882, |
| "step": 40200 |
| }, |
| { |
| "epoch": 4.332149391884619, |
| "grad_norm": 0.6662764549255371, |
| "learning_rate": 0.00034052796034910024, |
| "loss": 3.3868, |
| "step": 40250 |
| }, |
| { |
| "epoch": 4.337530943924228, |
| "grad_norm": 0.6494865417480469, |
| "learning_rate": 0.0003402047193190389, |
| "loss": 3.3874, |
| "step": 40300 |
| }, |
| { |
| "epoch": 4.342912495963836, |
| "grad_norm": 0.7330858111381531, |
| "learning_rate": 0.0003398814782889775, |
| "loss": 3.3871, |
| "step": 40350 |
| }, |
| { |
| "epoch": 4.348294048003444, |
| "grad_norm": 0.6532103419303894, |
| "learning_rate": 0.000339558237258916, |
| "loss": 3.395, |
| "step": 40400 |
| }, |
| { |
| "epoch": 4.3536756000430525, |
| "grad_norm": 0.652320384979248, |
| "learning_rate": 0.0003392349962288546, |
| "loss": 3.3932, |
| "step": 40450 |
| }, |
| { |
| "epoch": 4.359057152082661, |
| "grad_norm": 0.6824901700019836, |
| "learning_rate": 0.0003389117551987932, |
| "loss": 3.364, |
| "step": 40500 |
| }, |
| { |
| "epoch": 4.364438704122269, |
| "grad_norm": 0.6938865780830383, |
| "learning_rate": 0.00033858851416873175, |
| "loss": 3.3868, |
| "step": 40550 |
| }, |
| { |
| "epoch": 4.369820256161877, |
| "grad_norm": 0.7002043724060059, |
| "learning_rate": 0.0003382652731386704, |
| "loss": 3.3905, |
| "step": 40600 |
| }, |
| { |
| "epoch": 4.375201808201485, |
| "grad_norm": 0.6329423785209656, |
| "learning_rate": 0.000337942032108609, |
| "loss": 3.3952, |
| "step": 40650 |
| }, |
| { |
| "epoch": 4.3805833602410935, |
| "grad_norm": 0.6581116914749146, |
| "learning_rate": 0.00033761879107854754, |
| "loss": 3.3918, |
| "step": 40700 |
| }, |
| { |
| "epoch": 4.385964912280702, |
| "grad_norm": 0.6352985501289368, |
| "learning_rate": 0.00033729555004848613, |
| "loss": 3.3855, |
| "step": 40750 |
| }, |
| { |
| "epoch": 4.39134646432031, |
| "grad_norm": 0.6230785846710205, |
| "learning_rate": 0.00033697230901842467, |
| "loss": 3.4084, |
| "step": 40800 |
| }, |
| { |
| "epoch": 4.396728016359918, |
| "grad_norm": 0.6846745610237122, |
| "learning_rate": 0.00033664906798836327, |
| "loss": 3.3767, |
| "step": 40850 |
| }, |
| { |
| "epoch": 4.402109568399527, |
| "grad_norm": 0.6545602083206177, |
| "learning_rate": 0.0003363258269583019, |
| "loss": 3.3906, |
| "step": 40900 |
| }, |
| { |
| "epoch": 4.4074911204391345, |
| "grad_norm": 0.6751523017883301, |
| "learning_rate": 0.00033600258592824046, |
| "loss": 3.403, |
| "step": 40950 |
| }, |
| { |
| "epoch": 4.412872672478743, |
| "grad_norm": 0.6432384252548218, |
| "learning_rate": 0.00033567934489817905, |
| "loss": 3.3969, |
| "step": 41000 |
| }, |
| { |
| "epoch": 4.412872672478743, |
| "eval_accuracy": 0.37734398233043276, |
| "eval_loss": 3.4374191761016846, |
| "eval_runtime": 186.1551, |
| "eval_samples_per_second": 96.753, |
| "eval_steps_per_second": 6.049, |
| "step": 41000 |
| }, |
| { |
| "epoch": 4.418254224518351, |
| "grad_norm": 0.6344847679138184, |
| "learning_rate": 0.00033535610386811764, |
| "loss": 3.4161, |
| "step": 41050 |
| }, |
| { |
| "epoch": 4.423635776557959, |
| "grad_norm": 0.6220810413360596, |
| "learning_rate": 0.00033503932765865745, |
| "loss": 3.402, |
| "step": 41100 |
| }, |
| { |
| "epoch": 4.429017328597568, |
| "grad_norm": 0.6427842974662781, |
| "learning_rate": 0.000334716086628596, |
| "loss": 3.3949, |
| "step": 41150 |
| }, |
| { |
| "epoch": 4.4343988806371755, |
| "grad_norm": 0.6630915999412537, |
| "learning_rate": 0.0003343928455985346, |
| "loss": 3.3724, |
| "step": 41200 |
| }, |
| { |
| "epoch": 4.439780432676784, |
| "grad_norm": 0.6749280691146851, |
| "learning_rate": 0.00033406960456847324, |
| "loss": 3.4069, |
| "step": 41250 |
| }, |
| { |
| "epoch": 4.445161984716393, |
| "grad_norm": 0.6909694075584412, |
| "learning_rate": 0.0003337463635384118, |
| "loss": 3.3812, |
| "step": 41300 |
| }, |
| { |
| "epoch": 4.450543536756, |
| "grad_norm": 0.6568173170089722, |
| "learning_rate": 0.00033342312250835037, |
| "loss": 3.4003, |
| "step": 41350 |
| }, |
| { |
| "epoch": 4.455925088795609, |
| "grad_norm": 0.6784096956253052, |
| "learning_rate": 0.00033309988147828896, |
| "loss": 3.3896, |
| "step": 41400 |
| }, |
| { |
| "epoch": 4.461306640835216, |
| "grad_norm": 0.6052407622337341, |
| "learning_rate": 0.0003327766404482275, |
| "loss": 3.4065, |
| "step": 41450 |
| }, |
| { |
| "epoch": 4.466688192874825, |
| "grad_norm": 0.6263163089752197, |
| "learning_rate": 0.0003324533994181661, |
| "loss": 3.4116, |
| "step": 41500 |
| }, |
| { |
| "epoch": 4.4720697449144335, |
| "grad_norm": 0.6299830675125122, |
| "learning_rate": 0.00033213015838810475, |
| "loss": 3.3998, |
| "step": 41550 |
| }, |
| { |
| "epoch": 4.477451296954041, |
| "grad_norm": 0.6599326729774475, |
| "learning_rate": 0.0003318069173580433, |
| "loss": 3.396, |
| "step": 41600 |
| }, |
| { |
| "epoch": 4.48283284899365, |
| "grad_norm": 0.68622887134552, |
| "learning_rate": 0.0003314901411485831, |
| "loss": 3.3886, |
| "step": 41650 |
| }, |
| { |
| "epoch": 4.488214401033258, |
| "grad_norm": 0.6320132613182068, |
| "learning_rate": 0.0003311669001185217, |
| "loss": 3.3932, |
| "step": 41700 |
| }, |
| { |
| "epoch": 4.493595953072866, |
| "grad_norm": 0.6479349136352539, |
| "learning_rate": 0.0003308436590884603, |
| "loss": 3.3889, |
| "step": 41750 |
| }, |
| { |
| "epoch": 4.4989775051124745, |
| "grad_norm": 0.6223969459533691, |
| "learning_rate": 0.0003305204180583988, |
| "loss": 3.3739, |
| "step": 41800 |
| }, |
| { |
| "epoch": 4.504359057152083, |
| "grad_norm": 0.7559693455696106, |
| "learning_rate": 0.0003301971770283374, |
| "loss": 3.3848, |
| "step": 41850 |
| }, |
| { |
| "epoch": 4.509740609191691, |
| "grad_norm": 0.7084282040596008, |
| "learning_rate": 0.00032987393599827607, |
| "loss": 3.3905, |
| "step": 41900 |
| }, |
| { |
| "epoch": 4.515122161231299, |
| "grad_norm": 0.6577198505401611, |
| "learning_rate": 0.0003295506949682146, |
| "loss": 3.3904, |
| "step": 41950 |
| }, |
| { |
| "epoch": 4.520503713270907, |
| "grad_norm": 0.7152112126350403, |
| "learning_rate": 0.0003292274539381532, |
| "loss": 3.4062, |
| "step": 42000 |
| }, |
| { |
| "epoch": 4.520503713270907, |
| "eval_accuracy": 0.3782491695388152, |
| "eval_loss": 3.432380199432373, |
| "eval_runtime": 184.5835, |
| "eval_samples_per_second": 97.576, |
| "eval_steps_per_second": 6.1, |
| "step": 42000 |
| }, |
| { |
| "epoch": 4.5258852653105155, |
| "grad_norm": 0.6525520086288452, |
| "learning_rate": 0.00032890421290809174, |
| "loss": 3.4211, |
| "step": 42050 |
| }, |
| { |
| "epoch": 4.531266817350124, |
| "grad_norm": 0.6888226270675659, |
| "learning_rate": 0.00032858097187803034, |
| "loss": 3.4038, |
| "step": 42100 |
| }, |
| { |
| "epoch": 4.536648369389732, |
| "grad_norm": 0.6215112209320068, |
| "learning_rate": 0.00032825773084796893, |
| "loss": 3.3989, |
| "step": 42150 |
| }, |
| { |
| "epoch": 4.54202992142934, |
| "grad_norm": 0.6531518697738647, |
| "learning_rate": 0.00032793448981790753, |
| "loss": 3.4025, |
| "step": 42200 |
| }, |
| { |
| "epoch": 4.547411473468949, |
| "grad_norm": 0.6420837640762329, |
| "learning_rate": 0.0003276112487878461, |
| "loss": 3.3903, |
| "step": 42250 |
| }, |
| { |
| "epoch": 4.5527930255085565, |
| "grad_norm": 0.5992629528045654, |
| "learning_rate": 0.0003272880077577847, |
| "loss": 3.3813, |
| "step": 42300 |
| }, |
| { |
| "epoch": 4.558174577548165, |
| "grad_norm": 0.7291805148124695, |
| "learning_rate": 0.00032696476672772326, |
| "loss": 3.3928, |
| "step": 42350 |
| }, |
| { |
| "epoch": 4.563556129587774, |
| "grad_norm": 0.6690374612808228, |
| "learning_rate": 0.00032664152569766185, |
| "loss": 3.3947, |
| "step": 42400 |
| }, |
| { |
| "epoch": 4.568937681627381, |
| "grad_norm": 0.6393318176269531, |
| "learning_rate": 0.0003263182846676004, |
| "loss": 3.3933, |
| "step": 42450 |
| }, |
| { |
| "epoch": 4.57431923366699, |
| "grad_norm": 0.7334935665130615, |
| "learning_rate": 0.00032599504363753904, |
| "loss": 3.405, |
| "step": 42500 |
| }, |
| { |
| "epoch": 4.579700785706597, |
| "grad_norm": 0.6425918936729431, |
| "learning_rate": 0.00032567180260747764, |
| "loss": 3.3993, |
| "step": 42550 |
| }, |
| { |
| "epoch": 4.585082337746206, |
| "grad_norm": 0.625737726688385, |
| "learning_rate": 0.0003253485615774162, |
| "loss": 3.4049, |
| "step": 42600 |
| }, |
| { |
| "epoch": 4.5904638897858145, |
| "grad_norm": 0.6439979076385498, |
| "learning_rate": 0.00032502532054735477, |
| "loss": 3.4023, |
| "step": 42650 |
| }, |
| { |
| "epoch": 4.595845441825422, |
| "grad_norm": 0.666843831539154, |
| "learning_rate": 0.00032470207951729337, |
| "loss": 3.3937, |
| "step": 42700 |
| }, |
| { |
| "epoch": 4.601226993865031, |
| "grad_norm": 0.6379914283752441, |
| "learning_rate": 0.0003243788384872319, |
| "loss": 3.4084, |
| "step": 42750 |
| }, |
| { |
| "epoch": 4.606608545904638, |
| "grad_norm": 0.7040283679962158, |
| "learning_rate": 0.00032405559745717056, |
| "loss": 3.3822, |
| "step": 42800 |
| }, |
| { |
| "epoch": 4.611990097944247, |
| "grad_norm": 0.6507775187492371, |
| "learning_rate": 0.00032373235642710915, |
| "loss": 3.4123, |
| "step": 42850 |
| }, |
| { |
| "epoch": 4.6173716499838555, |
| "grad_norm": 0.6584706902503967, |
| "learning_rate": 0.0003234091153970477, |
| "loss": 3.3987, |
| "step": 42900 |
| }, |
| { |
| "epoch": 4.622753202023463, |
| "grad_norm": 0.7082355618476868, |
| "learning_rate": 0.0003230858743669863, |
| "loss": 3.3717, |
| "step": 42950 |
| }, |
| { |
| "epoch": 4.628134754063072, |
| "grad_norm": 0.6571763753890991, |
| "learning_rate": 0.0003227626333369248, |
| "loss": 3.3706, |
| "step": 43000 |
| }, |
| { |
| "epoch": 4.628134754063072, |
| "eval_accuracy": 0.3783615166248154, |
| "eval_loss": 3.4268484115600586, |
| "eval_runtime": 185.1004, |
| "eval_samples_per_second": 97.304, |
| "eval_steps_per_second": 6.083, |
| "step": 43000 |
| }, |
| { |
| "epoch": 4.63351630610268, |
| "grad_norm": 0.6413716077804565, |
| "learning_rate": 0.0003224393923068635, |
| "loss": 3.4092, |
| "step": 43050 |
| }, |
| { |
| "epoch": 4.638897858142288, |
| "grad_norm": 0.6072510480880737, |
| "learning_rate": 0.00032211615127680207, |
| "loss": 3.3956, |
| "step": 43100 |
| }, |
| { |
| "epoch": 4.6442794101818965, |
| "grad_norm": 0.672419011592865, |
| "learning_rate": 0.0003217929102467406, |
| "loss": 3.4054, |
| "step": 43150 |
| }, |
| { |
| "epoch": 4.649660962221505, |
| "grad_norm": 0.6509859561920166, |
| "learning_rate": 0.0003214696692166792, |
| "loss": 3.3917, |
| "step": 43200 |
| }, |
| { |
| "epoch": 4.655042514261113, |
| "grad_norm": 0.689818263053894, |
| "learning_rate": 0.0003211464281866178, |
| "loss": 3.3953, |
| "step": 43250 |
| }, |
| { |
| "epoch": 4.660424066300721, |
| "grad_norm": 0.6377256512641907, |
| "learning_rate": 0.00032082318715655634, |
| "loss": 3.3956, |
| "step": 43300 |
| }, |
| { |
| "epoch": 4.665805618340329, |
| "grad_norm": 0.680729329586029, |
| "learning_rate": 0.000320499946126495, |
| "loss": 3.383, |
| "step": 43350 |
| }, |
| { |
| "epoch": 4.6711871703799375, |
| "grad_norm": 0.7217500805854797, |
| "learning_rate": 0.0003201767050964336, |
| "loss": 3.4109, |
| "step": 43400 |
| }, |
| { |
| "epoch": 4.676568722419546, |
| "grad_norm": 0.6791247129440308, |
| "learning_rate": 0.0003198534640663721, |
| "loss": 3.4091, |
| "step": 43450 |
| }, |
| { |
| "epoch": 4.681950274459154, |
| "grad_norm": 0.6612512469291687, |
| "learning_rate": 0.0003195302230363107, |
| "loss": 3.3867, |
| "step": 43500 |
| }, |
| { |
| "epoch": 4.687331826498762, |
| "grad_norm": 0.6829741597175598, |
| "learning_rate": 0.00031920698200624926, |
| "loss": 3.4031, |
| "step": 43550 |
| }, |
| { |
| "epoch": 4.692713378538371, |
| "grad_norm": 0.6728960871696472, |
| "learning_rate": 0.00031888374097618785, |
| "loss": 3.3919, |
| "step": 43600 |
| }, |
| { |
| "epoch": 4.6980949305779784, |
| "grad_norm": 0.6731211543083191, |
| "learning_rate": 0.0003185604999461265, |
| "loss": 3.3839, |
| "step": 43650 |
| }, |
| { |
| "epoch": 4.703476482617587, |
| "grad_norm": 0.6127499341964722, |
| "learning_rate": 0.00031823725891606504, |
| "loss": 3.3901, |
| "step": 43700 |
| }, |
| { |
| "epoch": 4.7088580346571955, |
| "grad_norm": 0.6518672704696655, |
| "learning_rate": 0.00031791401788600364, |
| "loss": 3.392, |
| "step": 43750 |
| }, |
| { |
| "epoch": 4.714239586696803, |
| "grad_norm": 0.6531979441642761, |
| "learning_rate": 0.00031759077685594223, |
| "loss": 3.41, |
| "step": 43800 |
| }, |
| { |
| "epoch": 4.719621138736412, |
| "grad_norm": 0.6991766691207886, |
| "learning_rate": 0.00031726753582588077, |
| "loss": 3.3793, |
| "step": 43850 |
| }, |
| { |
| "epoch": 4.725002690776019, |
| "grad_norm": 0.704139232635498, |
| "learning_rate": 0.0003169442947958194, |
| "loss": 3.4015, |
| "step": 43900 |
| }, |
| { |
| "epoch": 4.730384242815628, |
| "grad_norm": 0.6892104148864746, |
| "learning_rate": 0.000316621053765758, |
| "loss": 3.3967, |
| "step": 43950 |
| }, |
| { |
| "epoch": 4.7357657948552365, |
| "grad_norm": 0.6997895240783691, |
| "learning_rate": 0.00031629781273569656, |
| "loss": 3.407, |
| "step": 44000 |
| }, |
| { |
| "epoch": 4.7357657948552365, |
| "eval_accuracy": 0.37935634246564476, |
| "eval_loss": 3.4216456413269043, |
| "eval_runtime": 184.8084, |
| "eval_samples_per_second": 97.458, |
| "eval_steps_per_second": 6.093, |
| "step": 44000 |
| }, |
| { |
| "epoch": 4.741147346894844, |
| "grad_norm": 0.6774421334266663, |
| "learning_rate": 0.00031597457170563515, |
| "loss": 3.3999, |
| "step": 44050 |
| }, |
| { |
| "epoch": 4.746528898934453, |
| "grad_norm": 0.7047556042671204, |
| "learning_rate": 0.0003156513306755737, |
| "loss": 3.3915, |
| "step": 44100 |
| }, |
| { |
| "epoch": 4.751910450974061, |
| "grad_norm": 0.6317178606987, |
| "learning_rate": 0.0003153280896455123, |
| "loss": 3.39, |
| "step": 44150 |
| }, |
| { |
| "epoch": 4.757292003013669, |
| "grad_norm": 0.6738790273666382, |
| "learning_rate": 0.00031500484861545094, |
| "loss": 3.3896, |
| "step": 44200 |
| }, |
| { |
| "epoch": 4.7626735550532775, |
| "grad_norm": 0.7213401794433594, |
| "learning_rate": 0.0003146816075853895, |
| "loss": 3.3808, |
| "step": 44250 |
| }, |
| { |
| "epoch": 4.768055107092886, |
| "grad_norm": 0.6467810273170471, |
| "learning_rate": 0.00031435836655532807, |
| "loss": 3.3878, |
| "step": 44300 |
| }, |
| { |
| "epoch": 4.773436659132494, |
| "grad_norm": 0.6356862187385559, |
| "learning_rate": 0.00031403512552526667, |
| "loss": 3.3864, |
| "step": 44350 |
| }, |
| { |
| "epoch": 4.778818211172102, |
| "grad_norm": 0.7657820582389832, |
| "learning_rate": 0.0003137118844952052, |
| "loss": 3.3988, |
| "step": 44400 |
| }, |
| { |
| "epoch": 4.78419976321171, |
| "grad_norm": 0.7307938933372498, |
| "learning_rate": 0.0003133886434651438, |
| "loss": 3.3934, |
| "step": 44450 |
| }, |
| { |
| "epoch": 4.7895813152513185, |
| "grad_norm": 0.6736586689949036, |
| "learning_rate": 0.00031306540243508245, |
| "loss": 3.4102, |
| "step": 44500 |
| }, |
| { |
| "epoch": 4.794962867290927, |
| "grad_norm": 0.6513236165046692, |
| "learning_rate": 0.000312742161405021, |
| "loss": 3.3931, |
| "step": 44550 |
| }, |
| { |
| "epoch": 4.800344419330535, |
| "grad_norm": 0.6173604130744934, |
| "learning_rate": 0.0003124189203749596, |
| "loss": 3.3887, |
| "step": 44600 |
| }, |
| { |
| "epoch": 4.805725971370143, |
| "grad_norm": 0.6931024789810181, |
| "learning_rate": 0.0003120956793448981, |
| "loss": 3.3973, |
| "step": 44650 |
| }, |
| { |
| "epoch": 4.811107523409751, |
| "grad_norm": 0.6727322936058044, |
| "learning_rate": 0.0003117724383148367, |
| "loss": 3.3839, |
| "step": 44700 |
| }, |
| { |
| "epoch": 4.8164890754493594, |
| "grad_norm": 0.6649513244628906, |
| "learning_rate": 0.00031144919728477526, |
| "loss": 3.3763, |
| "step": 44750 |
| }, |
| { |
| "epoch": 4.821870627488968, |
| "grad_norm": 0.6292411088943481, |
| "learning_rate": 0.0003111259562547139, |
| "loss": 3.3963, |
| "step": 44800 |
| }, |
| { |
| "epoch": 4.827252179528576, |
| "grad_norm": 0.6615663170814514, |
| "learning_rate": 0.0003108027152246525, |
| "loss": 3.3864, |
| "step": 44850 |
| }, |
| { |
| "epoch": 4.832633731568184, |
| "grad_norm": 0.6625530123710632, |
| "learning_rate": 0.0003104794741945911, |
| "loss": 3.4006, |
| "step": 44900 |
| }, |
| { |
| "epoch": 4.838015283607793, |
| "grad_norm": 0.6565783023834229, |
| "learning_rate": 0.00031015623316452964, |
| "loss": 3.3776, |
| "step": 44950 |
| }, |
| { |
| "epoch": 4.8433968356474, |
| "grad_norm": 0.7287229895591736, |
| "learning_rate": 0.00030983299213446823, |
| "loss": 3.385, |
| "step": 45000 |
| }, |
| { |
| "epoch": 4.8433968356474, |
| "eval_accuracy": 0.37947618660090415, |
| "eval_loss": 3.4171833992004395, |
| "eval_runtime": 184.7333, |
| "eval_samples_per_second": 97.497, |
| "eval_steps_per_second": 6.095, |
| "step": 45000 |
| }, |
| { |
| "epoch": 4.848778387687009, |
| "grad_norm": 0.6722894310951233, |
| "learning_rate": 0.0003095097511044069, |
| "loss": 3.3969, |
| "step": 45050 |
| }, |
| { |
| "epoch": 4.8541599397266175, |
| "grad_norm": 0.6436869502067566, |
| "learning_rate": 0.0003091865100743454, |
| "loss": 3.4128, |
| "step": 45100 |
| }, |
| { |
| "epoch": 4.859541491766225, |
| "grad_norm": 0.654567539691925, |
| "learning_rate": 0.000308863269044284, |
| "loss": 3.4056, |
| "step": 45150 |
| }, |
| { |
| "epoch": 4.864923043805834, |
| "grad_norm": 0.6916000247001648, |
| "learning_rate": 0.00030854002801422256, |
| "loss": 3.4162, |
| "step": 45200 |
| }, |
| { |
| "epoch": 4.870304595845441, |
| "grad_norm": 0.6123929023742676, |
| "learning_rate": 0.00030821678698416115, |
| "loss": 3.3915, |
| "step": 45250 |
| }, |
| { |
| "epoch": 4.87568614788505, |
| "grad_norm": 0.6771581172943115, |
| "learning_rate": 0.0003078935459540997, |
| "loss": 3.3876, |
| "step": 45300 |
| }, |
| { |
| "epoch": 4.8810676999246585, |
| "grad_norm": 0.6943243741989136, |
| "learning_rate": 0.00030757030492403834, |
| "loss": 3.3773, |
| "step": 45350 |
| }, |
| { |
| "epoch": 4.886449251964266, |
| "grad_norm": 0.7478858828544617, |
| "learning_rate": 0.00030724706389397694, |
| "loss": 3.3974, |
| "step": 45400 |
| }, |
| { |
| "epoch": 4.891830804003875, |
| "grad_norm": 0.6267790794372559, |
| "learning_rate": 0.0003069238228639155, |
| "loss": 3.3862, |
| "step": 45450 |
| }, |
| { |
| "epoch": 4.897212356043483, |
| "grad_norm": 0.6570143103599548, |
| "learning_rate": 0.00030660058183385407, |
| "loss": 3.3973, |
| "step": 45500 |
| }, |
| { |
| "epoch": 4.902593908083091, |
| "grad_norm": 0.6861966848373413, |
| "learning_rate": 0.00030627734080379267, |
| "loss": 3.399, |
| "step": 45550 |
| }, |
| { |
| "epoch": 4.9079754601226995, |
| "grad_norm": 0.6667155623435974, |
| "learning_rate": 0.0003059540997737312, |
| "loss": 3.4026, |
| "step": 45600 |
| }, |
| { |
| "epoch": 4.913357012162308, |
| "grad_norm": 0.7267022132873535, |
| "learning_rate": 0.00030563732356427107, |
| "loss": 3.3965, |
| "step": 45650 |
| }, |
| { |
| "epoch": 4.918738564201916, |
| "grad_norm": 0.7066245079040527, |
| "learning_rate": 0.00030531408253420966, |
| "loss": 3.4037, |
| "step": 45700 |
| }, |
| { |
| "epoch": 4.924120116241524, |
| "grad_norm": 0.693480372428894, |
| "learning_rate": 0.00030499084150414826, |
| "loss": 3.3913, |
| "step": 45750 |
| }, |
| { |
| "epoch": 4.929501668281132, |
| "grad_norm": 0.6911986470222473, |
| "learning_rate": 0.00030466760047408685, |
| "loss": 3.3884, |
| "step": 45800 |
| }, |
| { |
| "epoch": 4.9348832203207404, |
| "grad_norm": 0.6638094782829285, |
| "learning_rate": 0.0003043443594440254, |
| "loss": 3.3795, |
| "step": 45850 |
| }, |
| { |
| "epoch": 4.940264772360349, |
| "grad_norm": 0.6978832483291626, |
| "learning_rate": 0.000304021118413964, |
| "loss": 3.3869, |
| "step": 45900 |
| }, |
| { |
| "epoch": 4.945646324399957, |
| "grad_norm": 0.7054188251495361, |
| "learning_rate": 0.0003036978773839025, |
| "loss": 3.3958, |
| "step": 45950 |
| }, |
| { |
| "epoch": 4.951027876439565, |
| "grad_norm": 0.7118246555328369, |
| "learning_rate": 0.0003033746363538412, |
| "loss": 3.3928, |
| "step": 46000 |
| }, |
| { |
| "epoch": 4.951027876439565, |
| "eval_accuracy": 0.38003716146069033, |
| "eval_loss": 3.4111993312835693, |
| "eval_runtime": 185.018, |
| "eval_samples_per_second": 97.347, |
| "eval_steps_per_second": 6.086, |
| "step": 46000 |
| }, |
| { |
| "epoch": 4.956409428479174, |
| "grad_norm": 0.7064889073371887, |
| "learning_rate": 0.00030305139532377977, |
| "loss": 3.4139, |
| "step": 46050 |
| }, |
| { |
| "epoch": 4.961790980518781, |
| "grad_norm": 0.6596596240997314, |
| "learning_rate": 0.0003027281542937183, |
| "loss": 3.3863, |
| "step": 46100 |
| }, |
| { |
| "epoch": 4.96717253255839, |
| "grad_norm": 0.6461761593818665, |
| "learning_rate": 0.0003024049132636569, |
| "loss": 3.4073, |
| "step": 46150 |
| }, |
| { |
| "epoch": 4.9725540845979985, |
| "grad_norm": 0.7342297434806824, |
| "learning_rate": 0.0003020816722335955, |
| "loss": 3.3922, |
| "step": 46200 |
| }, |
| { |
| "epoch": 4.977935636637606, |
| "grad_norm": 0.7432610988616943, |
| "learning_rate": 0.00030175843120353404, |
| "loss": 3.4053, |
| "step": 46250 |
| }, |
| { |
| "epoch": 4.983317188677215, |
| "grad_norm": 0.6522065997123718, |
| "learning_rate": 0.0003014351901734727, |
| "loss": 3.406, |
| "step": 46300 |
| }, |
| { |
| "epoch": 4.988698740716822, |
| "grad_norm": 0.6344789862632751, |
| "learning_rate": 0.0003011119491434113, |
| "loss": 3.3822, |
| "step": 46350 |
| }, |
| { |
| "epoch": 4.994080292756431, |
| "grad_norm": 0.6894682049751282, |
| "learning_rate": 0.0003007887081133498, |
| "loss": 3.3982, |
| "step": 46400 |
| }, |
| { |
| "epoch": 4.9994618447960395, |
| "grad_norm": 0.6706180572509766, |
| "learning_rate": 0.0003004654670832884, |
| "loss": 3.3969, |
| "step": 46450 |
| }, |
| { |
| "epoch": 5.004843396835647, |
| "grad_norm": 0.6677836179733276, |
| "learning_rate": 0.00030014222605322696, |
| "loss": 3.3245, |
| "step": 46500 |
| }, |
| { |
| "epoch": 5.010224948875256, |
| "grad_norm": 0.6664152145385742, |
| "learning_rate": 0.00029981898502316555, |
| "loss": 3.3235, |
| "step": 46550 |
| }, |
| { |
| "epoch": 5.015606500914864, |
| "grad_norm": 0.6685128211975098, |
| "learning_rate": 0.00029949574399310415, |
| "loss": 3.3007, |
| "step": 46600 |
| }, |
| { |
| "epoch": 5.020988052954472, |
| "grad_norm": 0.6446495056152344, |
| "learning_rate": 0.00029917250296304274, |
| "loss": 3.2914, |
| "step": 46650 |
| }, |
| { |
| "epoch": 5.0263696049940805, |
| "grad_norm": 0.6748939752578735, |
| "learning_rate": 0.00029884926193298134, |
| "loss": 3.3175, |
| "step": 46700 |
| }, |
| { |
| "epoch": 5.031751157033688, |
| "grad_norm": 0.7316281199455261, |
| "learning_rate": 0.0002985260209029199, |
| "loss": 3.3109, |
| "step": 46750 |
| }, |
| { |
| "epoch": 5.037132709073297, |
| "grad_norm": 0.6529657244682312, |
| "learning_rate": 0.00029820277987285853, |
| "loss": 3.3074, |
| "step": 46800 |
| }, |
| { |
| "epoch": 5.042514261112905, |
| "grad_norm": 0.6704086065292358, |
| "learning_rate": 0.00029787953884279707, |
| "loss": 3.3047, |
| "step": 46850 |
| }, |
| { |
| "epoch": 5.047895813152513, |
| "grad_norm": 0.6949445009231567, |
| "learning_rate": 0.00029755629781273566, |
| "loss": 3.3009, |
| "step": 46900 |
| }, |
| { |
| "epoch": 5.0532773651921215, |
| "grad_norm": 0.6553006768226624, |
| "learning_rate": 0.00029723305678267426, |
| "loss": 3.3224, |
| "step": 46950 |
| }, |
| { |
| "epoch": 5.05865891723173, |
| "grad_norm": 0.7208836674690247, |
| "learning_rate": 0.00029690981575261285, |
| "loss": 3.3105, |
| "step": 47000 |
| }, |
| { |
| "epoch": 5.05865891723173, |
| "eval_accuracy": 0.38008953215261687, |
| "eval_loss": 3.415787935256958, |
| "eval_runtime": 184.7813, |
| "eval_samples_per_second": 97.472, |
| "eval_steps_per_second": 6.094, |
| "step": 47000 |
| }, |
| { |
| "epoch": 5.064040469271338, |
| "grad_norm": 0.7050295472145081, |
| "learning_rate": 0.0002965865747225514, |
| "loss": 3.3034, |
| "step": 47050 |
| }, |
| { |
| "epoch": 5.069422021310946, |
| "grad_norm": 0.6400262117385864, |
| "learning_rate": 0.00029626333369249, |
| "loss": 3.2999, |
| "step": 47100 |
| }, |
| { |
| "epoch": 5.074803573350554, |
| "grad_norm": 0.6437999606132507, |
| "learning_rate": 0.0002959400926624286, |
| "loss": 3.2865, |
| "step": 47150 |
| }, |
| { |
| "epoch": 5.080185125390162, |
| "grad_norm": 0.6724359393119812, |
| "learning_rate": 0.0002956168516323672, |
| "loss": 3.3, |
| "step": 47200 |
| }, |
| { |
| "epoch": 5.085566677429771, |
| "grad_norm": 0.6594730019569397, |
| "learning_rate": 0.00029529361060230577, |
| "loss": 3.3093, |
| "step": 47250 |
| }, |
| { |
| "epoch": 5.090948229469379, |
| "grad_norm": 0.687857449054718, |
| "learning_rate": 0.0002949703695722443, |
| "loss": 3.3153, |
| "step": 47300 |
| }, |
| { |
| "epoch": 5.096329781508987, |
| "grad_norm": 0.6695648431777954, |
| "learning_rate": 0.00029464712854218296, |
| "loss": 3.3269, |
| "step": 47350 |
| }, |
| { |
| "epoch": 5.101711333548596, |
| "grad_norm": 0.6603997945785522, |
| "learning_rate": 0.0002943238875121215, |
| "loss": 3.3302, |
| "step": 47400 |
| }, |
| { |
| "epoch": 5.107092885588203, |
| "grad_norm": 0.651153028011322, |
| "learning_rate": 0.0002940006464820601, |
| "loss": 3.3208, |
| "step": 47450 |
| }, |
| { |
| "epoch": 5.112474437627812, |
| "grad_norm": 0.7518373131752014, |
| "learning_rate": 0.0002936774054519987, |
| "loss": 3.3356, |
| "step": 47500 |
| }, |
| { |
| "epoch": 5.1178559896674205, |
| "grad_norm": 0.738702654838562, |
| "learning_rate": 0.0002933541644219373, |
| "loss": 3.3095, |
| "step": 47550 |
| }, |
| { |
| "epoch": 5.123237541707028, |
| "grad_norm": 0.6973130106925964, |
| "learning_rate": 0.0002930309233918758, |
| "loss": 3.3263, |
| "step": 47600 |
| }, |
| { |
| "epoch": 5.128619093746637, |
| "grad_norm": 0.6830120086669922, |
| "learning_rate": 0.0002927076823618144, |
| "loss": 3.3239, |
| "step": 47650 |
| }, |
| { |
| "epoch": 5.134000645786244, |
| "grad_norm": 0.6565729379653931, |
| "learning_rate": 0.0002923909061523542, |
| "loss": 3.3204, |
| "step": 47700 |
| }, |
| { |
| "epoch": 5.139382197825853, |
| "grad_norm": 0.7066057324409485, |
| "learning_rate": 0.0002920676651222928, |
| "loss": 3.3097, |
| "step": 47750 |
| }, |
| { |
| "epoch": 5.1447637498654615, |
| "grad_norm": 0.6757489442825317, |
| "learning_rate": 0.0002917444240922314, |
| "loss": 3.3308, |
| "step": 47800 |
| }, |
| { |
| "epoch": 5.150145301905069, |
| "grad_norm": 0.6399307250976562, |
| "learning_rate": 0.00029142118306216996, |
| "loss": 3.3083, |
| "step": 47850 |
| }, |
| { |
| "epoch": 5.155526853944678, |
| "grad_norm": 0.6833632588386536, |
| "learning_rate": 0.0002910979420321086, |
| "loss": 3.3093, |
| "step": 47900 |
| }, |
| { |
| "epoch": 5.160908405984286, |
| "grad_norm": 0.6751337051391602, |
| "learning_rate": 0.00029077470100204715, |
| "loss": 3.3321, |
| "step": 47950 |
| }, |
| { |
| "epoch": 5.166289958023894, |
| "grad_norm": 0.6862357258796692, |
| "learning_rate": 0.00029045145997198574, |
| "loss": 3.3207, |
| "step": 48000 |
| }, |
| { |
| "epoch": 5.166289958023894, |
| "eval_accuracy": 0.3803372607368339, |
| "eval_loss": 3.4145004749298096, |
| "eval_runtime": 184.805, |
| "eval_samples_per_second": 97.459, |
| "eval_steps_per_second": 6.093, |
| "step": 48000 |
| }, |
| { |
| "epoch": 5.1716715100635025, |
| "grad_norm": 0.6574402451515198, |
| "learning_rate": 0.00029012821894192433, |
| "loss": 3.3211, |
| "step": 48050 |
| }, |
| { |
| "epoch": 5.17705306210311, |
| "grad_norm": 0.691375732421875, |
| "learning_rate": 0.00028980497791186293, |
| "loss": 3.3103, |
| "step": 48100 |
| }, |
| { |
| "epoch": 5.182434614142719, |
| "grad_norm": 0.6582377552986145, |
| "learning_rate": 0.0002894817368818015, |
| "loss": 3.3237, |
| "step": 48150 |
| }, |
| { |
| "epoch": 5.187816166182327, |
| "grad_norm": 0.74177485704422, |
| "learning_rate": 0.00028915849585174006, |
| "loss": 3.3264, |
| "step": 48200 |
| }, |
| { |
| "epoch": 5.193197718221935, |
| "grad_norm": 0.6446400284767151, |
| "learning_rate": 0.00028883525482167866, |
| "loss": 3.3225, |
| "step": 48250 |
| }, |
| { |
| "epoch": 5.198579270261543, |
| "grad_norm": 0.6987661719322205, |
| "learning_rate": 0.00028851201379161725, |
| "loss": 3.3068, |
| "step": 48300 |
| }, |
| { |
| "epoch": 5.203960822301152, |
| "grad_norm": 0.6338952779769897, |
| "learning_rate": 0.00028818877276155585, |
| "loss": 3.3243, |
| "step": 48350 |
| }, |
| { |
| "epoch": 5.20934237434076, |
| "grad_norm": 0.6728731989860535, |
| "learning_rate": 0.0002878655317314944, |
| "loss": 3.3331, |
| "step": 48400 |
| }, |
| { |
| "epoch": 5.214723926380368, |
| "grad_norm": 0.7184383869171143, |
| "learning_rate": 0.00028754229070143304, |
| "loss": 3.3179, |
| "step": 48450 |
| }, |
| { |
| "epoch": 5.220105478419977, |
| "grad_norm": 0.6783042550086975, |
| "learning_rate": 0.0002872190496713716, |
| "loss": 3.3045, |
| "step": 48500 |
| }, |
| { |
| "epoch": 5.225487030459584, |
| "grad_norm": 0.6732574701309204, |
| "learning_rate": 0.0002868958086413102, |
| "loss": 3.3205, |
| "step": 48550 |
| }, |
| { |
| "epoch": 5.230868582499193, |
| "grad_norm": 0.7296816110610962, |
| "learning_rate": 0.00028657256761124877, |
| "loss": 3.3154, |
| "step": 48600 |
| }, |
| { |
| "epoch": 5.236250134538801, |
| "grad_norm": 0.6376650929450989, |
| "learning_rate": 0.00028624932658118736, |
| "loss": 3.3214, |
| "step": 48650 |
| }, |
| { |
| "epoch": 5.241631686578409, |
| "grad_norm": 0.6608768105506897, |
| "learning_rate": 0.0002859260855511259, |
| "loss": 3.3387, |
| "step": 48700 |
| }, |
| { |
| "epoch": 5.247013238618018, |
| "grad_norm": 0.6965283155441284, |
| "learning_rate": 0.0002856028445210645, |
| "loss": 3.3452, |
| "step": 48750 |
| }, |
| { |
| "epoch": 5.252394790657625, |
| "grad_norm": 0.6676567792892456, |
| "learning_rate": 0.0002852796034910031, |
| "loss": 3.3208, |
| "step": 48800 |
| }, |
| { |
| "epoch": 5.257776342697234, |
| "grad_norm": 0.6948692202568054, |
| "learning_rate": 0.0002849563624609417, |
| "loss": 3.3475, |
| "step": 48850 |
| }, |
| { |
| "epoch": 5.2631578947368425, |
| "grad_norm": 0.6433426737785339, |
| "learning_rate": 0.0002846331214308803, |
| "loss": 3.333, |
| "step": 48900 |
| }, |
| { |
| "epoch": 5.26853944677645, |
| "grad_norm": 0.6406604647636414, |
| "learning_rate": 0.0002843098804008188, |
| "loss": 3.3265, |
| "step": 48950 |
| }, |
| { |
| "epoch": 5.273920998816059, |
| "grad_norm": 0.7309980988502502, |
| "learning_rate": 0.00028398663937075747, |
| "loss": 3.3359, |
| "step": 49000 |
| }, |
| { |
| "epoch": 5.273920998816059, |
| "eval_accuracy": 0.380725477504995, |
| "eval_loss": 3.4107563495635986, |
| "eval_runtime": 184.8872, |
| "eval_samples_per_second": 97.416, |
| "eval_steps_per_second": 6.09, |
| "step": 49000 |
| }, |
| { |
| "epoch": 5.279302550855666, |
| "grad_norm": 0.6788413524627686, |
| "learning_rate": 0.000283663398340696, |
| "loss": 3.3424, |
| "step": 49050 |
| }, |
| { |
| "epoch": 5.284684102895275, |
| "grad_norm": 0.7102241516113281, |
| "learning_rate": 0.0002833401573106346, |
| "loss": 3.3221, |
| "step": 49100 |
| }, |
| { |
| "epoch": 5.2900656549348835, |
| "grad_norm": 0.675389289855957, |
| "learning_rate": 0.0002830169162805732, |
| "loss": 3.3225, |
| "step": 49150 |
| }, |
| { |
| "epoch": 5.295447206974491, |
| "grad_norm": 0.6687794327735901, |
| "learning_rate": 0.0002826936752505118, |
| "loss": 3.3299, |
| "step": 49200 |
| }, |
| { |
| "epoch": 5.3008287590141, |
| "grad_norm": 0.7226663827896118, |
| "learning_rate": 0.00028237043422045034, |
| "loss": 3.3284, |
| "step": 49250 |
| }, |
| { |
| "epoch": 5.306210311053708, |
| "grad_norm": 0.6959290504455566, |
| "learning_rate": 0.00028204719319038893, |
| "loss": 3.3362, |
| "step": 49300 |
| }, |
| { |
| "epoch": 5.311591863093316, |
| "grad_norm": 0.7100611925125122, |
| "learning_rate": 0.0002817239521603275, |
| "loss": 3.3298, |
| "step": 49350 |
| }, |
| { |
| "epoch": 5.316973415132924, |
| "grad_norm": 0.6422631144523621, |
| "learning_rate": 0.0002814007111302661, |
| "loss": 3.3177, |
| "step": 49400 |
| }, |
| { |
| "epoch": 5.322354967172533, |
| "grad_norm": 0.7156828045845032, |
| "learning_rate": 0.0002810774701002047, |
| "loss": 3.3182, |
| "step": 49450 |
| }, |
| { |
| "epoch": 5.327736519212141, |
| "grad_norm": 0.6757283806800842, |
| "learning_rate": 0.00028075422907014325, |
| "loss": 3.3563, |
| "step": 49500 |
| }, |
| { |
| "epoch": 5.333118071251749, |
| "grad_norm": 0.7245064973831177, |
| "learning_rate": 0.00028043098804008185, |
| "loss": 3.3438, |
| "step": 49550 |
| }, |
| { |
| "epoch": 5.338499623291357, |
| "grad_norm": 0.7414877414703369, |
| "learning_rate": 0.00028010774701002044, |
| "loss": 3.3352, |
| "step": 49600 |
| }, |
| { |
| "epoch": 5.343881175330965, |
| "grad_norm": 0.7045677900314331, |
| "learning_rate": 0.00027978450597995904, |
| "loss": 3.3261, |
| "step": 49650 |
| }, |
| { |
| "epoch": 5.349262727370574, |
| "grad_norm": 0.6863226294517517, |
| "learning_rate": 0.00027946772977049885, |
| "loss": 3.3259, |
| "step": 49700 |
| }, |
| { |
| "epoch": 5.354644279410182, |
| "grad_norm": 0.7018098831176758, |
| "learning_rate": 0.00027914448874043744, |
| "loss": 3.3174, |
| "step": 49750 |
| }, |
| { |
| "epoch": 5.36002583144979, |
| "grad_norm": 0.6944161057472229, |
| "learning_rate": 0.00027882124771037603, |
| "loss": 3.3304, |
| "step": 49800 |
| }, |
| { |
| "epoch": 5.365407383489399, |
| "grad_norm": 0.6603707671165466, |
| "learning_rate": 0.0002784980066803146, |
| "loss": 3.3176, |
| "step": 49850 |
| }, |
| { |
| "epoch": 5.370788935529006, |
| "grad_norm": 0.7112104892730713, |
| "learning_rate": 0.00027817476565025317, |
| "loss": 3.3343, |
| "step": 49900 |
| }, |
| { |
| "epoch": 5.376170487568615, |
| "grad_norm": 0.6899374127388, |
| "learning_rate": 0.00027785152462019176, |
| "loss": 3.34, |
| "step": 49950 |
| }, |
| { |
| "epoch": 5.3815520396082235, |
| "grad_norm": 0.7424962520599365, |
| "learning_rate": 0.00027752828359013036, |
| "loss": 3.3431, |
| "step": 50000 |
| }, |
| { |
| "epoch": 5.3815520396082235, |
| "eval_accuracy": 0.38131611460315457, |
| "eval_loss": 3.4052765369415283, |
| "eval_runtime": 184.8419, |
| "eval_samples_per_second": 97.44, |
| "eval_steps_per_second": 6.092, |
| "step": 50000 |
| }, |
| { |
| "epoch": 5.386933591647831, |
| "grad_norm": 0.694149911403656, |
| "learning_rate": 0.0002772050425600689, |
| "loss": 3.3057, |
| "step": 50050 |
| }, |
| { |
| "epoch": 5.39231514368744, |
| "grad_norm": 0.7085953950881958, |
| "learning_rate": 0.00027688180153000755, |
| "loss": 3.3414, |
| "step": 50100 |
| }, |
| { |
| "epoch": 5.397696695727047, |
| "grad_norm": 0.691828727722168, |
| "learning_rate": 0.0002765585604999461, |
| "loss": 3.3414, |
| "step": 50150 |
| }, |
| { |
| "epoch": 5.403078247766656, |
| "grad_norm": 0.7001767158508301, |
| "learning_rate": 0.0002762353194698847, |
| "loss": 3.3379, |
| "step": 50200 |
| }, |
| { |
| "epoch": 5.4084597998062645, |
| "grad_norm": 0.676909327507019, |
| "learning_rate": 0.0002759120784398233, |
| "loss": 3.3329, |
| "step": 50250 |
| }, |
| { |
| "epoch": 5.413841351845872, |
| "grad_norm": 0.7085362076759338, |
| "learning_rate": 0.00027558883740976187, |
| "loss": 3.3346, |
| "step": 50300 |
| }, |
| { |
| "epoch": 5.419222903885481, |
| "grad_norm": 0.7403079867362976, |
| "learning_rate": 0.0002752655963797004, |
| "loss": 3.3504, |
| "step": 50350 |
| }, |
| { |
| "epoch": 5.424604455925088, |
| "grad_norm": 0.6892008781433105, |
| "learning_rate": 0.000274942355349639, |
| "loss": 3.343, |
| "step": 50400 |
| }, |
| { |
| "epoch": 5.429986007964697, |
| "grad_norm": 0.6684857606887817, |
| "learning_rate": 0.0002746191143195776, |
| "loss": 3.3485, |
| "step": 50450 |
| }, |
| { |
| "epoch": 5.435367560004305, |
| "grad_norm": 0.7241749167442322, |
| "learning_rate": 0.0002742958732895162, |
| "loss": 3.3344, |
| "step": 50500 |
| }, |
| { |
| "epoch": 5.440749112043913, |
| "grad_norm": 0.6599351167678833, |
| "learning_rate": 0.0002739726322594548, |
| "loss": 3.3182, |
| "step": 50550 |
| }, |
| { |
| "epoch": 5.446130664083522, |
| "grad_norm": 0.6936551928520203, |
| "learning_rate": 0.00027364939122939333, |
| "loss": 3.321, |
| "step": 50600 |
| }, |
| { |
| "epoch": 5.45151221612313, |
| "grad_norm": 0.6684977412223816, |
| "learning_rate": 0.0002733261501993319, |
| "loss": 3.3305, |
| "step": 50650 |
| }, |
| { |
| "epoch": 5.456893768162738, |
| "grad_norm": 0.6817634701728821, |
| "learning_rate": 0.0002730029091692705, |
| "loss": 3.339, |
| "step": 50700 |
| }, |
| { |
| "epoch": 5.462275320202346, |
| "grad_norm": 0.6818822026252747, |
| "learning_rate": 0.0002726796681392091, |
| "loss": 3.3326, |
| "step": 50750 |
| }, |
| { |
| "epoch": 5.467656872241955, |
| "grad_norm": 0.6656324863433838, |
| "learning_rate": 0.0002723564271091477, |
| "loss": 3.3483, |
| "step": 50800 |
| }, |
| { |
| "epoch": 5.473038424281563, |
| "grad_norm": 0.6781908869743347, |
| "learning_rate": 0.0002720331860790863, |
| "loss": 3.333, |
| "step": 50850 |
| }, |
| { |
| "epoch": 5.478419976321171, |
| "grad_norm": 0.7217748165130615, |
| "learning_rate": 0.00027170994504902485, |
| "loss": 3.3446, |
| "step": 50900 |
| }, |
| { |
| "epoch": 5.483801528360779, |
| "grad_norm": 0.7343464493751526, |
| "learning_rate": 0.00027138670401896344, |
| "loss": 3.355, |
| "step": 50950 |
| }, |
| { |
| "epoch": 5.489183080400387, |
| "grad_norm": 0.6531841158866882, |
| "learning_rate": 0.00027106346298890204, |
| "loss": 3.3256, |
| "step": 51000 |
| }, |
| { |
| "epoch": 5.489183080400387, |
| "eval_accuracy": 0.3820636464713182, |
| "eval_loss": 3.3998801708221436, |
| "eval_runtime": 184.5563, |
| "eval_samples_per_second": 97.591, |
| "eval_steps_per_second": 6.101, |
| "step": 51000 |
| }, |
| { |
| "epoch": 5.494564632439996, |
| "grad_norm": 0.6916175484657288, |
| "learning_rate": 0.00027074022195884063, |
| "loss": 3.3435, |
| "step": 51050 |
| }, |
| { |
| "epoch": 5.499946184479604, |
| "grad_norm": 0.6753712296485901, |
| "learning_rate": 0.0002704169809287792, |
| "loss": 3.3426, |
| "step": 51100 |
| }, |
| { |
| "epoch": 5.505327736519212, |
| "grad_norm": 0.6608497500419617, |
| "learning_rate": 0.00027009373989871776, |
| "loss": 3.3304, |
| "step": 51150 |
| }, |
| { |
| "epoch": 5.510709288558821, |
| "grad_norm": 0.678605854511261, |
| "learning_rate": 0.00026977049886865636, |
| "loss": 3.3325, |
| "step": 51200 |
| }, |
| { |
| "epoch": 5.516090840598428, |
| "grad_norm": 0.7043835520744324, |
| "learning_rate": 0.00026944725783859495, |
| "loss": 3.3457, |
| "step": 51250 |
| }, |
| { |
| "epoch": 5.521472392638037, |
| "grad_norm": 0.7066404223442078, |
| "learning_rate": 0.00026912401680853355, |
| "loss": 3.3602, |
| "step": 51300 |
| }, |
| { |
| "epoch": 5.5268539446776455, |
| "grad_norm": 0.6896523833274841, |
| "learning_rate": 0.0002688007757784721, |
| "loss": 3.3234, |
| "step": 51350 |
| }, |
| { |
| "epoch": 5.532235496717253, |
| "grad_norm": 0.6568740606307983, |
| "learning_rate": 0.00026847753474841074, |
| "loss": 3.3371, |
| "step": 51400 |
| }, |
| { |
| "epoch": 5.537617048756862, |
| "grad_norm": 0.7369204163551331, |
| "learning_rate": 0.0002681542937183493, |
| "loss": 3.3357, |
| "step": 51450 |
| }, |
| { |
| "epoch": 5.542998600796469, |
| "grad_norm": 0.687563419342041, |
| "learning_rate": 0.0002678310526882879, |
| "loss": 3.3575, |
| "step": 51500 |
| }, |
| { |
| "epoch": 5.548380152836078, |
| "grad_norm": 0.6815269589424133, |
| "learning_rate": 0.00026750781165822647, |
| "loss": 3.3286, |
| "step": 51550 |
| }, |
| { |
| "epoch": 5.553761704875686, |
| "grad_norm": 0.7188419699668884, |
| "learning_rate": 0.00026718457062816506, |
| "loss": 3.3391, |
| "step": 51600 |
| }, |
| { |
| "epoch": 5.559143256915294, |
| "grad_norm": 0.7285634279251099, |
| "learning_rate": 0.00026686132959810366, |
| "loss": 3.3365, |
| "step": 51650 |
| }, |
| { |
| "epoch": 5.564524808954903, |
| "grad_norm": 0.6750437617301941, |
| "learning_rate": 0.0002665445533886434, |
| "loss": 3.3578, |
| "step": 51700 |
| }, |
| { |
| "epoch": 5.569906360994511, |
| "grad_norm": 0.6828053593635559, |
| "learning_rate": 0.00026622131235858206, |
| "loss": 3.327, |
| "step": 51750 |
| }, |
| { |
| "epoch": 5.575287913034119, |
| "grad_norm": 0.7370911836624146, |
| "learning_rate": 0.0002658980713285206, |
| "loss": 3.3642, |
| "step": 51800 |
| }, |
| { |
| "epoch": 5.580669465073727, |
| "grad_norm": 0.7274966835975647, |
| "learning_rate": 0.0002655748302984592, |
| "loss": 3.3163, |
| "step": 51850 |
| }, |
| { |
| "epoch": 5.586051017113336, |
| "grad_norm": 0.724192202091217, |
| "learning_rate": 0.0002652515892683978, |
| "loss": 3.3457, |
| "step": 51900 |
| }, |
| { |
| "epoch": 5.591432569152944, |
| "grad_norm": 0.7059313058853149, |
| "learning_rate": 0.0002649283482383364, |
| "loss": 3.3399, |
| "step": 51950 |
| }, |
| { |
| "epoch": 5.596814121192552, |
| "grad_norm": 0.6701476573944092, |
| "learning_rate": 0.0002646051072082749, |
| "loss": 3.3417, |
| "step": 52000 |
| }, |
| { |
| "epoch": 5.596814121192552, |
| "eval_accuracy": 0.38217947044972844, |
| "eval_loss": 3.397221088409424, |
| "eval_runtime": 184.8285, |
| "eval_samples_per_second": 97.447, |
| "eval_steps_per_second": 6.092, |
| "step": 52000 |
| }, |
| { |
| "epoch": 5.60219567323216, |
| "grad_norm": 0.7342854738235474, |
| "learning_rate": 0.0002642818661782135, |
| "loss": 3.3522, |
| "step": 52050 |
| }, |
| { |
| "epoch": 5.607577225271768, |
| "grad_norm": 0.7041910886764526, |
| "learning_rate": 0.0002639586251481521, |
| "loss": 3.3441, |
| "step": 52100 |
| }, |
| { |
| "epoch": 5.612958777311377, |
| "grad_norm": 0.6980746984481812, |
| "learning_rate": 0.0002636353841180907, |
| "loss": 3.3438, |
| "step": 52150 |
| }, |
| { |
| "epoch": 5.618340329350985, |
| "grad_norm": 0.7169170379638672, |
| "learning_rate": 0.0002633121430880293, |
| "loss": 3.3183, |
| "step": 52200 |
| }, |
| { |
| "epoch": 5.623721881390593, |
| "grad_norm": 0.6377230286598206, |
| "learning_rate": 0.00026298890205796784, |
| "loss": 3.3293, |
| "step": 52250 |
| }, |
| { |
| "epoch": 5.629103433430201, |
| "grad_norm": 0.6815519332885742, |
| "learning_rate": 0.00026266566102790644, |
| "loss": 3.3181, |
| "step": 52300 |
| }, |
| { |
| "epoch": 5.634484985469809, |
| "grad_norm": 0.660933256149292, |
| "learning_rate": 0.00026234241999784503, |
| "loss": 3.3399, |
| "step": 52350 |
| }, |
| { |
| "epoch": 5.639866537509418, |
| "grad_norm": 0.6373748779296875, |
| "learning_rate": 0.0002620191789677836, |
| "loss": 3.3355, |
| "step": 52400 |
| }, |
| { |
| "epoch": 5.645248089549026, |
| "grad_norm": 0.6971507668495178, |
| "learning_rate": 0.00026169593793772217, |
| "loss": 3.341, |
| "step": 52450 |
| }, |
| { |
| "epoch": 5.650629641588634, |
| "grad_norm": 0.7280732989311218, |
| "learning_rate": 0.0002613726969076608, |
| "loss": 3.356, |
| "step": 52500 |
| }, |
| { |
| "epoch": 5.656011193628243, |
| "grad_norm": 0.6992477178573608, |
| "learning_rate": 0.00026104945587759936, |
| "loss": 3.337, |
| "step": 52550 |
| }, |
| { |
| "epoch": 5.66139274566785, |
| "grad_norm": 0.696068525314331, |
| "learning_rate": 0.00026072621484753795, |
| "loss": 3.3223, |
| "step": 52600 |
| }, |
| { |
| "epoch": 5.666774297707459, |
| "grad_norm": 0.7363685965538025, |
| "learning_rate": 0.00026040297381747655, |
| "loss": 3.3556, |
| "step": 52650 |
| }, |
| { |
| "epoch": 5.672155849747067, |
| "grad_norm": 0.682386040687561, |
| "learning_rate": 0.00026007973278741514, |
| "loss": 3.3368, |
| "step": 52700 |
| }, |
| { |
| "epoch": 5.677537401786675, |
| "grad_norm": 0.6552140116691589, |
| "learning_rate": 0.00025975649175735373, |
| "loss": 3.3458, |
| "step": 52750 |
| }, |
| { |
| "epoch": 5.682918953826284, |
| "grad_norm": 0.7018849849700928, |
| "learning_rate": 0.0002594332507272923, |
| "loss": 3.3377, |
| "step": 52800 |
| }, |
| { |
| "epoch": 5.688300505865891, |
| "grad_norm": 0.7437209486961365, |
| "learning_rate": 0.00025911000969723087, |
| "loss": 3.3348, |
| "step": 52850 |
| }, |
| { |
| "epoch": 5.6936820579055, |
| "grad_norm": 0.7024593949317932, |
| "learning_rate": 0.00025878676866716946, |
| "loss": 3.3426, |
| "step": 52900 |
| }, |
| { |
| "epoch": 5.699063609945108, |
| "grad_norm": 0.6867077350616455, |
| "learning_rate": 0.00025846352763710806, |
| "loss": 3.3245, |
| "step": 52950 |
| }, |
| { |
| "epoch": 5.704445161984716, |
| "grad_norm": 0.7076590061187744, |
| "learning_rate": 0.0002581402866070466, |
| "loss": 3.3248, |
| "step": 53000 |
| }, |
| { |
| "epoch": 5.704445161984716, |
| "eval_accuracy": 0.38262896744661695, |
| "eval_loss": 3.3937745094299316, |
| "eval_runtime": 184.9379, |
| "eval_samples_per_second": 97.389, |
| "eval_steps_per_second": 6.089, |
| "step": 53000 |
| }, |
| { |
| "epoch": 5.709826714024325, |
| "grad_norm": 0.8061707019805908, |
| "learning_rate": 0.00025781704557698525, |
| "loss": 3.337, |
| "step": 53050 |
| }, |
| { |
| "epoch": 5.715208266063933, |
| "grad_norm": 0.7249623537063599, |
| "learning_rate": 0.0002574938045469238, |
| "loss": 3.3559, |
| "step": 53100 |
| }, |
| { |
| "epoch": 5.720589818103541, |
| "grad_norm": 0.7043891549110413, |
| "learning_rate": 0.0002571705635168624, |
| "loss": 3.3466, |
| "step": 53150 |
| }, |
| { |
| "epoch": 5.725971370143149, |
| "grad_norm": 0.706001877784729, |
| "learning_rate": 0.000256847322486801, |
| "loss": 3.3386, |
| "step": 53200 |
| }, |
| { |
| "epoch": 5.731352922182758, |
| "grad_norm": 0.7397270798683167, |
| "learning_rate": 0.0002565240814567396, |
| "loss": 3.336, |
| "step": 53250 |
| }, |
| { |
| "epoch": 5.736734474222366, |
| "grad_norm": 0.7014256715774536, |
| "learning_rate": 0.0002562008404266781, |
| "loss": 3.3473, |
| "step": 53300 |
| }, |
| { |
| "epoch": 5.742116026261974, |
| "grad_norm": 0.7000493407249451, |
| "learning_rate": 0.0002558775993966167, |
| "loss": 3.3464, |
| "step": 53350 |
| }, |
| { |
| "epoch": 5.747497578301582, |
| "grad_norm": 0.6832889318466187, |
| "learning_rate": 0.0002555543583665553, |
| "loss": 3.3479, |
| "step": 53400 |
| }, |
| { |
| "epoch": 5.75287913034119, |
| "grad_norm": 0.6802942156791687, |
| "learning_rate": 0.0002552311173364939, |
| "loss": 3.3327, |
| "step": 53450 |
| }, |
| { |
| "epoch": 5.758260682380799, |
| "grad_norm": 0.6792773604393005, |
| "learning_rate": 0.0002549078763064325, |
| "loss": 3.3537, |
| "step": 53500 |
| }, |
| { |
| "epoch": 5.763642234420407, |
| "grad_norm": 0.7272350788116455, |
| "learning_rate": 0.00025458463527637103, |
| "loss": 3.3429, |
| "step": 53550 |
| }, |
| { |
| "epoch": 5.769023786460015, |
| "grad_norm": 0.6929247975349426, |
| "learning_rate": 0.0002542613942463097, |
| "loss": 3.3457, |
| "step": 53600 |
| }, |
| { |
| "epoch": 5.774405338499624, |
| "grad_norm": 0.6766125559806824, |
| "learning_rate": 0.0002539381532162482, |
| "loss": 3.334, |
| "step": 53650 |
| }, |
| { |
| "epoch": 5.779786890539231, |
| "grad_norm": 0.6857972741127014, |
| "learning_rate": 0.00025362137700678803, |
| "loss": 3.3444, |
| "step": 53700 |
| }, |
| { |
| "epoch": 5.78516844257884, |
| "grad_norm": 0.6698341369628906, |
| "learning_rate": 0.0002532981359767266, |
| "loss": 3.3559, |
| "step": 53750 |
| }, |
| { |
| "epoch": 5.790549994618448, |
| "grad_norm": 0.6908291578292847, |
| "learning_rate": 0.0002529748949466652, |
| "loss": 3.3184, |
| "step": 53800 |
| }, |
| { |
| "epoch": 5.795931546658056, |
| "grad_norm": 0.6863468289375305, |
| "learning_rate": 0.0002526516539166038, |
| "loss": 3.3631, |
| "step": 53850 |
| }, |
| { |
| "epoch": 5.801313098697665, |
| "grad_norm": 0.7525630593299866, |
| "learning_rate": 0.00025232841288654235, |
| "loss": 3.3399, |
| "step": 53900 |
| }, |
| { |
| "epoch": 5.806694650737272, |
| "grad_norm": 0.7373597025871277, |
| "learning_rate": 0.00025200517185648095, |
| "loss": 3.343, |
| "step": 53950 |
| }, |
| { |
| "epoch": 5.812076202776881, |
| "grad_norm": 0.7358316779136658, |
| "learning_rate": 0.00025168193082641954, |
| "loss": 3.3469, |
| "step": 54000 |
| }, |
| { |
| "epoch": 5.812076202776881, |
| "eval_accuracy": 0.38326252243546316, |
| "eval_loss": 3.3860888481140137, |
| "eval_runtime": 184.3203, |
| "eval_samples_per_second": 97.716, |
| "eval_steps_per_second": 6.109, |
| "step": 54000 |
| }, |
| { |
| "epoch": 5.817457754816489, |
| "grad_norm": 0.6819138526916504, |
| "learning_rate": 0.00025135868979635814, |
| "loss": 3.3147, |
| "step": 54050 |
| }, |
| { |
| "epoch": 5.822839306856097, |
| "grad_norm": 0.7241456508636475, |
| "learning_rate": 0.0002510354487662967, |
| "loss": 3.3323, |
| "step": 54100 |
| }, |
| { |
| "epoch": 5.828220858895706, |
| "grad_norm": 0.6862399578094482, |
| "learning_rate": 0.0002507122077362353, |
| "loss": 3.3485, |
| "step": 54150 |
| }, |
| { |
| "epoch": 5.833602410935313, |
| "grad_norm": 0.7358222007751465, |
| "learning_rate": 0.00025038896670617387, |
| "loss": 3.3481, |
| "step": 54200 |
| }, |
| { |
| "epoch": 5.838983962974922, |
| "grad_norm": 0.6974659562110901, |
| "learning_rate": 0.00025006572567611246, |
| "loss": 3.3291, |
| "step": 54250 |
| }, |
| { |
| "epoch": 5.84436551501453, |
| "grad_norm": 0.7777621746063232, |
| "learning_rate": 0.00024974248464605106, |
| "loss": 3.3385, |
| "step": 54300 |
| }, |
| { |
| "epoch": 5.849747067054138, |
| "grad_norm": 0.727279543876648, |
| "learning_rate": 0.00024941924361598965, |
| "loss": 3.3501, |
| "step": 54350 |
| }, |
| { |
| "epoch": 5.855128619093747, |
| "grad_norm": 0.6796432137489319, |
| "learning_rate": 0.00024909600258592825, |
| "loss": 3.3435, |
| "step": 54400 |
| }, |
| { |
| "epoch": 5.860510171133355, |
| "grad_norm": 0.7217870950698853, |
| "learning_rate": 0.0002487727615558668, |
| "loss": 3.3396, |
| "step": 54450 |
| }, |
| { |
| "epoch": 5.865891723172963, |
| "grad_norm": 0.744452714920044, |
| "learning_rate": 0.0002484495205258054, |
| "loss": 3.336, |
| "step": 54500 |
| }, |
| { |
| "epoch": 5.871273275212571, |
| "grad_norm": 0.7230072617530823, |
| "learning_rate": 0.000248126279495744, |
| "loss": 3.3416, |
| "step": 54550 |
| }, |
| { |
| "epoch": 5.87665482725218, |
| "grad_norm": 0.7871853113174438, |
| "learning_rate": 0.00024780303846568257, |
| "loss": 3.3303, |
| "step": 54600 |
| }, |
| { |
| "epoch": 5.882036379291788, |
| "grad_norm": 0.7021961212158203, |
| "learning_rate": 0.0002474797974356211, |
| "loss": 3.3474, |
| "step": 54650 |
| }, |
| { |
| "epoch": 5.887417931331396, |
| "grad_norm": 0.7060352563858032, |
| "learning_rate": 0.00024715655640555976, |
| "loss": 3.3321, |
| "step": 54700 |
| }, |
| { |
| "epoch": 5.892799483371004, |
| "grad_norm": 0.71165931224823, |
| "learning_rate": 0.0002468333153754983, |
| "loss": 3.3456, |
| "step": 54750 |
| }, |
| { |
| "epoch": 5.898181035410612, |
| "grad_norm": 0.683138906955719, |
| "learning_rate": 0.0002465100743454369, |
| "loss": 3.3308, |
| "step": 54800 |
| }, |
| { |
| "epoch": 5.903562587450221, |
| "grad_norm": 0.7074604034423828, |
| "learning_rate": 0.0002461868333153755, |
| "loss": 3.3447, |
| "step": 54850 |
| }, |
| { |
| "epoch": 5.9089441394898286, |
| "grad_norm": 0.722017765045166, |
| "learning_rate": 0.0002458635922853141, |
| "loss": 3.3308, |
| "step": 54900 |
| }, |
| { |
| "epoch": 5.914325691529437, |
| "grad_norm": 0.7072309255599976, |
| "learning_rate": 0.0002455403512552526, |
| "loss": 3.3635, |
| "step": 54950 |
| }, |
| { |
| "epoch": 5.919707243569046, |
| "grad_norm": 0.7123068571090698, |
| "learning_rate": 0.0002452171102251912, |
| "loss": 3.3445, |
| "step": 55000 |
| }, |
| { |
| "epoch": 5.919707243569046, |
| "eval_accuracy": 0.38317081939814773, |
| "eval_loss": 3.3848025798797607, |
| "eval_runtime": 184.0505, |
| "eval_samples_per_second": 97.859, |
| "eval_steps_per_second": 6.118, |
| "step": 55000 |
| }, |
| { |
| "epoch": 5.925088795608653, |
| "grad_norm": 0.7227998971939087, |
| "learning_rate": 0.0002448938691951298, |
| "loss": 3.3381, |
| "step": 55050 |
| }, |
| { |
| "epoch": 5.930470347648262, |
| "grad_norm": 0.6625151634216309, |
| "learning_rate": 0.0002445706281650684, |
| "loss": 3.3566, |
| "step": 55100 |
| }, |
| { |
| "epoch": 5.93585189968787, |
| "grad_norm": 0.6930627226829529, |
| "learning_rate": 0.000244247387135007, |
| "loss": 3.3258, |
| "step": 55150 |
| }, |
| { |
| "epoch": 5.941233451727478, |
| "grad_norm": 0.7274208664894104, |
| "learning_rate": 0.00024392414610494557, |
| "loss": 3.327, |
| "step": 55200 |
| }, |
| { |
| "epoch": 5.946615003767087, |
| "grad_norm": 0.7061982154846191, |
| "learning_rate": 0.00024360090507488414, |
| "loss": 3.3288, |
| "step": 55250 |
| }, |
| { |
| "epoch": 5.951996555806694, |
| "grad_norm": 0.684239387512207, |
| "learning_rate": 0.00024327766404482273, |
| "loss": 3.3363, |
| "step": 55300 |
| }, |
| { |
| "epoch": 5.957378107846303, |
| "grad_norm": 0.6876581907272339, |
| "learning_rate": 0.0002429544230147613, |
| "loss": 3.3406, |
| "step": 55350 |
| }, |
| { |
| "epoch": 5.962759659885911, |
| "grad_norm": 0.7436810731887817, |
| "learning_rate": 0.00024263118198469992, |
| "loss": 3.344, |
| "step": 55400 |
| }, |
| { |
| "epoch": 5.968141211925519, |
| "grad_norm": 0.7856045365333557, |
| "learning_rate": 0.0002423079409546385, |
| "loss": 3.3348, |
| "step": 55450 |
| }, |
| { |
| "epoch": 5.973522763965128, |
| "grad_norm": 0.7595300674438477, |
| "learning_rate": 0.00024198469992457706, |
| "loss": 3.336, |
| "step": 55500 |
| }, |
| { |
| "epoch": 5.978904316004736, |
| "grad_norm": 0.8078129887580872, |
| "learning_rate": 0.00024166145889451568, |
| "loss": 3.3391, |
| "step": 55550 |
| }, |
| { |
| "epoch": 5.984285868044344, |
| "grad_norm": 0.7111245393753052, |
| "learning_rate": 0.00024133821786445425, |
| "loss": 3.3372, |
| "step": 55600 |
| }, |
| { |
| "epoch": 5.989667420083952, |
| "grad_norm": 0.7171187996864319, |
| "learning_rate": 0.0002410149768343928, |
| "loss": 3.3492, |
| "step": 55650 |
| }, |
| { |
| "epoch": 5.995048972123561, |
| "grad_norm": 0.7238572239875793, |
| "learning_rate": 0.00024069820062493265, |
| "loss": 3.3415, |
| "step": 55700 |
| }, |
| { |
| "epoch": 6.000430524163169, |
| "grad_norm": 0.7295876741409302, |
| "learning_rate": 0.00024037495959487121, |
| "loss": 3.33, |
| "step": 55750 |
| }, |
| { |
| "epoch": 6.005812076202777, |
| "grad_norm": 0.6892274618148804, |
| "learning_rate": 0.0002400517185648098, |
| "loss": 3.247, |
| "step": 55800 |
| }, |
| { |
| "epoch": 6.011193628242385, |
| "grad_norm": 0.708806037902832, |
| "learning_rate": 0.00023972847753474838, |
| "loss": 3.2382, |
| "step": 55850 |
| }, |
| { |
| "epoch": 6.016575180281993, |
| "grad_norm": 0.732413113117218, |
| "learning_rate": 0.00023940523650468697, |
| "loss": 3.2495, |
| "step": 55900 |
| }, |
| { |
| "epoch": 6.021956732321602, |
| "grad_norm": 0.7050142288208008, |
| "learning_rate": 0.00023908199547462557, |
| "loss": 3.2437, |
| "step": 55950 |
| }, |
| { |
| "epoch": 6.0273382843612096, |
| "grad_norm": 0.7132096886634827, |
| "learning_rate": 0.00023875875444456413, |
| "loss": 3.2471, |
| "step": 56000 |
| }, |
| { |
| "epoch": 6.0273382843612096, |
| "eval_accuracy": 0.3839456231411528, |
| "eval_loss": 3.3877761363983154, |
| "eval_runtime": 184.3787, |
| "eval_samples_per_second": 97.685, |
| "eval_steps_per_second": 6.107, |
| "step": 56000 |
| }, |
| { |
| "epoch": 6.032719836400818, |
| "grad_norm": 0.6656051278114319, |
| "learning_rate": 0.0002384355134145027, |
| "loss": 3.2627, |
| "step": 56050 |
| }, |
| { |
| "epoch": 6.038101388440427, |
| "grad_norm": 0.6595709919929504, |
| "learning_rate": 0.00023811227238444132, |
| "loss": 3.2699, |
| "step": 56100 |
| }, |
| { |
| "epoch": 6.043482940480034, |
| "grad_norm": 0.760237455368042, |
| "learning_rate": 0.0002377890313543799, |
| "loss": 3.2639, |
| "step": 56150 |
| }, |
| { |
| "epoch": 6.048864492519643, |
| "grad_norm": 0.6955994367599487, |
| "learning_rate": 0.00023746579032431849, |
| "loss": 3.2432, |
| "step": 56200 |
| }, |
| { |
| "epoch": 6.0542460445592505, |
| "grad_norm": 0.7197161912918091, |
| "learning_rate": 0.0002371490141148583, |
| "loss": 3.246, |
| "step": 56250 |
| }, |
| { |
| "epoch": 6.059627596598859, |
| "grad_norm": 0.6739785075187683, |
| "learning_rate": 0.00023682577308479689, |
| "loss": 3.274, |
| "step": 56300 |
| }, |
| { |
| "epoch": 6.065009148638468, |
| "grad_norm": 0.7220432758331299, |
| "learning_rate": 0.00023650253205473545, |
| "loss": 3.2443, |
| "step": 56350 |
| }, |
| { |
| "epoch": 6.070390700678075, |
| "grad_norm": 0.6965115666389465, |
| "learning_rate": 0.00023617929102467402, |
| "loss": 3.2513, |
| "step": 56400 |
| }, |
| { |
| "epoch": 6.075772252717684, |
| "grad_norm": 0.7645218372344971, |
| "learning_rate": 0.00023585604999461264, |
| "loss": 3.2559, |
| "step": 56450 |
| }, |
| { |
| "epoch": 6.081153804757292, |
| "grad_norm": 0.7593476176261902, |
| "learning_rate": 0.0002355328089645512, |
| "loss": 3.2585, |
| "step": 56500 |
| }, |
| { |
| "epoch": 6.0865353567969, |
| "grad_norm": 0.6853207945823669, |
| "learning_rate": 0.00023520956793448978, |
| "loss": 3.277, |
| "step": 56550 |
| }, |
| { |
| "epoch": 6.091916908836509, |
| "grad_norm": 0.680121898651123, |
| "learning_rate": 0.0002348863269044284, |
| "loss": 3.2643, |
| "step": 56600 |
| }, |
| { |
| "epoch": 6.097298460876116, |
| "grad_norm": 0.7078647017478943, |
| "learning_rate": 0.00023456308587436697, |
| "loss": 3.258, |
| "step": 56650 |
| }, |
| { |
| "epoch": 6.102680012915725, |
| "grad_norm": 0.6985390782356262, |
| "learning_rate": 0.00023423984484430554, |
| "loss": 3.2553, |
| "step": 56700 |
| }, |
| { |
| "epoch": 6.108061564955333, |
| "grad_norm": 0.7196840047836304, |
| "learning_rate": 0.00023391660381424413, |
| "loss": 3.2634, |
| "step": 56750 |
| }, |
| { |
| "epoch": 6.113443116994941, |
| "grad_norm": 0.7238477468490601, |
| "learning_rate": 0.00023359336278418272, |
| "loss": 3.2565, |
| "step": 56800 |
| }, |
| { |
| "epoch": 6.11882466903455, |
| "grad_norm": 0.7099041938781738, |
| "learning_rate": 0.0002332701217541213, |
| "loss": 3.2683, |
| "step": 56850 |
| }, |
| { |
| "epoch": 6.124206221074158, |
| "grad_norm": 0.7175658345222473, |
| "learning_rate": 0.0002329468807240599, |
| "loss": 3.275, |
| "step": 56900 |
| }, |
| { |
| "epoch": 6.129587773113766, |
| "grad_norm": 0.7077184319496155, |
| "learning_rate": 0.00023262363969399845, |
| "loss": 3.2594, |
| "step": 56950 |
| }, |
| { |
| "epoch": 6.134969325153374, |
| "grad_norm": 0.6981788873672485, |
| "learning_rate": 0.00023230039866393708, |
| "loss": 3.2697, |
| "step": 57000 |
| }, |
| { |
| "epoch": 6.134969325153374, |
| "eval_accuracy": 0.38385218165763235, |
| "eval_loss": 3.3861021995544434, |
| "eval_runtime": 184.0749, |
| "eval_samples_per_second": 97.846, |
| "eval_steps_per_second": 6.117, |
| "step": 57000 |
| }, |
| { |
| "epoch": 6.140350877192983, |
| "grad_norm": 0.7192840576171875, |
| "learning_rate": 0.00023197715763387564, |
| "loss": 3.2669, |
| "step": 57050 |
| }, |
| { |
| "epoch": 6.1457324292325906, |
| "grad_norm": 0.7335977554321289, |
| "learning_rate": 0.0002316539166038142, |
| "loss": 3.2646, |
| "step": 57100 |
| }, |
| { |
| "epoch": 6.151113981272199, |
| "grad_norm": 0.7761592268943787, |
| "learning_rate": 0.00023133067557375283, |
| "loss": 3.2544, |
| "step": 57150 |
| }, |
| { |
| "epoch": 6.156495533311807, |
| "grad_norm": 0.7091419696807861, |
| "learning_rate": 0.0002310074345436914, |
| "loss": 3.2679, |
| "step": 57200 |
| }, |
| { |
| "epoch": 6.161877085351415, |
| "grad_norm": 0.7654346823692322, |
| "learning_rate": 0.00023068419351362997, |
| "loss": 3.2741, |
| "step": 57250 |
| }, |
| { |
| "epoch": 6.167258637391024, |
| "grad_norm": 0.7190260291099548, |
| "learning_rate": 0.00023036095248356856, |
| "loss": 3.2763, |
| "step": 57300 |
| }, |
| { |
| "epoch": 6.1726401894306315, |
| "grad_norm": 0.7094807028770447, |
| "learning_rate": 0.00023003771145350716, |
| "loss": 3.2687, |
| "step": 57350 |
| }, |
| { |
| "epoch": 6.17802174147024, |
| "grad_norm": 0.6916371583938599, |
| "learning_rate": 0.00022971447042344572, |
| "loss": 3.2652, |
| "step": 57400 |
| }, |
| { |
| "epoch": 6.183403293509849, |
| "grad_norm": 0.766798734664917, |
| "learning_rate": 0.00022939122939338432, |
| "loss": 3.2899, |
| "step": 57450 |
| }, |
| { |
| "epoch": 6.188784845549456, |
| "grad_norm": 0.7512463927268982, |
| "learning_rate": 0.0002290679883633229, |
| "loss": 3.2665, |
| "step": 57500 |
| }, |
| { |
| "epoch": 6.194166397589065, |
| "grad_norm": 0.6769281625747681, |
| "learning_rate": 0.00022874474733326148, |
| "loss": 3.2634, |
| "step": 57550 |
| }, |
| { |
| "epoch": 6.1995479496286725, |
| "grad_norm": 0.6847677826881409, |
| "learning_rate": 0.00022842150630320008, |
| "loss": 3.2695, |
| "step": 57600 |
| }, |
| { |
| "epoch": 6.204929501668281, |
| "grad_norm": 0.7338321208953857, |
| "learning_rate": 0.00022809826527313864, |
| "loss": 3.2808, |
| "step": 57650 |
| }, |
| { |
| "epoch": 6.21031105370789, |
| "grad_norm": 0.7830347418785095, |
| "learning_rate": 0.0002277750242430772, |
| "loss": 3.255, |
| "step": 57700 |
| }, |
| { |
| "epoch": 6.215692605747497, |
| "grad_norm": 0.7542902827262878, |
| "learning_rate": 0.00022745178321301583, |
| "loss": 3.2679, |
| "step": 57750 |
| }, |
| { |
| "epoch": 6.221074157787106, |
| "grad_norm": 0.7709856033325195, |
| "learning_rate": 0.0002271285421829544, |
| "loss": 3.2809, |
| "step": 57800 |
| }, |
| { |
| "epoch": 6.226455709826714, |
| "grad_norm": 0.7455905675888062, |
| "learning_rate": 0.00022680530115289297, |
| "loss": 3.2939, |
| "step": 57850 |
| }, |
| { |
| "epoch": 6.231837261866322, |
| "grad_norm": 0.7592281699180603, |
| "learning_rate": 0.0002264820601228316, |
| "loss": 3.2679, |
| "step": 57900 |
| }, |
| { |
| "epoch": 6.237218813905931, |
| "grad_norm": 0.7581020593643188, |
| "learning_rate": 0.00022615881909277016, |
| "loss": 3.2665, |
| "step": 57950 |
| }, |
| { |
| "epoch": 6.242600365945538, |
| "grad_norm": 0.7076690196990967, |
| "learning_rate": 0.00022583557806270875, |
| "loss": 3.2705, |
| "step": 58000 |
| }, |
| { |
| "epoch": 6.242600365945538, |
| "eval_accuracy": 0.3839192204894139, |
| "eval_loss": 3.3843679428100586, |
| "eval_runtime": 184.5165, |
| "eval_samples_per_second": 97.612, |
| "eval_steps_per_second": 6.102, |
| "step": 58000 |
| }, |
| { |
| "epoch": 6.247981917985147, |
| "grad_norm": 0.755380392074585, |
| "learning_rate": 0.00022551233703264732, |
| "loss": 3.2734, |
| "step": 58050 |
| }, |
| { |
| "epoch": 6.253363470024755, |
| "grad_norm": 0.7202139496803284, |
| "learning_rate": 0.0002251890960025859, |
| "loss": 3.2963, |
| "step": 58100 |
| }, |
| { |
| "epoch": 6.258745022064363, |
| "grad_norm": 0.7247757911682129, |
| "learning_rate": 0.0002248658549725245, |
| "loss": 3.28, |
| "step": 58150 |
| }, |
| { |
| "epoch": 6.264126574103972, |
| "grad_norm": 0.7414571046829224, |
| "learning_rate": 0.00022454261394246308, |
| "loss": 3.2758, |
| "step": 58200 |
| }, |
| { |
| "epoch": 6.26950812614358, |
| "grad_norm": 0.6735288500785828, |
| "learning_rate": 0.00022421937291240164, |
| "loss": 3.2868, |
| "step": 58250 |
| }, |
| { |
| "epoch": 6.274889678183188, |
| "grad_norm": 0.7054392695426941, |
| "learning_rate": 0.00022389613188234027, |
| "loss": 3.2836, |
| "step": 58300 |
| }, |
| { |
| "epoch": 6.280271230222796, |
| "grad_norm": 0.710004985332489, |
| "learning_rate": 0.00022357289085227883, |
| "loss": 3.2973, |
| "step": 58350 |
| }, |
| { |
| "epoch": 6.285652782262405, |
| "grad_norm": 0.7532511949539185, |
| "learning_rate": 0.0002232496498222174, |
| "loss": 3.2859, |
| "step": 58400 |
| }, |
| { |
| "epoch": 6.2910343343020125, |
| "grad_norm": 0.7729986310005188, |
| "learning_rate": 0.000222926408792156, |
| "loss": 3.2755, |
| "step": 58450 |
| }, |
| { |
| "epoch": 6.296415886341621, |
| "grad_norm": 0.7198432683944702, |
| "learning_rate": 0.0002226031677620946, |
| "loss": 3.2798, |
| "step": 58500 |
| }, |
| { |
| "epoch": 6.301797438381229, |
| "grad_norm": 0.730705201625824, |
| "learning_rate": 0.00022227992673203316, |
| "loss": 3.2788, |
| "step": 58550 |
| }, |
| { |
| "epoch": 6.307178990420837, |
| "grad_norm": 0.6701096296310425, |
| "learning_rate": 0.00022195668570197175, |
| "loss": 3.2919, |
| "step": 58600 |
| }, |
| { |
| "epoch": 6.312560542460446, |
| "grad_norm": 0.791893720626831, |
| "learning_rate": 0.00022163344467191032, |
| "loss": 3.2969, |
| "step": 58650 |
| }, |
| { |
| "epoch": 6.3179420945000535, |
| "grad_norm": 0.7149146199226379, |
| "learning_rate": 0.00022131020364184891, |
| "loss": 3.2738, |
| "step": 58700 |
| }, |
| { |
| "epoch": 6.323323646539662, |
| "grad_norm": 0.6918314099311829, |
| "learning_rate": 0.0002209869626117875, |
| "loss": 3.2761, |
| "step": 58750 |
| }, |
| { |
| "epoch": 6.328705198579271, |
| "grad_norm": 0.7772107124328613, |
| "learning_rate": 0.00022066372158172608, |
| "loss": 3.287, |
| "step": 58800 |
| }, |
| { |
| "epoch": 6.334086750618878, |
| "grad_norm": 0.7206048369407654, |
| "learning_rate": 0.00022034048055166464, |
| "loss": 3.2821, |
| "step": 58850 |
| }, |
| { |
| "epoch": 6.339468302658487, |
| "grad_norm": 0.6840312480926514, |
| "learning_rate": 0.00022001723952160327, |
| "loss": 3.288, |
| "step": 58900 |
| }, |
| { |
| "epoch": 6.344849854698095, |
| "grad_norm": 0.7471789717674255, |
| "learning_rate": 0.00021969399849154183, |
| "loss": 3.2853, |
| "step": 58950 |
| }, |
| { |
| "epoch": 6.350231406737703, |
| "grad_norm": 0.7397593855857849, |
| "learning_rate": 0.00021937075746148043, |
| "loss": 3.2824, |
| "step": 59000 |
| }, |
| { |
| "epoch": 6.350231406737703, |
| "eval_accuracy": 0.38456961667587136, |
| "eval_loss": 3.3814733028411865, |
| "eval_runtime": 184.1551, |
| "eval_samples_per_second": 97.803, |
| "eval_steps_per_second": 6.114, |
| "step": 59000 |
| }, |
| { |
| "epoch": 6.355612958777312, |
| "grad_norm": 0.7210568189620972, |
| "learning_rate": 0.00021904751643141902, |
| "loss": 3.2614, |
| "step": 59050 |
| }, |
| { |
| "epoch": 6.360994510816919, |
| "grad_norm": 0.6994883418083191, |
| "learning_rate": 0.0002187242754013576, |
| "loss": 3.2758, |
| "step": 59100 |
| }, |
| { |
| "epoch": 6.366376062856528, |
| "grad_norm": 0.7228302359580994, |
| "learning_rate": 0.00021840103437129619, |
| "loss": 3.3003, |
| "step": 59150 |
| }, |
| { |
| "epoch": 6.371757614896136, |
| "grad_norm": 0.7233319282531738, |
| "learning_rate": 0.00021807779334123475, |
| "loss": 3.2775, |
| "step": 59200 |
| }, |
| { |
| "epoch": 6.377139166935744, |
| "grad_norm": 0.7645636796951294, |
| "learning_rate": 0.00021775455231117335, |
| "loss": 3.2802, |
| "step": 59250 |
| }, |
| { |
| "epoch": 6.382520718975353, |
| "grad_norm": 0.7440745234489441, |
| "learning_rate": 0.00021743131128111194, |
| "loss": 3.2961, |
| "step": 59300 |
| }, |
| { |
| "epoch": 6.387902271014961, |
| "grad_norm": 0.762554943561554, |
| "learning_rate": 0.00021711453507165172, |
| "loss": 3.2985, |
| "step": 59350 |
| }, |
| { |
| "epoch": 6.393283823054569, |
| "grad_norm": 0.7303590774536133, |
| "learning_rate": 0.00021679129404159034, |
| "loss": 3.2798, |
| "step": 59400 |
| }, |
| { |
| "epoch": 6.398665375094177, |
| "grad_norm": 0.7145769000053406, |
| "learning_rate": 0.0002164680530115289, |
| "loss": 3.2863, |
| "step": 59450 |
| }, |
| { |
| "epoch": 6.404046927133785, |
| "grad_norm": 0.7159892916679382, |
| "learning_rate": 0.00021614481198146748, |
| "loss": 3.2711, |
| "step": 59500 |
| }, |
| { |
| "epoch": 6.4094284791733935, |
| "grad_norm": 0.745442271232605, |
| "learning_rate": 0.00021582157095140607, |
| "loss": 3.2881, |
| "step": 59550 |
| }, |
| { |
| "epoch": 6.414810031213002, |
| "grad_norm": 0.7081004977226257, |
| "learning_rate": 0.00021549832992134467, |
| "loss": 3.2733, |
| "step": 59600 |
| }, |
| { |
| "epoch": 6.42019158325261, |
| "grad_norm": 0.6986952424049377, |
| "learning_rate": 0.00021517508889128324, |
| "loss": 3.2899, |
| "step": 59650 |
| }, |
| { |
| "epoch": 6.425573135292218, |
| "grad_norm": 0.7033957242965698, |
| "learning_rate": 0.00021485184786122183, |
| "loss": 3.3002, |
| "step": 59700 |
| }, |
| { |
| "epoch": 6.430954687331827, |
| "grad_norm": 0.7274535298347473, |
| "learning_rate": 0.0002145286068311604, |
| "loss": 3.2904, |
| "step": 59750 |
| }, |
| { |
| "epoch": 6.4363362393714345, |
| "grad_norm": 0.7411662936210632, |
| "learning_rate": 0.00021420536580109902, |
| "loss": 3.281, |
| "step": 59800 |
| }, |
| { |
| "epoch": 6.441717791411043, |
| "grad_norm": 0.774648129940033, |
| "learning_rate": 0.0002138821247710376, |
| "loss": 3.2846, |
| "step": 59850 |
| }, |
| { |
| "epoch": 6.447099343450651, |
| "grad_norm": 0.7479146122932434, |
| "learning_rate": 0.00021355888374097615, |
| "loss": 3.3097, |
| "step": 59900 |
| }, |
| { |
| "epoch": 6.452480895490259, |
| "grad_norm": 0.7600090503692627, |
| "learning_rate": 0.00021323564271091478, |
| "loss": 3.2781, |
| "step": 59950 |
| }, |
| { |
| "epoch": 6.457862447529868, |
| "grad_norm": 0.723557710647583, |
| "learning_rate": 0.00021291240168085334, |
| "loss": 3.2848, |
| "step": 60000 |
| }, |
| { |
| "epoch": 6.457862447529868, |
| "eval_accuracy": 0.3852529346873366, |
| "eval_loss": 3.373725652694702, |
| "eval_runtime": 184.4028, |
| "eval_samples_per_second": 97.672, |
| "eval_steps_per_second": 6.106, |
| "step": 60000 |
| }, |
| { |
| "epoch": 6.4632439995694755, |
| "grad_norm": 0.7417320609092712, |
| "learning_rate": 0.0002125891606507919, |
| "loss": 3.2836, |
| "step": 60050 |
| }, |
| { |
| "epoch": 6.468625551609084, |
| "grad_norm": 0.7455304861068726, |
| "learning_rate": 0.0002122659196207305, |
| "loss": 3.2872, |
| "step": 60100 |
| }, |
| { |
| "epoch": 6.474007103648693, |
| "grad_norm": 0.7251283526420593, |
| "learning_rate": 0.0002119426785906691, |
| "loss": 3.2821, |
| "step": 60150 |
| }, |
| { |
| "epoch": 6.4793886556883, |
| "grad_norm": 0.70667964220047, |
| "learning_rate": 0.00021161943756060767, |
| "loss": 3.2856, |
| "step": 60200 |
| }, |
| { |
| "epoch": 6.484770207727909, |
| "grad_norm": 0.7194522619247437, |
| "learning_rate": 0.00021129619653054626, |
| "loss": 3.297, |
| "step": 60250 |
| }, |
| { |
| "epoch": 6.490151759767517, |
| "grad_norm": 0.7400850057601929, |
| "learning_rate": 0.00021097295550048483, |
| "loss": 3.3113, |
| "step": 60300 |
| }, |
| { |
| "epoch": 6.495533311807125, |
| "grad_norm": 0.7215319871902466, |
| "learning_rate": 0.00021064971447042343, |
| "loss": 3.2953, |
| "step": 60350 |
| }, |
| { |
| "epoch": 6.500914863846734, |
| "grad_norm": 0.7370033860206604, |
| "learning_rate": 0.00021032647344036202, |
| "loss": 3.2912, |
| "step": 60400 |
| }, |
| { |
| "epoch": 6.506296415886341, |
| "grad_norm": 0.7307314872741699, |
| "learning_rate": 0.0002100032324103006, |
| "loss": 3.2959, |
| "step": 60450 |
| }, |
| { |
| "epoch": 6.51167796792595, |
| "grad_norm": 0.7162298560142517, |
| "learning_rate": 0.00020967999138023916, |
| "loss": 3.2938, |
| "step": 60500 |
| }, |
| { |
| "epoch": 6.517059519965558, |
| "grad_norm": 0.7373604774475098, |
| "learning_rate": 0.00020935675035017778, |
| "loss": 3.2815, |
| "step": 60550 |
| }, |
| { |
| "epoch": 6.522441072005166, |
| "grad_norm": 0.7468721866607666, |
| "learning_rate": 0.00020903350932011634, |
| "loss": 3.2863, |
| "step": 60600 |
| }, |
| { |
| "epoch": 6.5278226240447745, |
| "grad_norm": 0.7195272445678711, |
| "learning_rate": 0.0002087102682900549, |
| "loss": 3.2734, |
| "step": 60650 |
| }, |
| { |
| "epoch": 6.533204176084383, |
| "grad_norm": 0.7415298223495483, |
| "learning_rate": 0.00020838702725999353, |
| "loss": 3.2863, |
| "step": 60700 |
| }, |
| { |
| "epoch": 6.538585728123991, |
| "grad_norm": 0.7255674004554749, |
| "learning_rate": 0.0002080637862299321, |
| "loss": 3.2876, |
| "step": 60750 |
| }, |
| { |
| "epoch": 6.543967280163599, |
| "grad_norm": 0.7130120396614075, |
| "learning_rate": 0.0002077405451998707, |
| "loss": 3.2908, |
| "step": 60800 |
| }, |
| { |
| "epoch": 6.549348832203208, |
| "grad_norm": 0.7552233934402466, |
| "learning_rate": 0.00020741730416980926, |
| "loss": 3.2829, |
| "step": 60850 |
| }, |
| { |
| "epoch": 6.5547303842428155, |
| "grad_norm": 0.6897677779197693, |
| "learning_rate": 0.00020709406313974786, |
| "loss": 3.2917, |
| "step": 60900 |
| }, |
| { |
| "epoch": 6.560111936282424, |
| "grad_norm": 0.933058500289917, |
| "learning_rate": 0.00020677082210968645, |
| "loss": 3.289, |
| "step": 60950 |
| }, |
| { |
| "epoch": 6.565493488322032, |
| "grad_norm": 0.7120639681816101, |
| "learning_rate": 0.00020644758107962502, |
| "loss": 3.2926, |
| "step": 61000 |
| }, |
| { |
| "epoch": 6.565493488322032, |
| "eval_accuracy": 0.385314214916064, |
| "eval_loss": 3.3742878437042236, |
| "eval_runtime": 186.1309, |
| "eval_samples_per_second": 96.765, |
| "eval_steps_per_second": 6.05, |
| "step": 61000 |
| }, |
| { |
| "epoch": 6.57087504036164, |
| "grad_norm": 0.7277801036834717, |
| "learning_rate": 0.0002061243400495636, |
| "loss": 3.2919, |
| "step": 61050 |
| }, |
| { |
| "epoch": 6.576256592401249, |
| "grad_norm": 0.8280491828918457, |
| "learning_rate": 0.0002058010990195022, |
| "loss": 3.2829, |
| "step": 61100 |
| }, |
| { |
| "epoch": 6.5816381444408565, |
| "grad_norm": 0.7976410984992981, |
| "learning_rate": 0.00020547785798944078, |
| "loss": 3.2889, |
| "step": 61150 |
| }, |
| { |
| "epoch": 6.587019696480465, |
| "grad_norm": 0.7580078840255737, |
| "learning_rate": 0.00020515461695937934, |
| "loss": 3.2738, |
| "step": 61200 |
| }, |
| { |
| "epoch": 6.592401248520073, |
| "grad_norm": 0.8833340406417847, |
| "learning_rate": 0.00020483137592931797, |
| "loss": 3.2927, |
| "step": 61250 |
| }, |
| { |
| "epoch": 6.597782800559681, |
| "grad_norm": 0.7139707803726196, |
| "learning_rate": 0.00020450813489925653, |
| "loss": 3.2959, |
| "step": 61300 |
| }, |
| { |
| "epoch": 6.60316435259929, |
| "grad_norm": 0.7467029094696045, |
| "learning_rate": 0.0002041848938691951, |
| "loss": 3.2858, |
| "step": 61350 |
| }, |
| { |
| "epoch": 6.608545904638898, |
| "grad_norm": 0.7808436751365662, |
| "learning_rate": 0.0002038616528391337, |
| "loss": 3.2803, |
| "step": 61400 |
| }, |
| { |
| "epoch": 6.613927456678506, |
| "grad_norm": 0.7623021006584167, |
| "learning_rate": 0.00020353841180907226, |
| "loss": 3.3034, |
| "step": 61450 |
| }, |
| { |
| "epoch": 6.619309008718115, |
| "grad_norm": 0.7227434515953064, |
| "learning_rate": 0.00020321517077901086, |
| "loss": 3.3085, |
| "step": 61500 |
| }, |
| { |
| "epoch": 6.624690560757722, |
| "grad_norm": 0.7426319718360901, |
| "learning_rate": 0.00020289192974894945, |
| "loss": 3.2961, |
| "step": 61550 |
| }, |
| { |
| "epoch": 6.630072112797331, |
| "grad_norm": 0.7512562870979309, |
| "learning_rate": 0.00020256868871888802, |
| "loss": 3.2782, |
| "step": 61600 |
| }, |
| { |
| "epoch": 6.635453664836939, |
| "grad_norm": 0.7073656916618347, |
| "learning_rate": 0.00020224544768882664, |
| "loss": 3.28, |
| "step": 61650 |
| }, |
| { |
| "epoch": 6.640835216876547, |
| "grad_norm": 0.7550458908081055, |
| "learning_rate": 0.0002019222066587652, |
| "loss": 3.2992, |
| "step": 61700 |
| }, |
| { |
| "epoch": 6.6462167689161555, |
| "grad_norm": 0.7247327566146851, |
| "learning_rate": 0.00020159896562870378, |
| "loss": 3.2897, |
| "step": 61750 |
| }, |
| { |
| "epoch": 6.651598320955763, |
| "grad_norm": 0.7218628525733948, |
| "learning_rate": 0.00020127572459864237, |
| "loss": 3.2825, |
| "step": 61800 |
| }, |
| { |
| "epoch": 6.656979872995372, |
| "grad_norm": 0.7239559292793274, |
| "learning_rate": 0.00020095248356858097, |
| "loss": 3.2896, |
| "step": 61850 |
| }, |
| { |
| "epoch": 6.66236142503498, |
| "grad_norm": 0.7311209440231323, |
| "learning_rate": 0.00020062924253851953, |
| "loss": 3.2835, |
| "step": 61900 |
| }, |
| { |
| "epoch": 6.667742977074588, |
| "grad_norm": 0.6923702359199524, |
| "learning_rate": 0.00020030600150845813, |
| "loss": 3.2869, |
| "step": 61950 |
| }, |
| { |
| "epoch": 6.6731245291141965, |
| "grad_norm": 0.703928530216217, |
| "learning_rate": 0.0001999827604783967, |
| "loss": 3.3001, |
| "step": 62000 |
| }, |
| { |
| "epoch": 6.6731245291141965, |
| "eval_accuracy": 0.3858626946937515, |
| "eval_loss": 3.3671865463256836, |
| "eval_runtime": 185.6838, |
| "eval_samples_per_second": 96.998, |
| "eval_steps_per_second": 6.064, |
| "step": 62000 |
| }, |
| { |
| "epoch": 6.678506081153805, |
| "grad_norm": 0.7426213026046753, |
| "learning_rate": 0.0001996595194483353, |
| "loss": 3.3055, |
| "step": 62050 |
| }, |
| { |
| "epoch": 6.683887633193413, |
| "grad_norm": 0.7367209196090698, |
| "learning_rate": 0.00019933627841827389, |
| "loss": 3.2921, |
| "step": 62100 |
| }, |
| { |
| "epoch": 6.689269185233021, |
| "grad_norm": 0.7438285946846008, |
| "learning_rate": 0.00019901303738821245, |
| "loss": 3.2854, |
| "step": 62150 |
| }, |
| { |
| "epoch": 6.69465073727263, |
| "grad_norm": 0.750081479549408, |
| "learning_rate": 0.00019868979635815102, |
| "loss": 3.2727, |
| "step": 62200 |
| }, |
| { |
| "epoch": 6.7000322893122375, |
| "grad_norm": 0.7430285215377808, |
| "learning_rate": 0.00019836655532808964, |
| "loss": 3.2853, |
| "step": 62250 |
| }, |
| { |
| "epoch": 6.705413841351846, |
| "grad_norm": 0.7340359687805176, |
| "learning_rate": 0.0001980433142980282, |
| "loss": 3.2936, |
| "step": 62300 |
| }, |
| { |
| "epoch": 6.710795393391454, |
| "grad_norm": 0.7362263798713684, |
| "learning_rate": 0.00019772007326796678, |
| "loss": 3.2875, |
| "step": 62350 |
| }, |
| { |
| "epoch": 6.716176945431062, |
| "grad_norm": 0.8167773485183716, |
| "learning_rate": 0.0001973968322379054, |
| "loss": 3.2949, |
| "step": 62400 |
| }, |
| { |
| "epoch": 6.721558497470671, |
| "grad_norm": 0.7202277779579163, |
| "learning_rate": 0.00019707359120784397, |
| "loss": 3.293, |
| "step": 62450 |
| }, |
| { |
| "epoch": 6.7269400495102785, |
| "grad_norm": 0.6944708228111267, |
| "learning_rate": 0.00019675035017778253, |
| "loss": 3.2817, |
| "step": 62500 |
| }, |
| { |
| "epoch": 6.732321601549887, |
| "grad_norm": 0.715620756149292, |
| "learning_rate": 0.00019642710914772113, |
| "loss": 3.2855, |
| "step": 62550 |
| }, |
| { |
| "epoch": 6.737703153589496, |
| "grad_norm": 0.7600197792053223, |
| "learning_rate": 0.00019610386811765972, |
| "loss": 3.2856, |
| "step": 62600 |
| }, |
| { |
| "epoch": 6.743084705629103, |
| "grad_norm": 0.7726536393165588, |
| "learning_rate": 0.00019578062708759832, |
| "loss": 3.27, |
| "step": 62650 |
| }, |
| { |
| "epoch": 6.748466257668712, |
| "grad_norm": 0.7212598919868469, |
| "learning_rate": 0.00019545738605753689, |
| "loss": 3.2962, |
| "step": 62700 |
| }, |
| { |
| "epoch": 6.75384780970832, |
| "grad_norm": 0.7723383903503418, |
| "learning_rate": 0.00019513414502747545, |
| "loss": 3.3036, |
| "step": 62750 |
| }, |
| { |
| "epoch": 6.759229361747928, |
| "grad_norm": 0.7511371970176697, |
| "learning_rate": 0.00019481090399741408, |
| "loss": 3.2864, |
| "step": 62800 |
| }, |
| { |
| "epoch": 6.7646109137875365, |
| "grad_norm": 0.7517921924591064, |
| "learning_rate": 0.00019448766296735264, |
| "loss": 3.3052, |
| "step": 62850 |
| }, |
| { |
| "epoch": 6.769992465827144, |
| "grad_norm": 0.7510790824890137, |
| "learning_rate": 0.0001941644219372912, |
| "loss": 3.2835, |
| "step": 62900 |
| }, |
| { |
| "epoch": 6.775374017866753, |
| "grad_norm": 0.7443214058876038, |
| "learning_rate": 0.00019384118090722983, |
| "loss": 3.299, |
| "step": 62950 |
| }, |
| { |
| "epoch": 6.780755569906361, |
| "grad_norm": 0.7521921992301941, |
| "learning_rate": 0.0001935179398771684, |
| "loss": 3.285, |
| "step": 63000 |
| }, |
| { |
| "epoch": 6.780755569906361, |
| "eval_accuracy": 0.3861808303492723, |
| "eval_loss": 3.3636913299560547, |
| "eval_runtime": 185.9174, |
| "eval_samples_per_second": 96.876, |
| "eval_steps_per_second": 6.056, |
| "step": 63000 |
| }, |
| { |
| "epoch": 6.786137121945969, |
| "grad_norm": 0.7525462508201599, |
| "learning_rate": 0.00019319469884710697, |
| "loss": 3.288, |
| "step": 63050 |
| }, |
| { |
| "epoch": 6.7915186739855775, |
| "grad_norm": 0.7926684021949768, |
| "learning_rate": 0.00019287145781704556, |
| "loss": 3.2791, |
| "step": 63100 |
| }, |
| { |
| "epoch": 6.796900226025185, |
| "grad_norm": 0.7578175067901611, |
| "learning_rate": 0.00019254821678698416, |
| "loss": 3.2871, |
| "step": 63150 |
| }, |
| { |
| "epoch": 6.802281778064794, |
| "grad_norm": 0.7035473585128784, |
| "learning_rate": 0.00019222497575692272, |
| "loss": 3.2867, |
| "step": 63200 |
| }, |
| { |
| "epoch": 6.807663330104402, |
| "grad_norm": 0.7578956484794617, |
| "learning_rate": 0.00019190819954746253, |
| "loss": 3.2833, |
| "step": 63250 |
| }, |
| { |
| "epoch": 6.813044882144011, |
| "grad_norm": 0.7775467038154602, |
| "learning_rate": 0.0001915849585174011, |
| "loss": 3.3052, |
| "step": 63300 |
| }, |
| { |
| "epoch": 6.8184264341836185, |
| "grad_norm": 0.7119178771972656, |
| "learning_rate": 0.00019126171748733972, |
| "loss": 3.3001, |
| "step": 63350 |
| }, |
| { |
| "epoch": 6.823807986223227, |
| "grad_norm": 0.7036906480789185, |
| "learning_rate": 0.0001909384764572783, |
| "loss": 3.3091, |
| "step": 63400 |
| }, |
| { |
| "epoch": 6.829189538262835, |
| "grad_norm": 0.7361630201339722, |
| "learning_rate": 0.00019061523542721688, |
| "loss": 3.2839, |
| "step": 63450 |
| }, |
| { |
| "epoch": 6.834571090302443, |
| "grad_norm": 0.7481423616409302, |
| "learning_rate": 0.00019029199439715548, |
| "loss": 3.2791, |
| "step": 63500 |
| }, |
| { |
| "epoch": 6.839952642342052, |
| "grad_norm": 0.7738127708435059, |
| "learning_rate": 0.00018996875336709404, |
| "loss": 3.3003, |
| "step": 63550 |
| }, |
| { |
| "epoch": 6.8453341943816595, |
| "grad_norm": 0.7231500744819641, |
| "learning_rate": 0.00018964551233703264, |
| "loss": 3.2839, |
| "step": 63600 |
| }, |
| { |
| "epoch": 6.850715746421268, |
| "grad_norm": 0.803739070892334, |
| "learning_rate": 0.0001893222713069712, |
| "loss": 3.2965, |
| "step": 63650 |
| }, |
| { |
| "epoch": 6.856097298460876, |
| "grad_norm": 0.7869791388511658, |
| "learning_rate": 0.0001889990302769098, |
| "loss": 3.2808, |
| "step": 63700 |
| }, |
| { |
| "epoch": 6.861478850500484, |
| "grad_norm": 0.7621977925300598, |
| "learning_rate": 0.0001886757892468484, |
| "loss": 3.2896, |
| "step": 63750 |
| }, |
| { |
| "epoch": 6.866860402540093, |
| "grad_norm": 0.7234262228012085, |
| "learning_rate": 0.00018835254821678696, |
| "loss": 3.2922, |
| "step": 63800 |
| }, |
| { |
| "epoch": 6.8722419545797, |
| "grad_norm": 0.7651438117027283, |
| "learning_rate": 0.00018802930718672553, |
| "loss": 3.2865, |
| "step": 63850 |
| }, |
| { |
| "epoch": 6.877623506619309, |
| "grad_norm": 0.752111554145813, |
| "learning_rate": 0.00018770606615666415, |
| "loss": 3.2797, |
| "step": 63900 |
| }, |
| { |
| "epoch": 6.8830050586589175, |
| "grad_norm": 0.7573847770690918, |
| "learning_rate": 0.00018738282512660272, |
| "loss": 3.2812, |
| "step": 63950 |
| }, |
| { |
| "epoch": 6.888386610698525, |
| "grad_norm": 0.7495539784431458, |
| "learning_rate": 0.0001870595840965413, |
| "loss": 3.2812, |
| "step": 64000 |
| }, |
| { |
| "epoch": 6.888386610698525, |
| "eval_accuracy": 0.3865049419136229, |
| "eval_loss": 3.359797477722168, |
| "eval_runtime": 185.9009, |
| "eval_samples_per_second": 96.885, |
| "eval_steps_per_second": 6.057, |
| "step": 64000 |
| }, |
| { |
| "epoch": 6.893768162738134, |
| "grad_norm": 0.7728069424629211, |
| "learning_rate": 0.0001867363430664799, |
| "loss": 3.2643, |
| "step": 64050 |
| }, |
| { |
| "epoch": 6.899149714777742, |
| "grad_norm": 0.7521748542785645, |
| "learning_rate": 0.00018641310203641848, |
| "loss": 3.2905, |
| "step": 64100 |
| }, |
| { |
| "epoch": 6.90453126681735, |
| "grad_norm": 0.7777418494224548, |
| "learning_rate": 0.00018608986100635705, |
| "loss": 3.2872, |
| "step": 64150 |
| }, |
| { |
| "epoch": 6.9099128188569585, |
| "grad_norm": 0.7546526789665222, |
| "learning_rate": 0.00018576661997629564, |
| "loss": 3.2953, |
| "step": 64200 |
| }, |
| { |
| "epoch": 6.915294370896566, |
| "grad_norm": 0.7118765711784363, |
| "learning_rate": 0.00018544337894623423, |
| "loss": 3.275, |
| "step": 64250 |
| }, |
| { |
| "epoch": 6.920675922936175, |
| "grad_norm": 0.7455823421478271, |
| "learning_rate": 0.0001851201379161728, |
| "loss": 3.2854, |
| "step": 64300 |
| }, |
| { |
| "epoch": 6.926057474975783, |
| "grad_norm": 0.747428834438324, |
| "learning_rate": 0.0001847968968861114, |
| "loss": 3.3062, |
| "step": 64350 |
| }, |
| { |
| "epoch": 6.931439027015391, |
| "grad_norm": 0.7414302229881287, |
| "learning_rate": 0.00018447365585604996, |
| "loss": 3.2951, |
| "step": 64400 |
| }, |
| { |
| "epoch": 6.9368205790549995, |
| "grad_norm": 0.6974478960037231, |
| "learning_rate": 0.00018415041482598859, |
| "loss": 3.2932, |
| "step": 64450 |
| }, |
| { |
| "epoch": 6.942202131094608, |
| "grad_norm": 0.7721847891807556, |
| "learning_rate": 0.00018382717379592715, |
| "loss": 3.2928, |
| "step": 64500 |
| }, |
| { |
| "epoch": 6.947583683134216, |
| "grad_norm": 0.7501277923583984, |
| "learning_rate": 0.00018350393276586572, |
| "loss": 3.2793, |
| "step": 64550 |
| }, |
| { |
| "epoch": 6.952965235173824, |
| "grad_norm": 0.7129700183868408, |
| "learning_rate": 0.00018318069173580434, |
| "loss": 3.2818, |
| "step": 64600 |
| }, |
| { |
| "epoch": 6.958346787213433, |
| "grad_norm": 0.81559157371521, |
| "learning_rate": 0.0001828574507057429, |
| "loss": 3.288, |
| "step": 64650 |
| }, |
| { |
| "epoch": 6.9637283392530405, |
| "grad_norm": 0.750301718711853, |
| "learning_rate": 0.00018253420967568148, |
| "loss": 3.2954, |
| "step": 64700 |
| }, |
| { |
| "epoch": 6.969109891292649, |
| "grad_norm": 0.7174051403999329, |
| "learning_rate": 0.00018221096864562007, |
| "loss": 3.2705, |
| "step": 64750 |
| }, |
| { |
| "epoch": 6.974491443332257, |
| "grad_norm": 0.7266606688499451, |
| "learning_rate": 0.00018188772761555867, |
| "loss": 3.2906, |
| "step": 64800 |
| }, |
| { |
| "epoch": 6.979872995371865, |
| "grad_norm": 0.7052478790283203, |
| "learning_rate": 0.00018156448658549723, |
| "loss": 3.2925, |
| "step": 64850 |
| }, |
| { |
| "epoch": 6.985254547411474, |
| "grad_norm": 0.8231821656227112, |
| "learning_rate": 0.00018124124555543583, |
| "loss": 3.2771, |
| "step": 64900 |
| }, |
| { |
| "epoch": 6.990636099451081, |
| "grad_norm": 0.8024710416793823, |
| "learning_rate": 0.0001809180045253744, |
| "loss": 3.3017, |
| "step": 64950 |
| }, |
| { |
| "epoch": 6.99601765149069, |
| "grad_norm": 0.8686951398849487, |
| "learning_rate": 0.00018059476349531296, |
| "loss": 3.2788, |
| "step": 65000 |
| }, |
| { |
| "epoch": 6.99601765149069, |
| "eval_accuracy": 0.3870885165740745, |
| "eval_loss": 3.35601806640625, |
| "eval_runtime": 185.3686, |
| "eval_samples_per_second": 97.163, |
| "eval_steps_per_second": 6.074, |
| "step": 65000 |
| }, |
| { |
| "epoch": 7.0013992035302985, |
| "grad_norm": 0.7666776180267334, |
| "learning_rate": 0.00018027152246525159, |
| "loss": 3.2632, |
| "step": 65050 |
| }, |
| { |
| "epoch": 7.006780755569906, |
| "grad_norm": 0.7738795280456543, |
| "learning_rate": 0.00017994828143519015, |
| "loss": 3.2048, |
| "step": 65100 |
| }, |
| { |
| "epoch": 7.012162307609515, |
| "grad_norm": 0.7600313425064087, |
| "learning_rate": 0.00017962504040512872, |
| "loss": 3.2028, |
| "step": 65150 |
| }, |
| { |
| "epoch": 7.017543859649122, |
| "grad_norm": 0.7462911605834961, |
| "learning_rate": 0.00017930179937506734, |
| "loss": 3.1916, |
| "step": 65200 |
| }, |
| { |
| "epoch": 7.022925411688731, |
| "grad_norm": 0.7242733836174011, |
| "learning_rate": 0.0001789785583450059, |
| "loss": 3.1975, |
| "step": 65250 |
| }, |
| { |
| "epoch": 7.0283069637283395, |
| "grad_norm": 0.7007710337638855, |
| "learning_rate": 0.00017865531731494448, |
| "loss": 3.1913, |
| "step": 65300 |
| }, |
| { |
| "epoch": 7.033688515767947, |
| "grad_norm": 0.7678819298744202, |
| "learning_rate": 0.00017833207628488307, |
| "loss": 3.213, |
| "step": 65350 |
| }, |
| { |
| "epoch": 7.039070067807556, |
| "grad_norm": 0.7424229979515076, |
| "learning_rate": 0.00017800883525482167, |
| "loss": 3.2039, |
| "step": 65400 |
| }, |
| { |
| "epoch": 7.044451619847164, |
| "grad_norm": 0.7469813823699951, |
| "learning_rate": 0.00017768559422476026, |
| "loss": 3.1909, |
| "step": 65450 |
| }, |
| { |
| "epoch": 7.049833171886772, |
| "grad_norm": 0.7923039793968201, |
| "learning_rate": 0.00017736235319469883, |
| "loss": 3.205, |
| "step": 65500 |
| }, |
| { |
| "epoch": 7.0552147239263805, |
| "grad_norm": 0.8204380869865417, |
| "learning_rate": 0.0001770391121646374, |
| "loss": 3.2063, |
| "step": 65550 |
| }, |
| { |
| "epoch": 7.060596275965988, |
| "grad_norm": 0.7640592455863953, |
| "learning_rate": 0.00017671587113457602, |
| "loss": 3.1972, |
| "step": 65600 |
| }, |
| { |
| "epoch": 7.065977828005597, |
| "grad_norm": 0.797031581401825, |
| "learning_rate": 0.0001763926301045146, |
| "loss": 3.2047, |
| "step": 65650 |
| }, |
| { |
| "epoch": 7.071359380045205, |
| "grad_norm": 0.7652319669723511, |
| "learning_rate": 0.00017606938907445315, |
| "loss": 3.2061, |
| "step": 65700 |
| }, |
| { |
| "epoch": 7.076740932084813, |
| "grad_norm": 0.7750967144966125, |
| "learning_rate": 0.00017574614804439178, |
| "loss": 3.2077, |
| "step": 65750 |
| }, |
| { |
| "epoch": 7.0821224841244215, |
| "grad_norm": 0.7342427968978882, |
| "learning_rate": 0.00017542290701433034, |
| "loss": 3.2122, |
| "step": 65800 |
| }, |
| { |
| "epoch": 7.08750403616403, |
| "grad_norm": 0.7213613390922546, |
| "learning_rate": 0.0001750996659842689, |
| "loss": 3.2194, |
| "step": 65850 |
| }, |
| { |
| "epoch": 7.092885588203638, |
| "grad_norm": 0.7914056181907654, |
| "learning_rate": 0.0001747764249542075, |
| "loss": 3.2036, |
| "step": 65900 |
| }, |
| { |
| "epoch": 7.098267140243246, |
| "grad_norm": 0.7841677069664001, |
| "learning_rate": 0.0001744531839241461, |
| "loss": 3.2226, |
| "step": 65950 |
| }, |
| { |
| "epoch": 7.103648692282855, |
| "grad_norm": 0.7444742918014526, |
| "learning_rate": 0.00017412994289408467, |
| "loss": 3.2324, |
| "step": 66000 |
| }, |
| { |
| "epoch": 7.103648692282855, |
| "eval_accuracy": 0.38705374764997386, |
| "eval_loss": 3.3600642681121826, |
| "eval_runtime": 185.3852, |
| "eval_samples_per_second": 97.154, |
| "eval_steps_per_second": 6.074, |
| "step": 66000 |
| }, |
| { |
| "epoch": 7.109030244322462, |
| "grad_norm": 0.7269986867904663, |
| "learning_rate": 0.00017380670186402326, |
| "loss": 3.222, |
| "step": 66050 |
| }, |
| { |
| "epoch": 7.114411796362071, |
| "grad_norm": 0.7909741401672363, |
| "learning_rate": 0.00017348346083396183, |
| "loss": 3.2207, |
| "step": 66100 |
| }, |
| { |
| "epoch": 7.119793348401679, |
| "grad_norm": 0.7954514026641846, |
| "learning_rate": 0.00017316021980390042, |
| "loss": 3.2228, |
| "step": 66150 |
| }, |
| { |
| "epoch": 7.125174900441287, |
| "grad_norm": 0.7739808559417725, |
| "learning_rate": 0.00017283697877383902, |
| "loss": 3.2292, |
| "step": 66200 |
| }, |
| { |
| "epoch": 7.130556452480896, |
| "grad_norm": 0.7709057927131653, |
| "learning_rate": 0.0001725137377437776, |
| "loss": 3.2097, |
| "step": 66250 |
| }, |
| { |
| "epoch": 7.135938004520503, |
| "grad_norm": 0.7626157402992249, |
| "learning_rate": 0.00017219049671371615, |
| "loss": 3.2097, |
| "step": 66300 |
| }, |
| { |
| "epoch": 7.141319556560112, |
| "grad_norm": 0.7553810477256775, |
| "learning_rate": 0.00017186725568365478, |
| "loss": 3.2228, |
| "step": 66350 |
| }, |
| { |
| "epoch": 7.1467011085997205, |
| "grad_norm": 0.7737429141998291, |
| "learning_rate": 0.00017154401465359334, |
| "loss": 3.2253, |
| "step": 66400 |
| }, |
| { |
| "epoch": 7.152082660639328, |
| "grad_norm": 0.828437328338623, |
| "learning_rate": 0.00017122077362353194, |
| "loss": 3.2324, |
| "step": 66450 |
| }, |
| { |
| "epoch": 7.157464212678937, |
| "grad_norm": 0.7904108166694641, |
| "learning_rate": 0.00017089753259347053, |
| "loss": 3.236, |
| "step": 66500 |
| }, |
| { |
| "epoch": 7.162845764718545, |
| "grad_norm": 0.8120779395103455, |
| "learning_rate": 0.0001705742915634091, |
| "loss": 3.2139, |
| "step": 66550 |
| }, |
| { |
| "epoch": 7.168227316758153, |
| "grad_norm": 0.7788743376731873, |
| "learning_rate": 0.0001702510505333477, |
| "loss": 3.2075, |
| "step": 66600 |
| }, |
| { |
| "epoch": 7.1736088687977615, |
| "grad_norm": 0.7754313945770264, |
| "learning_rate": 0.00016992780950328626, |
| "loss": 3.2088, |
| "step": 66650 |
| }, |
| { |
| "epoch": 7.178990420837369, |
| "grad_norm": 0.7604042291641235, |
| "learning_rate": 0.00016960456847322486, |
| "loss": 3.2142, |
| "step": 66700 |
| }, |
| { |
| "epoch": 7.184371972876978, |
| "grad_norm": 0.7660232186317444, |
| "learning_rate": 0.00016928132744316345, |
| "loss": 3.2324, |
| "step": 66750 |
| }, |
| { |
| "epoch": 7.189753524916586, |
| "grad_norm": 0.8033111095428467, |
| "learning_rate": 0.00016895808641310202, |
| "loss": 3.2294, |
| "step": 66800 |
| }, |
| { |
| "epoch": 7.195135076956194, |
| "grad_norm": 0.7543028593063354, |
| "learning_rate": 0.0001686348453830406, |
| "loss": 3.2363, |
| "step": 66850 |
| }, |
| { |
| "epoch": 7.2005166289958025, |
| "grad_norm": 0.7488883137702942, |
| "learning_rate": 0.0001683116043529792, |
| "loss": 3.2186, |
| "step": 66900 |
| }, |
| { |
| "epoch": 7.205898181035411, |
| "grad_norm": 0.7356674075126648, |
| "learning_rate": 0.000167994828143519, |
| "loss": 3.2357, |
| "step": 66950 |
| }, |
| { |
| "epoch": 7.211279733075019, |
| "grad_norm": 0.8123546838760376, |
| "learning_rate": 0.00016767158711345758, |
| "loss": 3.2299, |
| "step": 67000 |
| }, |
| { |
| "epoch": 7.211279733075019, |
| "eval_accuracy": 0.3872414998401173, |
| "eval_loss": 3.3597795963287354, |
| "eval_runtime": 185.6336, |
| "eval_samples_per_second": 97.024, |
| "eval_steps_per_second": 6.066, |
| "step": 67000 |
| }, |
| { |
| "epoch": 7.216661285114627, |
| "grad_norm": 0.7803553938865662, |
| "learning_rate": 0.00016734834608339618, |
| "loss": 3.2275, |
| "step": 67050 |
| }, |
| { |
| "epoch": 7.222042837154235, |
| "grad_norm": 0.7761536240577698, |
| "learning_rate": 0.00016702510505333475, |
| "loss": 3.2226, |
| "step": 67100 |
| }, |
| { |
| "epoch": 7.2274243891938434, |
| "grad_norm": 0.7861804962158203, |
| "learning_rate": 0.00016670186402327334, |
| "loss": 3.2236, |
| "step": 67150 |
| }, |
| { |
| "epoch": 7.232805941233452, |
| "grad_norm": 0.7782896161079407, |
| "learning_rate": 0.0001663786229932119, |
| "loss": 3.2238, |
| "step": 67200 |
| }, |
| { |
| "epoch": 7.23818749327306, |
| "grad_norm": 0.788238525390625, |
| "learning_rate": 0.00016605538196315053, |
| "loss": 3.2344, |
| "step": 67250 |
| }, |
| { |
| "epoch": 7.243569045312668, |
| "grad_norm": 0.7603457570075989, |
| "learning_rate": 0.0001657321409330891, |
| "loss": 3.2303, |
| "step": 67300 |
| }, |
| { |
| "epoch": 7.248950597352277, |
| "grad_norm": 0.8009949922561646, |
| "learning_rate": 0.00016540889990302766, |
| "loss": 3.2343, |
| "step": 67350 |
| }, |
| { |
| "epoch": 7.254332149391884, |
| "grad_norm": 0.8290765285491943, |
| "learning_rate": 0.00016508565887296629, |
| "loss": 3.2391, |
| "step": 67400 |
| }, |
| { |
| "epoch": 7.259713701431493, |
| "grad_norm": 0.7603195905685425, |
| "learning_rate": 0.00016476241784290485, |
| "loss": 3.2145, |
| "step": 67450 |
| }, |
| { |
| "epoch": 7.265095253471101, |
| "grad_norm": 0.7419003248214722, |
| "learning_rate": 0.00016443917681284342, |
| "loss": 3.2315, |
| "step": 67500 |
| }, |
| { |
| "epoch": 7.270476805510709, |
| "grad_norm": 0.7349012494087219, |
| "learning_rate": 0.00016411593578278202, |
| "loss": 3.2342, |
| "step": 67550 |
| }, |
| { |
| "epoch": 7.275858357550318, |
| "grad_norm": 0.7769827842712402, |
| "learning_rate": 0.0001637926947527206, |
| "loss": 3.2128, |
| "step": 67600 |
| }, |
| { |
| "epoch": 7.281239909589925, |
| "grad_norm": 0.779949426651001, |
| "learning_rate": 0.00016346945372265918, |
| "loss": 3.224, |
| "step": 67650 |
| }, |
| { |
| "epoch": 7.286621461629534, |
| "grad_norm": 0.7693045139312744, |
| "learning_rate": 0.00016314621269259777, |
| "loss": 3.2385, |
| "step": 67700 |
| }, |
| { |
| "epoch": 7.2920030136691425, |
| "grad_norm": 0.7848260402679443, |
| "learning_rate": 0.00016282297166253634, |
| "loss": 3.2205, |
| "step": 67750 |
| }, |
| { |
| "epoch": 7.29738456570875, |
| "grad_norm": 0.7522407174110413, |
| "learning_rate": 0.00016249973063247494, |
| "loss": 3.2086, |
| "step": 67800 |
| }, |
| { |
| "epoch": 7.302766117748359, |
| "grad_norm": 0.799242377281189, |
| "learning_rate": 0.00016217648960241353, |
| "loss": 3.2311, |
| "step": 67850 |
| }, |
| { |
| "epoch": 7.308147669787967, |
| "grad_norm": 0.7705872654914856, |
| "learning_rate": 0.0001618532485723521, |
| "loss": 3.2347, |
| "step": 67900 |
| }, |
| { |
| "epoch": 7.313529221827575, |
| "grad_norm": 0.7869008183479309, |
| "learning_rate": 0.00016153000754229067, |
| "loss": 3.2402, |
| "step": 67950 |
| }, |
| { |
| "epoch": 7.3189107738671835, |
| "grad_norm": 0.7905930280685425, |
| "learning_rate": 0.0001612067665122293, |
| "loss": 3.2138, |
| "step": 68000 |
| }, |
| { |
| "epoch": 7.3189107738671835, |
| "eval_accuracy": 0.38763482329400567, |
| "eval_loss": 3.3565242290496826, |
| "eval_runtime": 185.2952, |
| "eval_samples_per_second": 97.202, |
| "eval_steps_per_second": 6.077, |
| "step": 68000 |
| }, |
| { |
| "epoch": 7.324292325906791, |
| "grad_norm": 0.7964014410972595, |
| "learning_rate": 0.00016088352548216785, |
| "loss": 3.2405, |
| "step": 68050 |
| }, |
| { |
| "epoch": 7.3296738779464, |
| "grad_norm": 0.779831051826477, |
| "learning_rate": 0.00016056028445210642, |
| "loss": 3.2321, |
| "step": 68100 |
| }, |
| { |
| "epoch": 7.335055429986008, |
| "grad_norm": 0.7957444190979004, |
| "learning_rate": 0.00016023704342204504, |
| "loss": 3.2332, |
| "step": 68150 |
| }, |
| { |
| "epoch": 7.340436982025616, |
| "grad_norm": 0.7503647208213806, |
| "learning_rate": 0.0001599138023919836, |
| "loss": 3.247, |
| "step": 68200 |
| }, |
| { |
| "epoch": 7.3458185340652244, |
| "grad_norm": 0.8236241340637207, |
| "learning_rate": 0.0001595905613619222, |
| "loss": 3.2457, |
| "step": 68250 |
| }, |
| { |
| "epoch": 7.351200086104833, |
| "grad_norm": 0.7641984820365906, |
| "learning_rate": 0.00015926732033186077, |
| "loss": 3.2195, |
| "step": 68300 |
| }, |
| { |
| "epoch": 7.356581638144441, |
| "grad_norm": 0.7976327538490295, |
| "learning_rate": 0.00015894407930179934, |
| "loss": 3.227, |
| "step": 68350 |
| }, |
| { |
| "epoch": 7.361963190184049, |
| "grad_norm": 0.8243280649185181, |
| "learning_rate": 0.00015862083827173796, |
| "loss": 3.2295, |
| "step": 68400 |
| }, |
| { |
| "epoch": 7.367344742223658, |
| "grad_norm": 0.7711389660835266, |
| "learning_rate": 0.00015829759724167653, |
| "loss": 3.2366, |
| "step": 68450 |
| }, |
| { |
| "epoch": 7.372726294263265, |
| "grad_norm": 0.7629539966583252, |
| "learning_rate": 0.0001579743562116151, |
| "loss": 3.2337, |
| "step": 68500 |
| }, |
| { |
| "epoch": 7.378107846302874, |
| "grad_norm": 0.7828649878501892, |
| "learning_rate": 0.00015765111518155372, |
| "loss": 3.2379, |
| "step": 68550 |
| }, |
| { |
| "epoch": 7.383489398342482, |
| "grad_norm": 0.7742972373962402, |
| "learning_rate": 0.0001573278741514923, |
| "loss": 3.2245, |
| "step": 68600 |
| }, |
| { |
| "epoch": 7.38887095038209, |
| "grad_norm": 0.7311100959777832, |
| "learning_rate": 0.00015700463312143085, |
| "loss": 3.2345, |
| "step": 68650 |
| }, |
| { |
| "epoch": 7.394252502421699, |
| "grad_norm": 0.811392605304718, |
| "learning_rate": 0.00015668139209136945, |
| "loss": 3.2358, |
| "step": 68700 |
| }, |
| { |
| "epoch": 7.399634054461306, |
| "grad_norm": 0.7674997448921204, |
| "learning_rate": 0.00015635815106130804, |
| "loss": 3.2327, |
| "step": 68750 |
| }, |
| { |
| "epoch": 7.405015606500915, |
| "grad_norm": 0.7893558740615845, |
| "learning_rate": 0.0001560349100312466, |
| "loss": 3.2267, |
| "step": 68800 |
| }, |
| { |
| "epoch": 7.4103971585405235, |
| "grad_norm": 0.7651101350784302, |
| "learning_rate": 0.0001557116690011852, |
| "loss": 3.2281, |
| "step": 68850 |
| }, |
| { |
| "epoch": 7.415778710580131, |
| "grad_norm": 0.8185087442398071, |
| "learning_rate": 0.00015538842797112377, |
| "loss": 3.2409, |
| "step": 68900 |
| }, |
| { |
| "epoch": 7.42116026261974, |
| "grad_norm": 0.7831454277038574, |
| "learning_rate": 0.00015506518694106237, |
| "loss": 3.2436, |
| "step": 68950 |
| }, |
| { |
| "epoch": 7.426541814659347, |
| "grad_norm": 0.8313812613487244, |
| "learning_rate": 0.00015474194591100096, |
| "loss": 3.2307, |
| "step": 69000 |
| }, |
| { |
| "epoch": 7.426541814659347, |
| "eval_accuracy": 0.38814179766854806, |
| "eval_loss": 3.351215124130249, |
| "eval_runtime": 185.4234, |
| "eval_samples_per_second": 97.134, |
| "eval_steps_per_second": 6.073, |
| "step": 69000 |
| }, |
| { |
| "epoch": 7.431923366698956, |
| "grad_norm": 0.762865424156189, |
| "learning_rate": 0.00015441870488093953, |
| "loss": 3.2376, |
| "step": 69050 |
| }, |
| { |
| "epoch": 7.4373049187385645, |
| "grad_norm": 0.7708355188369751, |
| "learning_rate": 0.0001540954638508781, |
| "loss": 3.2271, |
| "step": 69100 |
| }, |
| { |
| "epoch": 7.442686470778172, |
| "grad_norm": 0.7354069352149963, |
| "learning_rate": 0.00015377222282081672, |
| "loss": 3.2354, |
| "step": 69150 |
| }, |
| { |
| "epoch": 7.448068022817781, |
| "grad_norm": 0.7922617793083191, |
| "learning_rate": 0.0001534489817907553, |
| "loss": 3.2525, |
| "step": 69200 |
| }, |
| { |
| "epoch": 7.453449574857389, |
| "grad_norm": 0.7759140133857727, |
| "learning_rate": 0.00015312574076069388, |
| "loss": 3.229, |
| "step": 69250 |
| }, |
| { |
| "epoch": 7.458831126896997, |
| "grad_norm": 0.7401615977287292, |
| "learning_rate": 0.00015280249973063248, |
| "loss": 3.2231, |
| "step": 69300 |
| }, |
| { |
| "epoch": 7.4642126789366054, |
| "grad_norm": 0.8062000870704651, |
| "learning_rate": 0.00015247925870057104, |
| "loss": 3.2304, |
| "step": 69350 |
| }, |
| { |
| "epoch": 7.469594230976213, |
| "grad_norm": 0.7673658132553101, |
| "learning_rate": 0.00015215601767050964, |
| "loss": 3.2379, |
| "step": 69400 |
| }, |
| { |
| "epoch": 7.474975783015822, |
| "grad_norm": 0.7671270370483398, |
| "learning_rate": 0.0001518327766404482, |
| "loss": 3.2405, |
| "step": 69450 |
| }, |
| { |
| "epoch": 7.48035733505543, |
| "grad_norm": 0.8334165811538696, |
| "learning_rate": 0.0001515095356103868, |
| "loss": 3.2467, |
| "step": 69500 |
| }, |
| { |
| "epoch": 7.485738887095038, |
| "grad_norm": 0.7375505566596985, |
| "learning_rate": 0.0001511862945803254, |
| "loss": 3.2518, |
| "step": 69550 |
| }, |
| { |
| "epoch": 7.491120439134646, |
| "grad_norm": 0.7733789682388306, |
| "learning_rate": 0.00015086305355026396, |
| "loss": 3.2537, |
| "step": 69600 |
| }, |
| { |
| "epoch": 7.496501991174255, |
| "grad_norm": 0.7866163849830627, |
| "learning_rate": 0.00015053981252020253, |
| "loss": 3.2355, |
| "step": 69650 |
| }, |
| { |
| "epoch": 7.501883543213863, |
| "grad_norm": 0.7521885633468628, |
| "learning_rate": 0.00015021657149014115, |
| "loss": 3.2372, |
| "step": 69700 |
| }, |
| { |
| "epoch": 7.507265095253471, |
| "grad_norm": 0.7642475962638855, |
| "learning_rate": 0.00014989333046007972, |
| "loss": 3.234, |
| "step": 69750 |
| }, |
| { |
| "epoch": 7.51264664729308, |
| "grad_norm": 0.8261973261833191, |
| "learning_rate": 0.00014957008943001832, |
| "loss": 3.2418, |
| "step": 69800 |
| }, |
| { |
| "epoch": 7.518028199332687, |
| "grad_norm": 0.7892765998840332, |
| "learning_rate": 0.00014924684839995688, |
| "loss": 3.2449, |
| "step": 69850 |
| }, |
| { |
| "epoch": 7.523409751372296, |
| "grad_norm": 0.876653790473938, |
| "learning_rate": 0.00014892360736989548, |
| "loss": 3.2268, |
| "step": 69900 |
| }, |
| { |
| "epoch": 7.528791303411904, |
| "grad_norm": 0.7362908124923706, |
| "learning_rate": 0.00014860036633983407, |
| "loss": 3.2202, |
| "step": 69950 |
| }, |
| { |
| "epoch": 7.534172855451512, |
| "grad_norm": 0.7930489182472229, |
| "learning_rate": 0.00014827712530977264, |
| "loss": 3.2237, |
| "step": 70000 |
| }, |
| { |
| "epoch": 7.534172855451512, |
| "eval_accuracy": 0.3885306663540361, |
| "eval_loss": 3.347409248352051, |
| "eval_runtime": 184.9947, |
| "eval_samples_per_second": 97.36, |
| "eval_steps_per_second": 6.087, |
| "step": 70000 |
| }, |
| { |
| "epoch": 7.539554407491121, |
| "grad_norm": 0.8036817312240601, |
| "learning_rate": 0.00014795388427971123, |
| "loss": 3.2342, |
| "step": 70050 |
| }, |
| { |
| "epoch": 7.544935959530728, |
| "grad_norm": 0.7615011930465698, |
| "learning_rate": 0.0001476306432496498, |
| "loss": 3.2196, |
| "step": 70100 |
| }, |
| { |
| "epoch": 7.550317511570337, |
| "grad_norm": 0.7539505958557129, |
| "learning_rate": 0.0001473074022195884, |
| "loss": 3.2305, |
| "step": 70150 |
| }, |
| { |
| "epoch": 7.5556990636099455, |
| "grad_norm": 0.8157840967178345, |
| "learning_rate": 0.00014698416118952696, |
| "loss": 3.2403, |
| "step": 70200 |
| }, |
| { |
| "epoch": 7.561080615649553, |
| "grad_norm": 0.7653469443321228, |
| "learning_rate": 0.00014666092015946556, |
| "loss": 3.245, |
| "step": 70250 |
| }, |
| { |
| "epoch": 7.566462167689162, |
| "grad_norm": 0.798906147480011, |
| "learning_rate": 0.00014633767912940415, |
| "loss": 3.2421, |
| "step": 70300 |
| }, |
| { |
| "epoch": 7.57184371972877, |
| "grad_norm": 0.7808455228805542, |
| "learning_rate": 0.00014601443809934272, |
| "loss": 3.2254, |
| "step": 70350 |
| }, |
| { |
| "epoch": 7.577225271768378, |
| "grad_norm": 0.8109095692634583, |
| "learning_rate": 0.00014569119706928132, |
| "loss": 3.2344, |
| "step": 70400 |
| }, |
| { |
| "epoch": 7.5826068238079865, |
| "grad_norm": 0.8093851804733276, |
| "learning_rate": 0.0001453679560392199, |
| "loss": 3.2406, |
| "step": 70450 |
| }, |
| { |
| "epoch": 7.587988375847594, |
| "grad_norm": 0.7726173400878906, |
| "learning_rate": 0.00014504471500915848, |
| "loss": 3.2277, |
| "step": 70500 |
| }, |
| { |
| "epoch": 7.593369927887203, |
| "grad_norm": 0.7869266271591187, |
| "learning_rate": 0.00014472147397909707, |
| "loss": 3.2524, |
| "step": 70550 |
| }, |
| { |
| "epoch": 7.598751479926811, |
| "grad_norm": 0.8031958341598511, |
| "learning_rate": 0.00014440469776963688, |
| "loss": 3.2369, |
| "step": 70600 |
| }, |
| { |
| "epoch": 7.604133031966419, |
| "grad_norm": 0.8056939840316772, |
| "learning_rate": 0.00014408145673957545, |
| "loss": 3.2352, |
| "step": 70650 |
| }, |
| { |
| "epoch": 7.609514584006027, |
| "grad_norm": 0.8427897095680237, |
| "learning_rate": 0.00014375821570951404, |
| "loss": 3.2355, |
| "step": 70700 |
| }, |
| { |
| "epoch": 7.614896136045635, |
| "grad_norm": 0.8212729692459106, |
| "learning_rate": 0.00014343497467945264, |
| "loss": 3.2493, |
| "step": 70750 |
| }, |
| { |
| "epoch": 7.620277688085244, |
| "grad_norm": 0.8241793513298035, |
| "learning_rate": 0.00014311173364939123, |
| "loss": 3.2543, |
| "step": 70800 |
| }, |
| { |
| "epoch": 7.625659240124852, |
| "grad_norm": 0.7650026679039001, |
| "learning_rate": 0.0001427884926193298, |
| "loss": 3.2323, |
| "step": 70850 |
| }, |
| { |
| "epoch": 7.63104079216446, |
| "grad_norm": 0.8155238032341003, |
| "learning_rate": 0.0001424652515892684, |
| "loss": 3.242, |
| "step": 70900 |
| }, |
| { |
| "epoch": 7.636422344204068, |
| "grad_norm": 0.7850176692008972, |
| "learning_rate": 0.000142142010559207, |
| "loss": 3.2581, |
| "step": 70950 |
| }, |
| { |
| "epoch": 7.641803896243677, |
| "grad_norm": 0.785377562046051, |
| "learning_rate": 0.00014181876952914555, |
| "loss": 3.2459, |
| "step": 71000 |
| }, |
| { |
| "epoch": 7.641803896243677, |
| "eval_accuracy": 0.3885870572028118, |
| "eval_loss": 3.3447868824005127, |
| "eval_runtime": 185.3202, |
| "eval_samples_per_second": 97.189, |
| "eval_steps_per_second": 6.076, |
| "step": 71000 |
| }, |
| { |
| "epoch": 7.647185448283285, |
| "grad_norm": 0.7669951319694519, |
| "learning_rate": 0.00014149552849908415, |
| "loss": 3.23, |
| "step": 71050 |
| }, |
| { |
| "epoch": 7.652567000322893, |
| "grad_norm": 0.8189201354980469, |
| "learning_rate": 0.00014117228746902272, |
| "loss": 3.244, |
| "step": 71100 |
| }, |
| { |
| "epoch": 7.657948552362502, |
| "grad_norm": 0.7769984602928162, |
| "learning_rate": 0.0001408490464389613, |
| "loss": 3.2315, |
| "step": 71150 |
| }, |
| { |
| "epoch": 7.663330104402109, |
| "grad_norm": 0.7977750301361084, |
| "learning_rate": 0.00014052580540889988, |
| "loss": 3.2309, |
| "step": 71200 |
| }, |
| { |
| "epoch": 7.668711656441718, |
| "grad_norm": 0.7878773808479309, |
| "learning_rate": 0.00014020256437883847, |
| "loss": 3.2351, |
| "step": 71250 |
| }, |
| { |
| "epoch": 7.674093208481326, |
| "grad_norm": 0.7823585867881775, |
| "learning_rate": 0.00013987932334877707, |
| "loss": 3.2523, |
| "step": 71300 |
| }, |
| { |
| "epoch": 7.679474760520934, |
| "grad_norm": 0.7985871434211731, |
| "learning_rate": 0.00013955608231871564, |
| "loss": 3.2374, |
| "step": 71350 |
| }, |
| { |
| "epoch": 7.684856312560543, |
| "grad_norm": 0.8083027005195618, |
| "learning_rate": 0.00013923284128865423, |
| "loss": 3.2564, |
| "step": 71400 |
| }, |
| { |
| "epoch": 7.69023786460015, |
| "grad_norm": 0.8058404922485352, |
| "learning_rate": 0.00013890960025859283, |
| "loss": 3.2503, |
| "step": 71450 |
| }, |
| { |
| "epoch": 7.695619416639759, |
| "grad_norm": 0.8511492013931274, |
| "learning_rate": 0.0001385863592285314, |
| "loss": 3.2375, |
| "step": 71500 |
| }, |
| { |
| "epoch": 7.7010009686793675, |
| "grad_norm": 0.7996093034744263, |
| "learning_rate": 0.00013826311819847, |
| "loss": 3.2436, |
| "step": 71550 |
| }, |
| { |
| "epoch": 7.706382520718975, |
| "grad_norm": 0.8188735246658325, |
| "learning_rate": 0.00013793987716840858, |
| "loss": 3.2353, |
| "step": 71600 |
| }, |
| { |
| "epoch": 7.711764072758584, |
| "grad_norm": 0.8392016291618347, |
| "learning_rate": 0.00013761663613834715, |
| "loss": 3.2451, |
| "step": 71650 |
| }, |
| { |
| "epoch": 7.717145624798192, |
| "grad_norm": 0.7832477688789368, |
| "learning_rate": 0.00013729339510828572, |
| "loss": 3.252, |
| "step": 71700 |
| }, |
| { |
| "epoch": 7.7225271768378, |
| "grad_norm": 0.8292062282562256, |
| "learning_rate": 0.0001369701540782243, |
| "loss": 3.2424, |
| "step": 71750 |
| }, |
| { |
| "epoch": 7.727908728877408, |
| "grad_norm": 0.786185085773468, |
| "learning_rate": 0.0001366469130481629, |
| "loss": 3.2195, |
| "step": 71800 |
| }, |
| { |
| "epoch": 7.733290280917016, |
| "grad_norm": 0.7632396817207336, |
| "learning_rate": 0.00013632367201810147, |
| "loss": 3.2411, |
| "step": 71850 |
| }, |
| { |
| "epoch": 7.738671832956625, |
| "grad_norm": 0.7751836776733398, |
| "learning_rate": 0.00013600043098804007, |
| "loss": 3.2415, |
| "step": 71900 |
| }, |
| { |
| "epoch": 7.744053384996233, |
| "grad_norm": 0.7656434774398804, |
| "learning_rate": 0.00013567718995797866, |
| "loss": 3.2421, |
| "step": 71950 |
| }, |
| { |
| "epoch": 7.749434937035841, |
| "grad_norm": 0.7917135953903198, |
| "learning_rate": 0.00013535394892791723, |
| "loss": 3.2524, |
| "step": 72000 |
| }, |
| { |
| "epoch": 7.749434937035841, |
| "eval_accuracy": 0.3892775463048728, |
| "eval_loss": 3.3401355743408203, |
| "eval_runtime": 185.1628, |
| "eval_samples_per_second": 97.271, |
| "eval_steps_per_second": 6.081, |
| "step": 72000 |
| }, |
| { |
| "epoch": 7.754816489075449, |
| "grad_norm": 0.801364541053772, |
| "learning_rate": 0.00013503070789785583, |
| "loss": 3.2339, |
| "step": 72050 |
| }, |
| { |
| "epoch": 7.760198041115058, |
| "grad_norm": 0.8408780097961426, |
| "learning_rate": 0.00013470746686779442, |
| "loss": 3.2441, |
| "step": 72100 |
| }, |
| { |
| "epoch": 7.765579593154666, |
| "grad_norm": 0.8276086449623108, |
| "learning_rate": 0.000134384225837733, |
| "loss": 3.2443, |
| "step": 72150 |
| }, |
| { |
| "epoch": 7.770961145194274, |
| "grad_norm": 0.8465146422386169, |
| "learning_rate": 0.00013406098480767158, |
| "loss": 3.2322, |
| "step": 72200 |
| }, |
| { |
| "epoch": 7.776342697233883, |
| "grad_norm": 0.8557305335998535, |
| "learning_rate": 0.00013373774377761015, |
| "loss": 3.2372, |
| "step": 72250 |
| }, |
| { |
| "epoch": 7.78172424927349, |
| "grad_norm": 0.8262193202972412, |
| "learning_rate": 0.00013341450274754874, |
| "loss": 3.2308, |
| "step": 72300 |
| }, |
| { |
| "epoch": 7.787105801313099, |
| "grad_norm": 0.8111556768417358, |
| "learning_rate": 0.0001330912617174873, |
| "loss": 3.238, |
| "step": 72350 |
| }, |
| { |
| "epoch": 7.792487353352707, |
| "grad_norm": 0.7550597190856934, |
| "learning_rate": 0.0001327680206874259, |
| "loss": 3.24, |
| "step": 72400 |
| }, |
| { |
| "epoch": 7.797868905392315, |
| "grad_norm": 0.7710828185081482, |
| "learning_rate": 0.0001324447796573645, |
| "loss": 3.2414, |
| "step": 72450 |
| }, |
| { |
| "epoch": 7.803250457431924, |
| "grad_norm": 0.8323572278022766, |
| "learning_rate": 0.00013212153862730307, |
| "loss": 3.2446, |
| "step": 72500 |
| }, |
| { |
| "epoch": 7.808632009471531, |
| "grad_norm": 0.7747483849525452, |
| "learning_rate": 0.00013179829759724166, |
| "loss": 3.2418, |
| "step": 72550 |
| }, |
| { |
| "epoch": 7.81401356151114, |
| "grad_norm": 0.8585454821586609, |
| "learning_rate": 0.00013147505656718026, |
| "loss": 3.2396, |
| "step": 72600 |
| }, |
| { |
| "epoch": 7.819395113550748, |
| "grad_norm": 0.7940674424171448, |
| "learning_rate": 0.00013115181553711883, |
| "loss": 3.235, |
| "step": 72650 |
| }, |
| { |
| "epoch": 7.824776665590356, |
| "grad_norm": 0.8253437280654907, |
| "learning_rate": 0.00013082857450705742, |
| "loss": 3.2329, |
| "step": 72700 |
| }, |
| { |
| "epoch": 7.830158217629965, |
| "grad_norm": 0.7802859544754028, |
| "learning_rate": 0.00013050533347699602, |
| "loss": 3.2448, |
| "step": 72750 |
| }, |
| { |
| "epoch": 7.835539769669572, |
| "grad_norm": 0.8375236988067627, |
| "learning_rate": 0.00013018209244693458, |
| "loss": 3.2488, |
| "step": 72800 |
| }, |
| { |
| "epoch": 7.840921321709181, |
| "grad_norm": 0.7941522002220154, |
| "learning_rate": 0.00012985885141687318, |
| "loss": 3.2526, |
| "step": 72850 |
| }, |
| { |
| "epoch": 7.846302873748789, |
| "grad_norm": 0.7875718474388123, |
| "learning_rate": 0.00012953561038681175, |
| "loss": 3.2378, |
| "step": 72900 |
| }, |
| { |
| "epoch": 7.851684425788397, |
| "grad_norm": 0.7901372313499451, |
| "learning_rate": 0.00012921236935675034, |
| "loss": 3.2463, |
| "step": 72950 |
| }, |
| { |
| "epoch": 7.857065977828006, |
| "grad_norm": 0.7917222380638123, |
| "learning_rate": 0.0001288891283266889, |
| "loss": 3.2448, |
| "step": 73000 |
| }, |
| { |
| "epoch": 7.857065977828006, |
| "eval_accuracy": 0.3890915325609344, |
| "eval_loss": 3.338547468185425, |
| "eval_runtime": 184.9196, |
| "eval_samples_per_second": 97.399, |
| "eval_steps_per_second": 6.089, |
| "step": 73000 |
| }, |
| { |
| "epoch": 7.862447529867614, |
| "grad_norm": 0.7870434522628784, |
| "learning_rate": 0.0001285658872966275, |
| "loss": 3.2372, |
| "step": 73050 |
| }, |
| { |
| "epoch": 7.867829081907222, |
| "grad_norm": 0.8214866518974304, |
| "learning_rate": 0.0001282426462665661, |
| "loss": 3.2345, |
| "step": 73100 |
| }, |
| { |
| "epoch": 7.87321063394683, |
| "grad_norm": 0.7990794777870178, |
| "learning_rate": 0.00012791940523650466, |
| "loss": 3.2531, |
| "step": 73150 |
| }, |
| { |
| "epoch": 7.878592185986438, |
| "grad_norm": 0.7697991132736206, |
| "learning_rate": 0.00012759616420644326, |
| "loss": 3.235, |
| "step": 73200 |
| }, |
| { |
| "epoch": 7.883973738026047, |
| "grad_norm": 0.8018404245376587, |
| "learning_rate": 0.00012727292317638185, |
| "loss": 3.2478, |
| "step": 73250 |
| }, |
| { |
| "epoch": 7.889355290065655, |
| "grad_norm": 0.7848626971244812, |
| "learning_rate": 0.00012694968214632045, |
| "loss": 3.2291, |
| "step": 73300 |
| }, |
| { |
| "epoch": 7.894736842105263, |
| "grad_norm": 0.8082217574119568, |
| "learning_rate": 0.00012662644111625902, |
| "loss": 3.2291, |
| "step": 73350 |
| }, |
| { |
| "epoch": 7.900118394144871, |
| "grad_norm": 0.8116781115531921, |
| "learning_rate": 0.0001263032000861976, |
| "loss": 3.243, |
| "step": 73400 |
| }, |
| { |
| "epoch": 7.90549994618448, |
| "grad_norm": 0.8129122257232666, |
| "learning_rate": 0.00012597995905613618, |
| "loss": 3.2448, |
| "step": 73450 |
| }, |
| { |
| "epoch": 7.910881498224088, |
| "grad_norm": 0.7559465765953064, |
| "learning_rate": 0.00012565671802607477, |
| "loss": 3.2353, |
| "step": 73500 |
| }, |
| { |
| "epoch": 7.916263050263696, |
| "grad_norm": 0.8014430403709412, |
| "learning_rate": 0.00012533347699601334, |
| "loss": 3.2305, |
| "step": 73550 |
| }, |
| { |
| "epoch": 7.921644602303305, |
| "grad_norm": 0.8275867104530334, |
| "learning_rate": 0.00012501023596595193, |
| "loss": 3.2334, |
| "step": 73600 |
| }, |
| { |
| "epoch": 7.927026154342912, |
| "grad_norm": 0.8044885993003845, |
| "learning_rate": 0.0001246869949358905, |
| "loss": 3.2502, |
| "step": 73650 |
| }, |
| { |
| "epoch": 7.932407706382521, |
| "grad_norm": 0.7925671339035034, |
| "learning_rate": 0.0001243637539058291, |
| "loss": 3.254, |
| "step": 73700 |
| }, |
| { |
| "epoch": 7.937789258422129, |
| "grad_norm": 0.7803763747215271, |
| "learning_rate": 0.0001240405128757677, |
| "loss": 3.2513, |
| "step": 73750 |
| }, |
| { |
| "epoch": 7.943170810461737, |
| "grad_norm": 0.8105255365371704, |
| "learning_rate": 0.00012371727184570629, |
| "loss": 3.2366, |
| "step": 73800 |
| }, |
| { |
| "epoch": 7.948552362501346, |
| "grad_norm": 0.8192880153656006, |
| "learning_rate": 0.00012339403081564485, |
| "loss": 3.249, |
| "step": 73850 |
| }, |
| { |
| "epoch": 7.953933914540953, |
| "grad_norm": 0.8156821727752686, |
| "learning_rate": 0.00012307078978558345, |
| "loss": 3.2375, |
| "step": 73900 |
| }, |
| { |
| "epoch": 7.959315466580562, |
| "grad_norm": 0.759685754776001, |
| "learning_rate": 0.00012274754875552202, |
| "loss": 3.2364, |
| "step": 73950 |
| }, |
| { |
| "epoch": 7.96469701862017, |
| "grad_norm": 0.8244112133979797, |
| "learning_rate": 0.0001224243077254606, |
| "loss": 3.2453, |
| "step": 74000 |
| }, |
| { |
| "epoch": 7.96469701862017, |
| "eval_accuracy": 0.3897680054404674, |
| "eval_loss": 3.334453821182251, |
| "eval_runtime": 185.0218, |
| "eval_samples_per_second": 97.345, |
| "eval_steps_per_second": 6.086, |
| "step": 74000 |
| }, |
| { |
| "epoch": 7.970078570659778, |
| "grad_norm": 0.8168348073959351, |
| "learning_rate": 0.00012210106669539918, |
| "loss": 3.2585, |
| "step": 74050 |
| }, |
| { |
| "epoch": 7.975460122699387, |
| "grad_norm": 0.8332703113555908, |
| "learning_rate": 0.00012177782566533779, |
| "loss": 3.2674, |
| "step": 74100 |
| }, |
| { |
| "epoch": 7.980841674738995, |
| "grad_norm": 0.805260956287384, |
| "learning_rate": 0.00012145458463527635, |
| "loss": 3.2452, |
| "step": 74150 |
| }, |
| { |
| "epoch": 7.986223226778603, |
| "grad_norm": 0.7721812129020691, |
| "learning_rate": 0.00012113134360521495, |
| "loss": 3.2402, |
| "step": 74200 |
| }, |
| { |
| "epoch": 7.991604778818211, |
| "grad_norm": 0.7534312009811401, |
| "learning_rate": 0.00012080810257515353, |
| "loss": 3.2406, |
| "step": 74250 |
| }, |
| { |
| "epoch": 7.996986330857819, |
| "grad_norm": 0.8354355096817017, |
| "learning_rate": 0.00012048486154509212, |
| "loss": 3.2524, |
| "step": 74300 |
| }, |
| { |
| "epoch": 8.002367882897428, |
| "grad_norm": 0.7531010508537292, |
| "learning_rate": 0.00012016162051503069, |
| "loss": 3.2025, |
| "step": 74350 |
| }, |
| { |
| "epoch": 8.007749434937036, |
| "grad_norm": 0.7959233522415161, |
| "learning_rate": 0.00011983837948496929, |
| "loss": 3.1688, |
| "step": 74400 |
| }, |
| { |
| "epoch": 8.013130986976645, |
| "grad_norm": 0.8078753352165222, |
| "learning_rate": 0.00011951513845490787, |
| "loss": 3.1317, |
| "step": 74450 |
| }, |
| { |
| "epoch": 8.018512539016251, |
| "grad_norm": 0.8000760078430176, |
| "learning_rate": 0.00011919189742484645, |
| "loss": 3.1572, |
| "step": 74500 |
| }, |
| { |
| "epoch": 8.02389409105586, |
| "grad_norm": 0.8154542446136475, |
| "learning_rate": 0.00011886865639478503, |
| "loss": 3.1778, |
| "step": 74550 |
| }, |
| { |
| "epoch": 8.029275643095469, |
| "grad_norm": 0.787812352180481, |
| "learning_rate": 0.00011855188018532485, |
| "loss": 3.1613, |
| "step": 74600 |
| }, |
| { |
| "epoch": 8.034657195135077, |
| "grad_norm": 0.7934473156929016, |
| "learning_rate": 0.00011822863915526343, |
| "loss": 3.1818, |
| "step": 74650 |
| }, |
| { |
| "epoch": 8.040038747174686, |
| "grad_norm": 0.7993143796920776, |
| "learning_rate": 0.00011790539812520201, |
| "loss": 3.1479, |
| "step": 74700 |
| }, |
| { |
| "epoch": 8.045420299214294, |
| "grad_norm": 0.8254367709159851, |
| "learning_rate": 0.00011758215709514061, |
| "loss": 3.1685, |
| "step": 74750 |
| }, |
| { |
| "epoch": 8.050801851253901, |
| "grad_norm": 0.858529269695282, |
| "learning_rate": 0.00011725891606507917, |
| "loss": 3.1728, |
| "step": 74800 |
| }, |
| { |
| "epoch": 8.05618340329351, |
| "grad_norm": 1.1026115417480469, |
| "learning_rate": 0.00011693567503501777, |
| "loss": 3.1755, |
| "step": 74850 |
| }, |
| { |
| "epoch": 8.061564955333118, |
| "grad_norm": 0.7949498891830444, |
| "learning_rate": 0.00011661243400495635, |
| "loss": 3.1708, |
| "step": 74900 |
| }, |
| { |
| "epoch": 8.066946507372727, |
| "grad_norm": 0.7937607765197754, |
| "learning_rate": 0.00011628919297489493, |
| "loss": 3.1623, |
| "step": 74950 |
| }, |
| { |
| "epoch": 8.072328059412335, |
| "grad_norm": 0.852537214756012, |
| "learning_rate": 0.00011596595194483351, |
| "loss": 3.1701, |
| "step": 75000 |
| }, |
| { |
| "epoch": 8.072328059412335, |
| "eval_accuracy": 0.3898147261822276, |
| "eval_loss": 3.3382134437561035, |
| "eval_runtime": 184.9993, |
| "eval_samples_per_second": 97.357, |
| "eval_steps_per_second": 6.087, |
| "step": 75000 |
| }, |
| { |
| "epoch": 8.077709611451942, |
| "grad_norm": 0.8193413019180298, |
| "learning_rate": 0.00011564271091477211, |
| "loss": 3.174, |
| "step": 75050 |
| }, |
| { |
| "epoch": 8.08309116349155, |
| "grad_norm": 0.8629480004310608, |
| "learning_rate": 0.0001153194698847107, |
| "loss": 3.173, |
| "step": 75100 |
| }, |
| { |
| "epoch": 8.088472715531159, |
| "grad_norm": 0.8033722043037415, |
| "learning_rate": 0.00011499622885464927, |
| "loss": 3.1637, |
| "step": 75150 |
| }, |
| { |
| "epoch": 8.093854267570768, |
| "grad_norm": 0.8152185082435608, |
| "learning_rate": 0.00011467298782458786, |
| "loss": 3.1754, |
| "step": 75200 |
| }, |
| { |
| "epoch": 8.099235819610376, |
| "grad_norm": 0.8210130929946899, |
| "learning_rate": 0.00011434974679452645, |
| "loss": 3.173, |
| "step": 75250 |
| }, |
| { |
| "epoch": 8.104617371649983, |
| "grad_norm": 0.7966681718826294, |
| "learning_rate": 0.00011402650576446503, |
| "loss": 3.1861, |
| "step": 75300 |
| }, |
| { |
| "epoch": 8.109998923689592, |
| "grad_norm": 0.8028509616851807, |
| "learning_rate": 0.00011370326473440361, |
| "loss": 3.1731, |
| "step": 75350 |
| }, |
| { |
| "epoch": 8.1153804757292, |
| "grad_norm": 0.8054764866828918, |
| "learning_rate": 0.0001133800237043422, |
| "loss": 3.1566, |
| "step": 75400 |
| }, |
| { |
| "epoch": 8.120762027768809, |
| "grad_norm": 0.8137140274047852, |
| "learning_rate": 0.00011305678267428077, |
| "loss": 3.1735, |
| "step": 75450 |
| }, |
| { |
| "epoch": 8.126143579808417, |
| "grad_norm": 0.8169977068901062, |
| "learning_rate": 0.00011273354164421936, |
| "loss": 3.1752, |
| "step": 75500 |
| }, |
| { |
| "epoch": 8.131525131848026, |
| "grad_norm": 0.7922982573509216, |
| "learning_rate": 0.00011241030061415795, |
| "loss": 3.1688, |
| "step": 75550 |
| }, |
| { |
| "epoch": 8.136906683887632, |
| "grad_norm": 0.8188258409500122, |
| "learning_rate": 0.00011208705958409654, |
| "loss": 3.1865, |
| "step": 75600 |
| }, |
| { |
| "epoch": 8.142288235927241, |
| "grad_norm": 0.8402687311172485, |
| "learning_rate": 0.00011176381855403511, |
| "loss": 3.1962, |
| "step": 75650 |
| }, |
| { |
| "epoch": 8.14766978796685, |
| "grad_norm": 0.8304385542869568, |
| "learning_rate": 0.0001114405775239737, |
| "loss": 3.1542, |
| "step": 75700 |
| }, |
| { |
| "epoch": 8.153051340006458, |
| "grad_norm": 0.8148844242095947, |
| "learning_rate": 0.00011111733649391228, |
| "loss": 3.1785, |
| "step": 75750 |
| }, |
| { |
| "epoch": 8.158432892046067, |
| "grad_norm": 0.8117761611938477, |
| "learning_rate": 0.00011079409546385086, |
| "loss": 3.1802, |
| "step": 75800 |
| }, |
| { |
| "epoch": 8.163814444085673, |
| "grad_norm": 0.821329653263092, |
| "learning_rate": 0.00011047731925439068, |
| "loss": 3.1793, |
| "step": 75850 |
| }, |
| { |
| "epoch": 8.169195996125282, |
| "grad_norm": 0.8583831191062927, |
| "learning_rate": 0.00011015407822432928, |
| "loss": 3.1839, |
| "step": 75900 |
| }, |
| { |
| "epoch": 8.17457754816489, |
| "grad_norm": 0.8019603490829468, |
| "learning_rate": 0.00010983083719426785, |
| "loss": 3.1867, |
| "step": 75950 |
| }, |
| { |
| "epoch": 8.1799591002045, |
| "grad_norm": 0.8235881328582764, |
| "learning_rate": 0.00010950759616420644, |
| "loss": 3.1943, |
| "step": 76000 |
| }, |
| { |
| "epoch": 8.1799591002045, |
| "eval_accuracy": 0.38993489627615047, |
| "eval_loss": 3.338373899459839, |
| "eval_runtime": 186.0672, |
| "eval_samples_per_second": 96.798, |
| "eval_steps_per_second": 6.052, |
| "step": 76000 |
| }, |
| { |
| "epoch": 8.185340652244108, |
| "grad_norm": 0.83317631483078, |
| "learning_rate": 0.00010918435513414502, |
| "loss": 3.1712, |
| "step": 76050 |
| }, |
| { |
| "epoch": 8.190722204283716, |
| "grad_norm": 0.758842408657074, |
| "learning_rate": 0.00010886111410408359, |
| "loss": 3.1701, |
| "step": 76100 |
| }, |
| { |
| "epoch": 8.196103756323323, |
| "grad_norm": 0.803379476070404, |
| "learning_rate": 0.00010853787307402218, |
| "loss": 3.1913, |
| "step": 76150 |
| }, |
| { |
| "epoch": 8.201485308362932, |
| "grad_norm": 0.8195414543151855, |
| "learning_rate": 0.00010821463204396078, |
| "loss": 3.1836, |
| "step": 76200 |
| }, |
| { |
| "epoch": 8.20686686040254, |
| "grad_norm": 0.8033959865570068, |
| "learning_rate": 0.00010789139101389935, |
| "loss": 3.1796, |
| "step": 76250 |
| }, |
| { |
| "epoch": 8.212248412442149, |
| "grad_norm": 0.7738995552062988, |
| "learning_rate": 0.00010756814998383794, |
| "loss": 3.1808, |
| "step": 76300 |
| }, |
| { |
| "epoch": 8.217629964481757, |
| "grad_norm": 0.8362164497375488, |
| "learning_rate": 0.00010724490895377652, |
| "loss": 3.1813, |
| "step": 76350 |
| }, |
| { |
| "epoch": 8.223011516521364, |
| "grad_norm": 0.8498926162719727, |
| "learning_rate": 0.00010692166792371512, |
| "loss": 3.1778, |
| "step": 76400 |
| }, |
| { |
| "epoch": 8.228393068560973, |
| "grad_norm": 0.8118336796760559, |
| "learning_rate": 0.00010659842689365368, |
| "loss": 3.202, |
| "step": 76450 |
| }, |
| { |
| "epoch": 8.233774620600581, |
| "grad_norm": 0.8127890229225159, |
| "learning_rate": 0.00010627518586359228, |
| "loss": 3.1597, |
| "step": 76500 |
| }, |
| { |
| "epoch": 8.23915617264019, |
| "grad_norm": 0.8220996260643005, |
| "learning_rate": 0.00010595194483353086, |
| "loss": 3.1678, |
| "step": 76550 |
| }, |
| { |
| "epoch": 8.244537724679798, |
| "grad_norm": 0.7945106029510498, |
| "learning_rate": 0.00010562870380346944, |
| "loss": 3.1966, |
| "step": 76600 |
| }, |
| { |
| "epoch": 8.249919276719407, |
| "grad_norm": 0.8012769818305969, |
| "learning_rate": 0.00010530546277340802, |
| "loss": 3.1733, |
| "step": 76650 |
| }, |
| { |
| "epoch": 8.255300828759013, |
| "grad_norm": 0.8267130851745605, |
| "learning_rate": 0.00010498222174334662, |
| "loss": 3.1683, |
| "step": 76700 |
| }, |
| { |
| "epoch": 8.260682380798622, |
| "grad_norm": 0.8328855037689209, |
| "learning_rate": 0.00010465898071328519, |
| "loss": 3.181, |
| "step": 76750 |
| }, |
| { |
| "epoch": 8.26606393283823, |
| "grad_norm": 0.7733327746391296, |
| "learning_rate": 0.00010433573968322378, |
| "loss": 3.1747, |
| "step": 76800 |
| }, |
| { |
| "epoch": 8.27144548487784, |
| "grad_norm": 0.8119755387306213, |
| "learning_rate": 0.00010401249865316237, |
| "loss": 3.2023, |
| "step": 76850 |
| }, |
| { |
| "epoch": 8.276827036917448, |
| "grad_norm": 0.8023077845573425, |
| "learning_rate": 0.00010368925762310096, |
| "loss": 3.1885, |
| "step": 76900 |
| }, |
| { |
| "epoch": 8.282208588957054, |
| "grad_norm": 0.7848573327064514, |
| "learning_rate": 0.00010336601659303954, |
| "loss": 3.2007, |
| "step": 76950 |
| }, |
| { |
| "epoch": 8.287590140996663, |
| "grad_norm": 0.7905171513557434, |
| "learning_rate": 0.00010304277556297812, |
| "loss": 3.1906, |
| "step": 77000 |
| }, |
| { |
| "epoch": 8.287590140996663, |
| "eval_accuracy": 0.39015557029130166, |
| "eval_loss": 3.3344521522521973, |
| "eval_runtime": 184.7107, |
| "eval_samples_per_second": 97.509, |
| "eval_steps_per_second": 6.096, |
| "step": 77000 |
| }, |
| { |
| "epoch": 8.292971693036272, |
| "grad_norm": 0.8187376260757446, |
| "learning_rate": 0.00010271953453291671, |
| "loss": 3.1843, |
| "step": 77050 |
| }, |
| { |
| "epoch": 8.29835324507588, |
| "grad_norm": 0.8553321361541748, |
| "learning_rate": 0.00010239629350285528, |
| "loss": 3.1856, |
| "step": 77100 |
| }, |
| { |
| "epoch": 8.303734797115489, |
| "grad_norm": 0.7859914302825928, |
| "learning_rate": 0.00010207305247279387, |
| "loss": 3.1817, |
| "step": 77150 |
| }, |
| { |
| "epoch": 8.309116349155097, |
| "grad_norm": 0.8525058627128601, |
| "learning_rate": 0.00010174981144273246, |
| "loss": 3.1928, |
| "step": 77200 |
| }, |
| { |
| "epoch": 8.314497901194704, |
| "grad_norm": 0.8467106819152832, |
| "learning_rate": 0.00010142657041267104, |
| "loss": 3.1828, |
| "step": 77250 |
| }, |
| { |
| "epoch": 8.319879453234313, |
| "grad_norm": 0.854286789894104, |
| "learning_rate": 0.00010110332938260962, |
| "loss": 3.1902, |
| "step": 77300 |
| }, |
| { |
| "epoch": 8.325261005273921, |
| "grad_norm": 0.8719087243080139, |
| "learning_rate": 0.00010078008835254821, |
| "loss": 3.1861, |
| "step": 77350 |
| }, |
| { |
| "epoch": 8.33064255731353, |
| "grad_norm": 0.8280075788497925, |
| "learning_rate": 0.0001004568473224868, |
| "loss": 3.1662, |
| "step": 77400 |
| }, |
| { |
| "epoch": 8.336024109353138, |
| "grad_norm": 0.8436982035636902, |
| "learning_rate": 0.00010013360629242537, |
| "loss": 3.1986, |
| "step": 77450 |
| }, |
| { |
| "epoch": 8.341405661392745, |
| "grad_norm": 0.8524004817008972, |
| "learning_rate": 9.981036526236396e-05, |
| "loss": 3.1873, |
| "step": 77500 |
| }, |
| { |
| "epoch": 8.346787213432354, |
| "grad_norm": 0.7993379831314087, |
| "learning_rate": 9.948712423230255e-05, |
| "loss": 3.1881, |
| "step": 77550 |
| }, |
| { |
| "epoch": 8.352168765471962, |
| "grad_norm": 0.822976291179657, |
| "learning_rate": 9.916388320224112e-05, |
| "loss": 3.1826, |
| "step": 77600 |
| }, |
| { |
| "epoch": 8.35755031751157, |
| "grad_norm": 0.8318803310394287, |
| "learning_rate": 9.884064217217971e-05, |
| "loss": 3.1908, |
| "step": 77650 |
| }, |
| { |
| "epoch": 8.36293186955118, |
| "grad_norm": 0.8046072125434875, |
| "learning_rate": 9.851740114211831e-05, |
| "loss": 3.1683, |
| "step": 77700 |
| }, |
| { |
| "epoch": 8.368313421590786, |
| "grad_norm": 0.8725241422653198, |
| "learning_rate": 9.819416011205688e-05, |
| "loss": 3.1804, |
| "step": 77750 |
| }, |
| { |
| "epoch": 8.373694973630395, |
| "grad_norm": 0.8241723775863647, |
| "learning_rate": 9.787091908199547e-05, |
| "loss": 3.1809, |
| "step": 77800 |
| }, |
| { |
| "epoch": 8.379076525670003, |
| "grad_norm": 0.8239020109176636, |
| "learning_rate": 9.754767805193405e-05, |
| "loss": 3.18, |
| "step": 77850 |
| }, |
| { |
| "epoch": 8.384458077709612, |
| "grad_norm": 0.8390723466873169, |
| "learning_rate": 9.722443702187265e-05, |
| "loss": 3.1922, |
| "step": 77900 |
| }, |
| { |
| "epoch": 8.38983962974922, |
| "grad_norm": 0.8563929796218872, |
| "learning_rate": 9.690119599181121e-05, |
| "loss": 3.1725, |
| "step": 77950 |
| }, |
| { |
| "epoch": 8.395221181788829, |
| "grad_norm": 0.7796242833137512, |
| "learning_rate": 9.657795496174981e-05, |
| "loss": 3.1759, |
| "step": 78000 |
| }, |
| { |
| "epoch": 8.395221181788829, |
| "eval_accuracy": 0.3906213652213623, |
| "eval_loss": 3.331331729888916, |
| "eval_runtime": 184.5428, |
| "eval_samples_per_second": 97.598, |
| "eval_steps_per_second": 6.102, |
| "step": 78000 |
| }, |
| { |
| "epoch": 8.400602733828435, |
| "grad_norm": 0.8204180598258972, |
| "learning_rate": 9.625471393168839e-05, |
| "loss": 3.1812, |
| "step": 78050 |
| }, |
| { |
| "epoch": 8.405984285868044, |
| "grad_norm": 0.7976886034011841, |
| "learning_rate": 9.593147290162697e-05, |
| "loss": 3.1918, |
| "step": 78100 |
| }, |
| { |
| "epoch": 8.411365837907653, |
| "grad_norm": 0.8261345624923706, |
| "learning_rate": 9.560823187156555e-05, |
| "loss": 3.1952, |
| "step": 78150 |
| }, |
| { |
| "epoch": 8.416747389947261, |
| "grad_norm": 0.8616110682487488, |
| "learning_rate": 9.528499084150415e-05, |
| "loss": 3.1779, |
| "step": 78200 |
| }, |
| { |
| "epoch": 8.42212894198687, |
| "grad_norm": 0.7845104932785034, |
| "learning_rate": 9.496174981144271e-05, |
| "loss": 3.1962, |
| "step": 78250 |
| }, |
| { |
| "epoch": 8.427510494026476, |
| "grad_norm": 0.873723566532135, |
| "learning_rate": 9.463850878138131e-05, |
| "loss": 3.1786, |
| "step": 78300 |
| }, |
| { |
| "epoch": 8.432892046066085, |
| "grad_norm": 0.8098871111869812, |
| "learning_rate": 9.431526775131989e-05, |
| "loss": 3.1858, |
| "step": 78350 |
| }, |
| { |
| "epoch": 8.438273598105694, |
| "grad_norm": 0.8407077789306641, |
| "learning_rate": 9.399202672125848e-05, |
| "loss": 3.1914, |
| "step": 78400 |
| }, |
| { |
| "epoch": 8.443655150145302, |
| "grad_norm": 0.8190973401069641, |
| "learning_rate": 9.366878569119705e-05, |
| "loss": 3.1817, |
| "step": 78450 |
| }, |
| { |
| "epoch": 8.44903670218491, |
| "grad_norm": 0.8429524302482605, |
| "learning_rate": 9.334554466113565e-05, |
| "loss": 3.1981, |
| "step": 78500 |
| }, |
| { |
| "epoch": 8.45441825422452, |
| "grad_norm": 0.7990863919258118, |
| "learning_rate": 9.302230363107424e-05, |
| "loss": 3.1671, |
| "step": 78550 |
| }, |
| { |
| "epoch": 8.459799806264126, |
| "grad_norm": 0.8443186283111572, |
| "learning_rate": 9.269906260101281e-05, |
| "loss": 3.2079, |
| "step": 78600 |
| }, |
| { |
| "epoch": 8.465181358303735, |
| "grad_norm": 0.8222746849060059, |
| "learning_rate": 9.23758215709514e-05, |
| "loss": 3.2124, |
| "step": 78650 |
| }, |
| { |
| "epoch": 8.470562910343343, |
| "grad_norm": 0.8287258744239807, |
| "learning_rate": 9.205258054088998e-05, |
| "loss": 3.1891, |
| "step": 78700 |
| }, |
| { |
| "epoch": 8.475944462382952, |
| "grad_norm": 0.8466224670410156, |
| "learning_rate": 9.172933951082856e-05, |
| "loss": 3.1759, |
| "step": 78750 |
| }, |
| { |
| "epoch": 8.48132601442256, |
| "grad_norm": 0.8561066389083862, |
| "learning_rate": 9.140609848076715e-05, |
| "loss": 3.187, |
| "step": 78800 |
| }, |
| { |
| "epoch": 8.486707566462167, |
| "grad_norm": 0.7976627349853516, |
| "learning_rate": 9.108285745070574e-05, |
| "loss": 3.1839, |
| "step": 78850 |
| }, |
| { |
| "epoch": 8.492089118501776, |
| "grad_norm": 0.8352649807929993, |
| "learning_rate": 9.075961642064432e-05, |
| "loss": 3.1816, |
| "step": 78900 |
| }, |
| { |
| "epoch": 8.497470670541384, |
| "grad_norm": 0.818343460559845, |
| "learning_rate": 9.04363753905829e-05, |
| "loss": 3.1888, |
| "step": 78950 |
| }, |
| { |
| "epoch": 8.502852222580993, |
| "grad_norm": 0.8282904624938965, |
| "learning_rate": 9.011313436052148e-05, |
| "loss": 3.1915, |
| "step": 79000 |
| }, |
| { |
| "epoch": 8.502852222580993, |
| "eval_accuracy": 0.3909855697013165, |
| "eval_loss": 3.3280069828033447, |
| "eval_runtime": 184.8504, |
| "eval_samples_per_second": 97.436, |
| "eval_steps_per_second": 6.091, |
| "step": 79000 |
| }, |
| { |
| "epoch": 8.508233774620601, |
| "grad_norm": 0.779547393321991, |
| "learning_rate": 8.978989333046008e-05, |
| "loss": 3.2, |
| "step": 79050 |
| }, |
| { |
| "epoch": 8.513615326660208, |
| "grad_norm": 0.8538345098495483, |
| "learning_rate": 8.946665230039865e-05, |
| "loss": 3.1964, |
| "step": 79100 |
| }, |
| { |
| "epoch": 8.518996878699816, |
| "grad_norm": 0.7864282727241516, |
| "learning_rate": 8.914341127033724e-05, |
| "loss": 3.1979, |
| "step": 79150 |
| }, |
| { |
| "epoch": 8.524378430739425, |
| "grad_norm": 0.8223557472229004, |
| "learning_rate": 8.882017024027582e-05, |
| "loss": 3.186, |
| "step": 79200 |
| }, |
| { |
| "epoch": 8.529759982779034, |
| "grad_norm": 0.8408324122428894, |
| "learning_rate": 8.84969292102144e-05, |
| "loss": 3.1804, |
| "step": 79250 |
| }, |
| { |
| "epoch": 8.535141534818642, |
| "grad_norm": 0.8488681316375732, |
| "learning_rate": 8.817368818015298e-05, |
| "loss": 3.1583, |
| "step": 79300 |
| }, |
| { |
| "epoch": 8.54052308685825, |
| "grad_norm": 0.9019160270690918, |
| "learning_rate": 8.785044715009158e-05, |
| "loss": 3.183, |
| "step": 79350 |
| }, |
| { |
| "epoch": 8.545904638897857, |
| "grad_norm": 0.8224218487739563, |
| "learning_rate": 8.752720612003017e-05, |
| "loss": 3.1904, |
| "step": 79400 |
| }, |
| { |
| "epoch": 8.551286190937466, |
| "grad_norm": 0.8656252026557922, |
| "learning_rate": 8.720396508996874e-05, |
| "loss": 3.1863, |
| "step": 79450 |
| }, |
| { |
| "epoch": 8.556667742977075, |
| "grad_norm": 0.8077045679092407, |
| "learning_rate": 8.688072405990734e-05, |
| "loss": 3.1836, |
| "step": 79500 |
| }, |
| { |
| "epoch": 8.562049295016683, |
| "grad_norm": 0.8300409913063049, |
| "learning_rate": 8.656394785044713e-05, |
| "loss": 3.1885, |
| "step": 79550 |
| }, |
| { |
| "epoch": 8.567430847056292, |
| "grad_norm": 0.8392258882522583, |
| "learning_rate": 8.624070682038572e-05, |
| "loss": 3.1947, |
| "step": 79600 |
| }, |
| { |
| "epoch": 8.572812399095898, |
| "grad_norm": 0.8185868263244629, |
| "learning_rate": 8.591746579032432e-05, |
| "loss": 3.1776, |
| "step": 79650 |
| }, |
| { |
| "epoch": 8.578193951135507, |
| "grad_norm": 0.8567453622817993, |
| "learning_rate": 8.55942247602629e-05, |
| "loss": 3.1977, |
| "step": 79700 |
| }, |
| { |
| "epoch": 8.583575503175116, |
| "grad_norm": 0.8524957299232483, |
| "learning_rate": 8.527098373020148e-05, |
| "loss": 3.2041, |
| "step": 79750 |
| }, |
| { |
| "epoch": 8.588957055214724, |
| "grad_norm": 0.8346071243286133, |
| "learning_rate": 8.494774270014006e-05, |
| "loss": 3.1833, |
| "step": 79800 |
| }, |
| { |
| "epoch": 8.594338607254333, |
| "grad_norm": 0.8146092295646667, |
| "learning_rate": 8.462450167007866e-05, |
| "loss": 3.2014, |
| "step": 79850 |
| }, |
| { |
| "epoch": 8.599720159293941, |
| "grad_norm": 0.8573397994041443, |
| "learning_rate": 8.430126064001722e-05, |
| "loss": 3.1881, |
| "step": 79900 |
| }, |
| { |
| "epoch": 8.605101711333548, |
| "grad_norm": 0.8767163753509521, |
| "learning_rate": 8.397801960995582e-05, |
| "loss": 3.1895, |
| "step": 79950 |
| }, |
| { |
| "epoch": 8.610483263373157, |
| "grad_norm": 0.8038421273231506, |
| "learning_rate": 8.36547785798944e-05, |
| "loss": 3.1977, |
| "step": 80000 |
| }, |
| { |
| "epoch": 8.610483263373157, |
| "eval_accuracy": 0.3914517992429284, |
| "eval_loss": 3.3241639137268066, |
| "eval_runtime": 184.6396, |
| "eval_samples_per_second": 97.547, |
| "eval_steps_per_second": 6.098, |
| "step": 80000 |
| }, |
| { |
| "epoch": 8.615864815412765, |
| "grad_norm": 0.8378307819366455, |
| "learning_rate": 8.333153754983298e-05, |
| "loss": 3.1845, |
| "step": 80050 |
| }, |
| { |
| "epoch": 8.621246367452374, |
| "grad_norm": 0.846358060836792, |
| "learning_rate": 8.300829651977156e-05, |
| "loss": 3.1763, |
| "step": 80100 |
| }, |
| { |
| "epoch": 8.626627919491982, |
| "grad_norm": 0.9070448875427246, |
| "learning_rate": 8.268505548971016e-05, |
| "loss": 3.2014, |
| "step": 80150 |
| }, |
| { |
| "epoch": 8.632009471531589, |
| "grad_norm": 0.8281485438346863, |
| "learning_rate": 8.236181445964875e-05, |
| "loss": 3.1819, |
| "step": 80200 |
| }, |
| { |
| "epoch": 8.637391023571197, |
| "grad_norm": 0.8226301074028015, |
| "learning_rate": 8.203857342958732e-05, |
| "loss": 3.1888, |
| "step": 80250 |
| }, |
| { |
| "epoch": 8.642772575610806, |
| "grad_norm": 0.817070484161377, |
| "learning_rate": 8.171533239952591e-05, |
| "loss": 3.1826, |
| "step": 80300 |
| }, |
| { |
| "epoch": 8.648154127650415, |
| "grad_norm": 0.8424062132835388, |
| "learning_rate": 8.13920913694645e-05, |
| "loss": 3.1891, |
| "step": 80350 |
| }, |
| { |
| "epoch": 8.653535679690023, |
| "grad_norm": 1.12106192111969, |
| "learning_rate": 8.106885033940308e-05, |
| "loss": 3.1823, |
| "step": 80400 |
| }, |
| { |
| "epoch": 8.658917231729632, |
| "grad_norm": 0.7957353591918945, |
| "learning_rate": 8.074560930934166e-05, |
| "loss": 3.1792, |
| "step": 80450 |
| }, |
| { |
| "epoch": 8.664298783769238, |
| "grad_norm": 0.8199249505996704, |
| "learning_rate": 8.042236827928025e-05, |
| "loss": 3.2027, |
| "step": 80500 |
| }, |
| { |
| "epoch": 8.669680335808847, |
| "grad_norm": 0.8268219232559204, |
| "learning_rate": 8.009912724921882e-05, |
| "loss": 3.1962, |
| "step": 80550 |
| }, |
| { |
| "epoch": 8.675061887848456, |
| "grad_norm": 0.8875090479850769, |
| "learning_rate": 7.977588621915741e-05, |
| "loss": 3.1878, |
| "step": 80600 |
| }, |
| { |
| "epoch": 8.680443439888064, |
| "grad_norm": 0.8505129218101501, |
| "learning_rate": 7.9452645189096e-05, |
| "loss": 3.1794, |
| "step": 80650 |
| }, |
| { |
| "epoch": 8.685824991927673, |
| "grad_norm": 0.8001469969749451, |
| "learning_rate": 7.91358689796358e-05, |
| "loss": 3.1733, |
| "step": 80700 |
| }, |
| { |
| "epoch": 8.69120654396728, |
| "grad_norm": 0.7973533868789673, |
| "learning_rate": 7.88126279495744e-05, |
| "loss": 3.187, |
| "step": 80750 |
| }, |
| { |
| "epoch": 8.696588096006888, |
| "grad_norm": 0.8615847229957581, |
| "learning_rate": 7.848938691951298e-05, |
| "loss": 3.1679, |
| "step": 80800 |
| }, |
| { |
| "epoch": 8.701969648046497, |
| "grad_norm": 0.8752439618110657, |
| "learning_rate": 7.816614588945156e-05, |
| "loss": 3.1839, |
| "step": 80850 |
| }, |
| { |
| "epoch": 8.707351200086105, |
| "grad_norm": 0.846650242805481, |
| "learning_rate": 7.784290485939014e-05, |
| "loss": 3.1935, |
| "step": 80900 |
| }, |
| { |
| "epoch": 8.712732752125714, |
| "grad_norm": 0.831285297870636, |
| "learning_rate": 7.751966382932873e-05, |
| "loss": 3.1917, |
| "step": 80950 |
| }, |
| { |
| "epoch": 8.718114304165322, |
| "grad_norm": 0.8393616080284119, |
| "learning_rate": 7.719642279926731e-05, |
| "loss": 3.1806, |
| "step": 81000 |
| }, |
| { |
| "epoch": 8.718114304165322, |
| "eval_accuracy": 0.39170126627335045, |
| "eval_loss": 3.3225531578063965, |
| "eval_runtime": 184.5825, |
| "eval_samples_per_second": 97.577, |
| "eval_steps_per_second": 6.1, |
| "step": 81000 |
| }, |
| { |
| "epoch": 8.723495856204929, |
| "grad_norm": 0.8248891234397888, |
| "learning_rate": 7.68731817692059e-05, |
| "loss": 3.1781, |
| "step": 81050 |
| }, |
| { |
| "epoch": 8.728877408244538, |
| "grad_norm": 0.8192293643951416, |
| "learning_rate": 7.654994073914448e-05, |
| "loss": 3.1876, |
| "step": 81100 |
| }, |
| { |
| "epoch": 8.734258960284146, |
| "grad_norm": 0.8089962601661682, |
| "learning_rate": 7.622669970908307e-05, |
| "loss": 3.1831, |
| "step": 81150 |
| }, |
| { |
| "epoch": 8.739640512323755, |
| "grad_norm": 0.8610677123069763, |
| "learning_rate": 7.590345867902164e-05, |
| "loss": 3.1931, |
| "step": 81200 |
| }, |
| { |
| "epoch": 8.745022064363363, |
| "grad_norm": 0.8550341725349426, |
| "learning_rate": 7.558021764896023e-05, |
| "loss": 3.1866, |
| "step": 81250 |
| }, |
| { |
| "epoch": 8.75040361640297, |
| "grad_norm": 0.8591922521591187, |
| "learning_rate": 7.525697661889883e-05, |
| "loss": 3.1865, |
| "step": 81300 |
| }, |
| { |
| "epoch": 8.755785168442578, |
| "grad_norm": 0.8985189199447632, |
| "learning_rate": 7.493373558883741e-05, |
| "loss": 3.1987, |
| "step": 81350 |
| }, |
| { |
| "epoch": 8.761166720482187, |
| "grad_norm": 0.8182072639465332, |
| "learning_rate": 7.461049455877599e-05, |
| "loss": 3.1902, |
| "step": 81400 |
| }, |
| { |
| "epoch": 8.766548272521796, |
| "grad_norm": 0.8501378893852234, |
| "learning_rate": 7.428725352871457e-05, |
| "loss": 3.1873, |
| "step": 81450 |
| }, |
| { |
| "epoch": 8.771929824561404, |
| "grad_norm": 0.862420916557312, |
| "learning_rate": 7.396401249865315e-05, |
| "loss": 3.1824, |
| "step": 81500 |
| }, |
| { |
| "epoch": 8.777311376601011, |
| "grad_norm": 0.8630905151367188, |
| "learning_rate": 7.364077146859173e-05, |
| "loss": 3.194, |
| "step": 81550 |
| }, |
| { |
| "epoch": 8.78269292864062, |
| "grad_norm": 0.838539719581604, |
| "learning_rate": 7.331753043853033e-05, |
| "loss": 3.1888, |
| "step": 81600 |
| }, |
| { |
| "epoch": 8.788074480680228, |
| "grad_norm": 0.8231281638145447, |
| "learning_rate": 7.299428940846891e-05, |
| "loss": 3.1906, |
| "step": 81650 |
| }, |
| { |
| "epoch": 8.793456032719837, |
| "grad_norm": 0.8844528198242188, |
| "learning_rate": 7.267104837840749e-05, |
| "loss": 3.1923, |
| "step": 81700 |
| }, |
| { |
| "epoch": 8.798837584759445, |
| "grad_norm": 0.8540237545967102, |
| "learning_rate": 7.234780734834607e-05, |
| "loss": 3.1919, |
| "step": 81750 |
| }, |
| { |
| "epoch": 8.804219136799054, |
| "grad_norm": 0.9054052829742432, |
| "learning_rate": 7.202456631828465e-05, |
| "loss": 3.1701, |
| "step": 81800 |
| }, |
| { |
| "epoch": 8.80960068883866, |
| "grad_norm": 0.8349437117576599, |
| "learning_rate": 7.170132528822325e-05, |
| "loss": 3.1747, |
| "step": 81850 |
| }, |
| { |
| "epoch": 8.814982240878269, |
| "grad_norm": 0.8228332996368408, |
| "learning_rate": 7.137808425816183e-05, |
| "loss": 3.1834, |
| "step": 81900 |
| }, |
| { |
| "epoch": 8.820363792917878, |
| "grad_norm": 0.792899489402771, |
| "learning_rate": 7.105484322810041e-05, |
| "loss": 3.1799, |
| "step": 81950 |
| }, |
| { |
| "epoch": 8.825745344957486, |
| "grad_norm": 0.8734421730041504, |
| "learning_rate": 7.073160219803899e-05, |
| "loss": 3.2022, |
| "step": 82000 |
| }, |
| { |
| "epoch": 8.825745344957486, |
| "eval_accuracy": 0.39199636751665456, |
| "eval_loss": 3.3189969062805176, |
| "eval_runtime": 184.5397, |
| "eval_samples_per_second": 97.6, |
| "eval_steps_per_second": 6.102, |
| "step": 82000 |
| }, |
| { |
| "epoch": 8.831126896997095, |
| "grad_norm": 0.8499659299850464, |
| "learning_rate": 7.040836116797757e-05, |
| "loss": 3.1784, |
| "step": 82050 |
| }, |
| { |
| "epoch": 8.836508449036701, |
| "grad_norm": 0.8386522531509399, |
| "learning_rate": 7.008512013791617e-05, |
| "loss": 3.1999, |
| "step": 82100 |
| }, |
| { |
| "epoch": 8.84189000107631, |
| "grad_norm": 0.7806944847106934, |
| "learning_rate": 6.976187910785475e-05, |
| "loss": 3.187, |
| "step": 82150 |
| }, |
| { |
| "epoch": 8.847271553115919, |
| "grad_norm": 0.8628476858139038, |
| "learning_rate": 6.943863807779334e-05, |
| "loss": 3.186, |
| "step": 82200 |
| }, |
| { |
| "epoch": 8.852653105155527, |
| "grad_norm": 0.8431470394134521, |
| "learning_rate": 6.911539704773192e-05, |
| "loss": 3.1932, |
| "step": 82250 |
| }, |
| { |
| "epoch": 8.858034657195136, |
| "grad_norm": 0.8196133971214294, |
| "learning_rate": 6.87921560176705e-05, |
| "loss": 3.1826, |
| "step": 82300 |
| }, |
| { |
| "epoch": 8.863416209234742, |
| "grad_norm": 0.7967073321342468, |
| "learning_rate": 6.846891498760909e-05, |
| "loss": 3.1747, |
| "step": 82350 |
| }, |
| { |
| "epoch": 8.868797761274351, |
| "grad_norm": 0.7723034620285034, |
| "learning_rate": 6.814567395754767e-05, |
| "loss": 3.1753, |
| "step": 82400 |
| }, |
| { |
| "epoch": 8.87417931331396, |
| "grad_norm": 0.7974772453308105, |
| "learning_rate": 6.782243292748626e-05, |
| "loss": 3.2064, |
| "step": 82450 |
| }, |
| { |
| "epoch": 8.879560865353568, |
| "grad_norm": 0.8252553939819336, |
| "learning_rate": 6.749919189742484e-05, |
| "loss": 3.1969, |
| "step": 82500 |
| }, |
| { |
| "epoch": 8.884942417393177, |
| "grad_norm": 0.8222091197967529, |
| "learning_rate": 6.717595086736342e-05, |
| "loss": 3.1896, |
| "step": 82550 |
| }, |
| { |
| "epoch": 8.890323969432785, |
| "grad_norm": 0.8507568836212158, |
| "learning_rate": 6.6852709837302e-05, |
| "loss": 3.1828, |
| "step": 82600 |
| }, |
| { |
| "epoch": 8.895705521472392, |
| "grad_norm": 0.851717472076416, |
| "learning_rate": 6.652946880724059e-05, |
| "loss": 3.1905, |
| "step": 82650 |
| }, |
| { |
| "epoch": 8.901087073512, |
| "grad_norm": 0.8440418839454651, |
| "learning_rate": 6.620622777717918e-05, |
| "loss": 3.2065, |
| "step": 82700 |
| }, |
| { |
| "epoch": 8.906468625551609, |
| "grad_norm": 0.8326337933540344, |
| "learning_rate": 6.588298674711776e-05, |
| "loss": 3.177, |
| "step": 82750 |
| }, |
| { |
| "epoch": 8.911850177591218, |
| "grad_norm": 0.8331674933433533, |
| "learning_rate": 6.555974571705636e-05, |
| "loss": 3.1813, |
| "step": 82800 |
| }, |
| { |
| "epoch": 8.917231729630826, |
| "grad_norm": 0.8243589401245117, |
| "learning_rate": 6.523650468699494e-05, |
| "loss": 3.2014, |
| "step": 82850 |
| }, |
| { |
| "epoch": 8.922613281670433, |
| "grad_norm": 0.8393963575363159, |
| "learning_rate": 6.491326365693352e-05, |
| "loss": 3.2035, |
| "step": 82900 |
| }, |
| { |
| "epoch": 8.927994833710041, |
| "grad_norm": 0.8549140095710754, |
| "learning_rate": 6.45900226268721e-05, |
| "loss": 3.2009, |
| "step": 82950 |
| }, |
| { |
| "epoch": 8.93337638574965, |
| "grad_norm": 0.8491566181182861, |
| "learning_rate": 6.426678159681068e-05, |
| "loss": 3.1851, |
| "step": 83000 |
| }, |
| { |
| "epoch": 8.93337638574965, |
| "eval_accuracy": 0.3924175061098235, |
| "eval_loss": 3.3162879943847656, |
| "eval_runtime": 184.6451, |
| "eval_samples_per_second": 97.544, |
| "eval_steps_per_second": 6.098, |
| "step": 83000 |
| }, |
| { |
| "epoch": 8.938757937789259, |
| "grad_norm": 0.8390645384788513, |
| "learning_rate": 6.394354056674928e-05, |
| "loss": 3.1912, |
| "step": 83050 |
| }, |
| { |
| "epoch": 8.944139489828867, |
| "grad_norm": 0.83738112449646, |
| "learning_rate": 6.362029953668786e-05, |
| "loss": 3.1936, |
| "step": 83100 |
| }, |
| { |
| "epoch": 8.949521041868476, |
| "grad_norm": 0.8717303276062012, |
| "learning_rate": 6.329705850662644e-05, |
| "loss": 3.1696, |
| "step": 83150 |
| }, |
| { |
| "epoch": 8.954902593908082, |
| "grad_norm": 0.8360523581504822, |
| "learning_rate": 6.297381747656502e-05, |
| "loss": 3.1865, |
| "step": 83200 |
| }, |
| { |
| "epoch": 8.960284145947691, |
| "grad_norm": 0.8245041370391846, |
| "learning_rate": 6.26505764465036e-05, |
| "loss": 3.1871, |
| "step": 83250 |
| }, |
| { |
| "epoch": 8.9656656979873, |
| "grad_norm": 0.8196189999580383, |
| "learning_rate": 6.23273354164422e-05, |
| "loss": 3.2, |
| "step": 83300 |
| }, |
| { |
| "epoch": 8.971047250026908, |
| "grad_norm": 0.819350004196167, |
| "learning_rate": 6.200409438638078e-05, |
| "loss": 3.204, |
| "step": 83350 |
| }, |
| { |
| "epoch": 8.976428802066517, |
| "grad_norm": 0.8296644687652588, |
| "learning_rate": 6.168085335631936e-05, |
| "loss": 3.1848, |
| "step": 83400 |
| }, |
| { |
| "epoch": 8.981810354106123, |
| "grad_norm": 0.848055362701416, |
| "learning_rate": 6.135761232625794e-05, |
| "loss": 3.208, |
| "step": 83450 |
| }, |
| { |
| "epoch": 8.987191906145732, |
| "grad_norm": 0.831109344959259, |
| "learning_rate": 6.1034371296196526e-05, |
| "loss": 3.1622, |
| "step": 83500 |
| }, |
| { |
| "epoch": 8.99257345818534, |
| "grad_norm": 0.8709787726402283, |
| "learning_rate": 6.0711130266135114e-05, |
| "loss": 3.1891, |
| "step": 83550 |
| }, |
| { |
| "epoch": 8.997955010224949, |
| "grad_norm": 0.8544104695320129, |
| "learning_rate": 6.0387889236073695e-05, |
| "loss": 3.2029, |
| "step": 83600 |
| }, |
| { |
| "epoch": 9.003336562264558, |
| "grad_norm": 0.8114558458328247, |
| "learning_rate": 6.0064648206012276e-05, |
| "loss": 3.1443, |
| "step": 83650 |
| }, |
| { |
| "epoch": 9.008718114304166, |
| "grad_norm": 0.8070032596588135, |
| "learning_rate": 5.9741407175950864e-05, |
| "loss": 3.1218, |
| "step": 83700 |
| }, |
| { |
| "epoch": 9.014099666343773, |
| "grad_norm": 0.8356350064277649, |
| "learning_rate": 5.9418166145889445e-05, |
| "loss": 3.137, |
| "step": 83750 |
| }, |
| { |
| "epoch": 9.019481218383381, |
| "grad_norm": 0.850892186164856, |
| "learning_rate": 5.909492511582803e-05, |
| "loss": 3.1288, |
| "step": 83800 |
| }, |
| { |
| "epoch": 9.02486277042299, |
| "grad_norm": 0.8735014200210571, |
| "learning_rate": 5.8771684085766614e-05, |
| "loss": 3.1357, |
| "step": 83850 |
| }, |
| { |
| "epoch": 9.030244322462599, |
| "grad_norm": 0.8315800428390503, |
| "learning_rate": 5.8448443055705195e-05, |
| "loss": 3.1184, |
| "step": 83900 |
| }, |
| { |
| "epoch": 9.035625874502207, |
| "grad_norm": 0.8260695338249207, |
| "learning_rate": 5.812520202564378e-05, |
| "loss": 3.1394, |
| "step": 83950 |
| }, |
| { |
| "epoch": 9.041007426541814, |
| "grad_norm": 0.85528963804245, |
| "learning_rate": 5.7801960995582364e-05, |
| "loss": 3.1297, |
| "step": 84000 |
| }, |
| { |
| "epoch": 9.041007426541814, |
| "eval_accuracy": 0.39232688960138623, |
| "eval_loss": 3.318115472793579, |
| "eval_runtime": 184.4628, |
| "eval_samples_per_second": 97.64, |
| "eval_steps_per_second": 6.104, |
| "step": 84000 |
| }, |
| { |
| "epoch": 9.046388978581422, |
| "grad_norm": 0.8493125438690186, |
| "learning_rate": 5.747871996552096e-05, |
| "loss": 3.134, |
| "step": 84050 |
| }, |
| { |
| "epoch": 9.051770530621031, |
| "grad_norm": 0.8698522448539734, |
| "learning_rate": 5.715547893545954e-05, |
| "loss": 3.1345, |
| "step": 84100 |
| }, |
| { |
| "epoch": 9.05715208266064, |
| "grad_norm": 0.838257372379303, |
| "learning_rate": 5.683223790539812e-05, |
| "loss": 3.1236, |
| "step": 84150 |
| }, |
| { |
| "epoch": 9.062533634700248, |
| "grad_norm": 0.8330464363098145, |
| "learning_rate": 5.650899687533671e-05, |
| "loss": 3.1374, |
| "step": 84200 |
| }, |
| { |
| "epoch": 9.067915186739857, |
| "grad_norm": 0.8760532736778259, |
| "learning_rate": 5.618575584527529e-05, |
| "loss": 3.1379, |
| "step": 84250 |
| }, |
| { |
| "epoch": 9.073296738779463, |
| "grad_norm": 0.8745752573013306, |
| "learning_rate": 5.586251481521388e-05, |
| "loss": 3.1318, |
| "step": 84300 |
| }, |
| { |
| "epoch": 9.078678290819072, |
| "grad_norm": 0.8654460906982422, |
| "learning_rate": 5.553927378515246e-05, |
| "loss": 3.1433, |
| "step": 84350 |
| }, |
| { |
| "epoch": 9.08405984285868, |
| "grad_norm": 0.8354806900024414, |
| "learning_rate": 5.521603275509104e-05, |
| "loss": 3.1573, |
| "step": 84400 |
| }, |
| { |
| "epoch": 9.089441394898289, |
| "grad_norm": 0.826752781867981, |
| "learning_rate": 5.489279172502963e-05, |
| "loss": 3.1278, |
| "step": 84450 |
| }, |
| { |
| "epoch": 9.094822946937898, |
| "grad_norm": 0.8850333094596863, |
| "learning_rate": 5.456955069496821e-05, |
| "loss": 3.1349, |
| "step": 84500 |
| }, |
| { |
| "epoch": 9.100204498977504, |
| "grad_norm": 0.9492641687393188, |
| "learning_rate": 5.42463096649068e-05, |
| "loss": 3.126, |
| "step": 84550 |
| }, |
| { |
| "epoch": 9.105586051017113, |
| "grad_norm": 0.8297050595283508, |
| "learning_rate": 5.392306863484538e-05, |
| "loss": 3.1188, |
| "step": 84600 |
| }, |
| { |
| "epoch": 9.110967603056721, |
| "grad_norm": 0.8881982564926147, |
| "learning_rate": 5.359982760478396e-05, |
| "loss": 3.1177, |
| "step": 84650 |
| }, |
| { |
| "epoch": 9.11634915509633, |
| "grad_norm": 0.8270519971847534, |
| "learning_rate": 5.327658657472255e-05, |
| "loss": 3.1435, |
| "step": 84700 |
| }, |
| { |
| "epoch": 9.121730707135939, |
| "grad_norm": 0.8376482129096985, |
| "learning_rate": 5.295334554466113e-05, |
| "loss": 3.1478, |
| "step": 84750 |
| }, |
| { |
| "epoch": 9.127112259175545, |
| "grad_norm": 0.7974318265914917, |
| "learning_rate": 5.2630104514599716e-05, |
| "loss": 3.1418, |
| "step": 84800 |
| }, |
| { |
| "epoch": 9.132493811215154, |
| "grad_norm": 0.852908194065094, |
| "learning_rate": 5.23068634845383e-05, |
| "loss": 3.1387, |
| "step": 84850 |
| }, |
| { |
| "epoch": 9.137875363254762, |
| "grad_norm": 0.8228384256362915, |
| "learning_rate": 5.198362245447688e-05, |
| "loss": 3.143, |
| "step": 84900 |
| }, |
| { |
| "epoch": 9.143256915294371, |
| "grad_norm": 0.8243878483772278, |
| "learning_rate": 5.166038142441547e-05, |
| "loss": 3.1274, |
| "step": 84950 |
| }, |
| { |
| "epoch": 9.14863846733398, |
| "grad_norm": 0.8277369737625122, |
| "learning_rate": 5.1337140394354054e-05, |
| "loss": 3.1309, |
| "step": 85000 |
| }, |
| { |
| "epoch": 9.14863846733398, |
| "eval_accuracy": 0.3925001909574503, |
| "eval_loss": 3.3175320625305176, |
| "eval_runtime": 184.4687, |
| "eval_samples_per_second": 97.637, |
| "eval_steps_per_second": 6.104, |
| "step": 85000 |
| }, |
| { |
| "epoch": 9.154020019373588, |
| "grad_norm": 0.8655103445053101, |
| "learning_rate": 5.101389936429264e-05, |
| "loss": 3.1261, |
| "step": 85050 |
| }, |
| { |
| "epoch": 9.159401571413195, |
| "grad_norm": 0.8685607314109802, |
| "learning_rate": 5.069065833423122e-05, |
| "loss": 3.1284, |
| "step": 85100 |
| }, |
| { |
| "epoch": 9.164783123452803, |
| "grad_norm": 0.8760268092155457, |
| "learning_rate": 5.0367417304169804e-05, |
| "loss": 3.1325, |
| "step": 85150 |
| }, |
| { |
| "epoch": 9.170164675492412, |
| "grad_norm": 0.8456857204437256, |
| "learning_rate": 5.004417627410839e-05, |
| "loss": 3.1454, |
| "step": 85200 |
| }, |
| { |
| "epoch": 9.17554622753202, |
| "grad_norm": 0.8488378524780273, |
| "learning_rate": 4.972093524404697e-05, |
| "loss": 3.1419, |
| "step": 85250 |
| }, |
| { |
| "epoch": 9.180927779571629, |
| "grad_norm": 0.8161391615867615, |
| "learning_rate": 4.939769421398556e-05, |
| "loss": 3.1268, |
| "step": 85300 |
| }, |
| { |
| "epoch": 9.186309331611236, |
| "grad_norm": 0.831362247467041, |
| "learning_rate": 4.907445318392414e-05, |
| "loss": 3.1351, |
| "step": 85350 |
| }, |
| { |
| "epoch": 9.191690883650844, |
| "grad_norm": 0.8824942111968994, |
| "learning_rate": 4.875121215386272e-05, |
| "loss": 3.1441, |
| "step": 85400 |
| }, |
| { |
| "epoch": 9.197072435690453, |
| "grad_norm": 0.8387179970741272, |
| "learning_rate": 4.842797112380131e-05, |
| "loss": 3.1401, |
| "step": 85450 |
| }, |
| { |
| "epoch": 9.202453987730062, |
| "grad_norm": 0.8176819682121277, |
| "learning_rate": 4.810473009373989e-05, |
| "loss": 3.1501, |
| "step": 85500 |
| }, |
| { |
| "epoch": 9.20783553976967, |
| "grad_norm": 0.8273627161979675, |
| "learning_rate": 4.778148906367848e-05, |
| "loss": 3.1534, |
| "step": 85550 |
| }, |
| { |
| "epoch": 9.213217091809279, |
| "grad_norm": 0.8555848598480225, |
| "learning_rate": 4.745824803361706e-05, |
| "loss": 3.1264, |
| "step": 85600 |
| }, |
| { |
| "epoch": 9.218598643848885, |
| "grad_norm": 0.8379252552986145, |
| "learning_rate": 4.713500700355564e-05, |
| "loss": 3.1297, |
| "step": 85650 |
| }, |
| { |
| "epoch": 9.223980195888494, |
| "grad_norm": 0.8088021278381348, |
| "learning_rate": 4.681176597349424e-05, |
| "loss": 3.1484, |
| "step": 85700 |
| }, |
| { |
| "epoch": 9.229361747928102, |
| "grad_norm": 0.8791375160217285, |
| "learning_rate": 4.648852494343282e-05, |
| "loss": 3.108, |
| "step": 85750 |
| }, |
| { |
| "epoch": 9.234743299967711, |
| "grad_norm": 0.874676525592804, |
| "learning_rate": 4.6165283913371406e-05, |
| "loss": 3.1435, |
| "step": 85800 |
| }, |
| { |
| "epoch": 9.24012485200732, |
| "grad_norm": 0.8440104722976685, |
| "learning_rate": 4.584204288330999e-05, |
| "loss": 3.1426, |
| "step": 85850 |
| }, |
| { |
| "epoch": 9.245506404046926, |
| "grad_norm": 0.8481955528259277, |
| "learning_rate": 4.551880185324857e-05, |
| "loss": 3.1367, |
| "step": 85900 |
| }, |
| { |
| "epoch": 9.250887956086535, |
| "grad_norm": 0.8442366123199463, |
| "learning_rate": 4.5195560823187156e-05, |
| "loss": 3.1375, |
| "step": 85950 |
| }, |
| { |
| "epoch": 9.256269508126143, |
| "grad_norm": 0.8529907464981079, |
| "learning_rate": 4.487231979312574e-05, |
| "loss": 3.1235, |
| "step": 86000 |
| }, |
| { |
| "epoch": 9.256269508126143, |
| "eval_accuracy": 0.3927857307466268, |
| "eval_loss": 3.315546989440918, |
| "eval_runtime": 184.656, |
| "eval_samples_per_second": 97.538, |
| "eval_steps_per_second": 6.098, |
| "step": 86000 |
| }, |
| { |
| "epoch": 9.261651060165752, |
| "grad_norm": 0.8316681981086731, |
| "learning_rate": 4.4549078763064325e-05, |
| "loss": 3.1276, |
| "step": 86050 |
| }, |
| { |
| "epoch": 9.26703261220536, |
| "grad_norm": 0.8408940434455872, |
| "learning_rate": 4.4225837733002906e-05, |
| "loss": 3.1184, |
| "step": 86100 |
| }, |
| { |
| "epoch": 9.272414164244967, |
| "grad_norm": 0.8253659605979919, |
| "learning_rate": 4.390259670294149e-05, |
| "loss": 3.1218, |
| "step": 86150 |
| }, |
| { |
| "epoch": 9.277795716284576, |
| "grad_norm": 0.8289502859115601, |
| "learning_rate": 4.3579355672880075e-05, |
| "loss": 3.1409, |
| "step": 86200 |
| }, |
| { |
| "epoch": 9.283177268324184, |
| "grad_norm": 0.8685707449913025, |
| "learning_rate": 4.3256114642818656e-05, |
| "loss": 3.1158, |
| "step": 86250 |
| }, |
| { |
| "epoch": 9.288558820363793, |
| "grad_norm": 0.8403062224388123, |
| "learning_rate": 4.2932873612757244e-05, |
| "loss": 3.1321, |
| "step": 86300 |
| }, |
| { |
| "epoch": 9.293940372403402, |
| "grad_norm": 0.8375023007392883, |
| "learning_rate": 4.2609632582695825e-05, |
| "loss": 3.1274, |
| "step": 86350 |
| }, |
| { |
| "epoch": 9.29932192444301, |
| "grad_norm": 0.8399428725242615, |
| "learning_rate": 4.2286391552634406e-05, |
| "loss": 3.124, |
| "step": 86400 |
| }, |
| { |
| "epoch": 9.304703476482617, |
| "grad_norm": 0.8669325113296509, |
| "learning_rate": 4.1963150522572994e-05, |
| "loss": 3.1222, |
| "step": 86450 |
| }, |
| { |
| "epoch": 9.310085028522225, |
| "grad_norm": 0.8724686503410339, |
| "learning_rate": 4.1639909492511575e-05, |
| "loss": 3.1573, |
| "step": 86500 |
| }, |
| { |
| "epoch": 9.315466580561834, |
| "grad_norm": 0.8969683647155762, |
| "learning_rate": 4.131666846245017e-05, |
| "loss": 3.1455, |
| "step": 86550 |
| }, |
| { |
| "epoch": 9.320848132601443, |
| "grad_norm": 0.8751991987228394, |
| "learning_rate": 4.099342743238875e-05, |
| "loss": 3.1269, |
| "step": 86600 |
| }, |
| { |
| "epoch": 9.326229684641051, |
| "grad_norm": 0.8700101971626282, |
| "learning_rate": 4.067018640232733e-05, |
| "loss": 3.1314, |
| "step": 86650 |
| }, |
| { |
| "epoch": 9.331611236680658, |
| "grad_norm": 0.8855168223381042, |
| "learning_rate": 4.035341019286714e-05, |
| "loss": 3.1481, |
| "step": 86700 |
| }, |
| { |
| "epoch": 9.336992788720266, |
| "grad_norm": 0.8036063313484192, |
| "learning_rate": 4.003016916280573e-05, |
| "loss": 3.121, |
| "step": 86750 |
| }, |
| { |
| "epoch": 9.342374340759875, |
| "grad_norm": 0.8808650374412537, |
| "learning_rate": 3.9706928132744314e-05, |
| "loss": 3.1303, |
| "step": 86800 |
| }, |
| { |
| "epoch": 9.347755892799483, |
| "grad_norm": 0.8660295605659485, |
| "learning_rate": 3.93836871026829e-05, |
| "loss": 3.1483, |
| "step": 86850 |
| }, |
| { |
| "epoch": 9.353137444839092, |
| "grad_norm": 0.8451541066169739, |
| "learning_rate": 3.906044607262148e-05, |
| "loss": 3.1565, |
| "step": 86900 |
| }, |
| { |
| "epoch": 9.3585189968787, |
| "grad_norm": 0.818123996257782, |
| "learning_rate": 3.8737205042560064e-05, |
| "loss": 3.1348, |
| "step": 86950 |
| }, |
| { |
| "epoch": 9.363900548918307, |
| "grad_norm": 0.8490729928016663, |
| "learning_rate": 3.841396401249865e-05, |
| "loss": 3.1488, |
| "step": 87000 |
| }, |
| { |
| "epoch": 9.363900548918307, |
| "eval_accuracy": 0.39300531823289986, |
| "eval_loss": 3.313007116317749, |
| "eval_runtime": 184.3608, |
| "eval_samples_per_second": 97.694, |
| "eval_steps_per_second": 6.108, |
| "step": 87000 |
| }, |
| { |
| "epoch": 9.369282100957916, |
| "grad_norm": 0.8371396064758301, |
| "learning_rate": 3.809072298243723e-05, |
| "loss": 3.1313, |
| "step": 87050 |
| }, |
| { |
| "epoch": 9.374663652997524, |
| "grad_norm": 0.8060377240180969, |
| "learning_rate": 3.776748195237582e-05, |
| "loss": 3.1325, |
| "step": 87100 |
| }, |
| { |
| "epoch": 9.380045205037133, |
| "grad_norm": 0.8614924550056458, |
| "learning_rate": 3.74442409223144e-05, |
| "loss": 3.1245, |
| "step": 87150 |
| }, |
| { |
| "epoch": 9.385426757076742, |
| "grad_norm": 0.848238468170166, |
| "learning_rate": 3.7127464712854216e-05, |
| "loss": 3.1259, |
| "step": 87200 |
| }, |
| { |
| "epoch": 9.390808309116348, |
| "grad_norm": 0.8356297612190247, |
| "learning_rate": 3.6804223682792803e-05, |
| "loss": 3.1552, |
| "step": 87250 |
| }, |
| { |
| "epoch": 9.396189861155957, |
| "grad_norm": 0.8634541034698486, |
| "learning_rate": 3.6480982652731385e-05, |
| "loss": 3.1351, |
| "step": 87300 |
| }, |
| { |
| "epoch": 9.401571413195565, |
| "grad_norm": 0.9080988168716431, |
| "learning_rate": 3.6157741622669966e-05, |
| "loss": 3.1374, |
| "step": 87350 |
| }, |
| { |
| "epoch": 9.406952965235174, |
| "grad_norm": 0.8959322571754456, |
| "learning_rate": 3.5834500592608554e-05, |
| "loss": 3.1358, |
| "step": 87400 |
| }, |
| { |
| "epoch": 9.412334517274783, |
| "grad_norm": 0.8156682848930359, |
| "learning_rate": 3.5511259562547135e-05, |
| "loss": 3.1316, |
| "step": 87450 |
| }, |
| { |
| "epoch": 9.417716069314391, |
| "grad_norm": 0.8381271362304688, |
| "learning_rate": 3.518801853248572e-05, |
| "loss": 3.14, |
| "step": 87500 |
| }, |
| { |
| "epoch": 9.423097621353998, |
| "grad_norm": 0.8428494334220886, |
| "learning_rate": 3.4864777502424304e-05, |
| "loss": 3.145, |
| "step": 87550 |
| }, |
| { |
| "epoch": 9.428479173393606, |
| "grad_norm": 0.8386251330375671, |
| "learning_rate": 3.4541536472362885e-05, |
| "loss": 3.1372, |
| "step": 87600 |
| }, |
| { |
| "epoch": 9.433860725433215, |
| "grad_norm": 0.8217976093292236, |
| "learning_rate": 3.421829544230147e-05, |
| "loss": 3.1445, |
| "step": 87650 |
| }, |
| { |
| "epoch": 9.439242277472824, |
| "grad_norm": 0.8642744421958923, |
| "learning_rate": 3.389505441224006e-05, |
| "loss": 3.1265, |
| "step": 87700 |
| }, |
| { |
| "epoch": 9.444623829512432, |
| "grad_norm": 0.8612553477287292, |
| "learning_rate": 3.357181338217864e-05, |
| "loss": 3.1228, |
| "step": 87750 |
| }, |
| { |
| "epoch": 9.450005381552039, |
| "grad_norm": 0.8542048335075378, |
| "learning_rate": 3.324857235211722e-05, |
| "loss": 3.1372, |
| "step": 87800 |
| }, |
| { |
| "epoch": 9.455386933591647, |
| "grad_norm": 0.882955014705658, |
| "learning_rate": 3.292533132205581e-05, |
| "loss": 3.1426, |
| "step": 87850 |
| }, |
| { |
| "epoch": 9.460768485631256, |
| "grad_norm": 0.9134492874145508, |
| "learning_rate": 3.260209029199439e-05, |
| "loss": 3.1432, |
| "step": 87900 |
| }, |
| { |
| "epoch": 9.466150037670864, |
| "grad_norm": 0.902438223361969, |
| "learning_rate": 3.227884926193298e-05, |
| "loss": 3.1252, |
| "step": 87950 |
| }, |
| { |
| "epoch": 9.471531589710473, |
| "grad_norm": 0.8667796850204468, |
| "learning_rate": 3.195560823187157e-05, |
| "loss": 3.1455, |
| "step": 88000 |
| }, |
| { |
| "epoch": 9.471531589710473, |
| "eval_accuracy": 0.3932915099394032, |
| "eval_loss": 3.3105454444885254, |
| "eval_runtime": 184.2551, |
| "eval_samples_per_second": 97.75, |
| "eval_steps_per_second": 6.111, |
| "step": 88000 |
| }, |
| { |
| "epoch": 9.476913141750082, |
| "grad_norm": 0.8128182888031006, |
| "learning_rate": 3.163236720181014e-05, |
| "loss": 3.1338, |
| "step": 88050 |
| }, |
| { |
| "epoch": 9.482294693789688, |
| "grad_norm": 0.8941981792449951, |
| "learning_rate": 3.130912617174873e-05, |
| "loss": 3.1472, |
| "step": 88100 |
| }, |
| { |
| "epoch": 9.487676245829297, |
| "grad_norm": 0.808533251285553, |
| "learning_rate": 3.098588514168732e-05, |
| "loss": 3.151, |
| "step": 88150 |
| }, |
| { |
| "epoch": 9.493057797868905, |
| "grad_norm": 0.8551432490348816, |
| "learning_rate": 3.06626441116259e-05, |
| "loss": 3.1357, |
| "step": 88200 |
| }, |
| { |
| "epoch": 9.498439349908514, |
| "grad_norm": 0.848689615726471, |
| "learning_rate": 3.0339403081564487e-05, |
| "loss": 3.1529, |
| "step": 88250 |
| }, |
| { |
| "epoch": 9.503820901948123, |
| "grad_norm": 0.8379674553871155, |
| "learning_rate": 3.0016162051503068e-05, |
| "loss": 3.1468, |
| "step": 88300 |
| }, |
| { |
| "epoch": 9.50920245398773, |
| "grad_norm": 0.8962308168411255, |
| "learning_rate": 2.9692921021441652e-05, |
| "loss": 3.148, |
| "step": 88350 |
| }, |
| { |
| "epoch": 9.514584006027338, |
| "grad_norm": 0.8571812510490417, |
| "learning_rate": 2.9369679991380237e-05, |
| "loss": 3.147, |
| "step": 88400 |
| }, |
| { |
| "epoch": 9.519965558066946, |
| "grad_norm": 0.8555566668510437, |
| "learning_rate": 2.904643896131882e-05, |
| "loss": 3.1502, |
| "step": 88450 |
| }, |
| { |
| "epoch": 9.525347110106555, |
| "grad_norm": 0.8368306756019592, |
| "learning_rate": 2.8723197931257406e-05, |
| "loss": 3.1455, |
| "step": 88500 |
| }, |
| { |
| "epoch": 9.530728662146164, |
| "grad_norm": 0.8342770338058472, |
| "learning_rate": 2.8399956901195987e-05, |
| "loss": 3.1193, |
| "step": 88550 |
| }, |
| { |
| "epoch": 9.536110214185772, |
| "grad_norm": 0.829977810382843, |
| "learning_rate": 2.807671587113457e-05, |
| "loss": 3.1288, |
| "step": 88600 |
| }, |
| { |
| "epoch": 9.541491766225379, |
| "grad_norm": 0.8537443280220032, |
| "learning_rate": 2.775347484107316e-05, |
| "loss": 3.1352, |
| "step": 88650 |
| }, |
| { |
| "epoch": 9.546873318264987, |
| "grad_norm": 0.8419883847236633, |
| "learning_rate": 2.7430233811011744e-05, |
| "loss": 3.1337, |
| "step": 88700 |
| }, |
| { |
| "epoch": 9.552254870304596, |
| "grad_norm": 0.8567074537277222, |
| "learning_rate": 2.7106992780950328e-05, |
| "loss": 3.1389, |
| "step": 88750 |
| }, |
| { |
| "epoch": 9.557636422344205, |
| "grad_norm": 0.8512192964553833, |
| "learning_rate": 2.678375175088891e-05, |
| "loss": 3.1232, |
| "step": 88800 |
| }, |
| { |
| "epoch": 9.563017974383813, |
| "grad_norm": 0.8137180805206299, |
| "learning_rate": 2.6460510720827494e-05, |
| "loss": 3.1321, |
| "step": 88850 |
| }, |
| { |
| "epoch": 9.56839952642342, |
| "grad_norm": 0.8155874609947205, |
| "learning_rate": 2.6137269690766078e-05, |
| "loss": 3.1375, |
| "step": 88900 |
| }, |
| { |
| "epoch": 9.573781078463028, |
| "grad_norm": 0.8336650133132935, |
| "learning_rate": 2.5814028660704663e-05, |
| "loss": 3.1384, |
| "step": 88950 |
| }, |
| { |
| "epoch": 9.579162630502637, |
| "grad_norm": 0.880803644657135, |
| "learning_rate": 2.5490787630643247e-05, |
| "loss": 3.1375, |
| "step": 89000 |
| }, |
| { |
| "epoch": 9.579162630502637, |
| "eval_accuracy": 0.39352620017708245, |
| "eval_loss": 3.3093018531799316, |
| "eval_runtime": 184.3935, |
| "eval_samples_per_second": 97.677, |
| "eval_steps_per_second": 6.107, |
| "step": 89000 |
| }, |
| { |
| "epoch": 9.584544182542245, |
| "grad_norm": 0.8568444848060608, |
| "learning_rate": 2.5167546600581828e-05, |
| "loss": 3.1341, |
| "step": 89050 |
| }, |
| { |
| "epoch": 9.589925734581854, |
| "grad_norm": 0.8514227867126465, |
| "learning_rate": 2.4844305570520416e-05, |
| "loss": 3.127, |
| "step": 89100 |
| }, |
| { |
| "epoch": 9.59530728662146, |
| "grad_norm": 0.8377872109413147, |
| "learning_rate": 2.4521064540459e-05, |
| "loss": 3.1369, |
| "step": 89150 |
| }, |
| { |
| "epoch": 9.60068883866107, |
| "grad_norm": 0.8574662804603577, |
| "learning_rate": 2.4197823510397585e-05, |
| "loss": 3.1345, |
| "step": 89200 |
| }, |
| { |
| "epoch": 9.606070390700678, |
| "grad_norm": 0.8137006759643555, |
| "learning_rate": 2.387458248033617e-05, |
| "loss": 3.1332, |
| "step": 89250 |
| }, |
| { |
| "epoch": 9.611451942740286, |
| "grad_norm": 0.871265709400177, |
| "learning_rate": 2.355134145027475e-05, |
| "loss": 3.1301, |
| "step": 89300 |
| }, |
| { |
| "epoch": 9.616833494779895, |
| "grad_norm": 0.8703457713127136, |
| "learning_rate": 2.3228100420213335e-05, |
| "loss": 3.1333, |
| "step": 89350 |
| }, |
| { |
| "epoch": 9.622215046819504, |
| "grad_norm": 0.8446807861328125, |
| "learning_rate": 2.290485939015192e-05, |
| "loss": 3.1298, |
| "step": 89400 |
| }, |
| { |
| "epoch": 9.62759659885911, |
| "grad_norm": 0.8325860500335693, |
| "learning_rate": 2.2581618360090508e-05, |
| "loss": 3.132, |
| "step": 89450 |
| }, |
| { |
| "epoch": 9.632978150898719, |
| "grad_norm": 0.8514713644981384, |
| "learning_rate": 2.2264842150630318e-05, |
| "loss": 3.136, |
| "step": 89500 |
| }, |
| { |
| "epoch": 9.638359702938327, |
| "grad_norm": 0.876413881778717, |
| "learning_rate": 2.1941601120568905e-05, |
| "loss": 3.1412, |
| "step": 89550 |
| }, |
| { |
| "epoch": 9.643741254977936, |
| "grad_norm": 0.8323884010314941, |
| "learning_rate": 2.1618360090507483e-05, |
| "loss": 3.1538, |
| "step": 89600 |
| }, |
| { |
| "epoch": 9.649122807017545, |
| "grad_norm": 0.8908365964889526, |
| "learning_rate": 2.129511906044607e-05, |
| "loss": 3.1279, |
| "step": 89650 |
| }, |
| { |
| "epoch": 9.654504359057151, |
| "grad_norm": 0.8492385149002075, |
| "learning_rate": 2.0971878030384655e-05, |
| "loss": 3.1357, |
| "step": 89700 |
| }, |
| { |
| "epoch": 9.65988591109676, |
| "grad_norm": 0.8621118664741516, |
| "learning_rate": 2.064863700032324e-05, |
| "loss": 3.1436, |
| "step": 89750 |
| }, |
| { |
| "epoch": 9.665267463136368, |
| "grad_norm": 0.8707040548324585, |
| "learning_rate": 2.0325395970261824e-05, |
| "loss": 3.1463, |
| "step": 89800 |
| }, |
| { |
| "epoch": 9.670649015175977, |
| "grad_norm": 0.8553931713104248, |
| "learning_rate": 2.0002154940200406e-05, |
| "loss": 3.1213, |
| "step": 89850 |
| }, |
| { |
| "epoch": 9.676030567215586, |
| "grad_norm": 0.8200343251228333, |
| "learning_rate": 1.967891391013899e-05, |
| "loss": 3.1342, |
| "step": 89900 |
| }, |
| { |
| "epoch": 9.681412119255192, |
| "grad_norm": 0.860724925994873, |
| "learning_rate": 1.9355672880077575e-05, |
| "loss": 3.1524, |
| "step": 89950 |
| }, |
| { |
| "epoch": 9.6867936712948, |
| "grad_norm": 0.8658027648925781, |
| "learning_rate": 1.9032431850016162e-05, |
| "loss": 3.1281, |
| "step": 90000 |
| }, |
| { |
| "epoch": 9.6867936712948, |
| "eval_accuracy": 0.3939400590267678, |
| "eval_loss": 3.3066227436065674, |
| "eval_runtime": 184.6006, |
| "eval_samples_per_second": 97.567, |
| "eval_steps_per_second": 6.1, |
| "step": 90000 |
| }, |
| { |
| "epoch": 9.69217522333441, |
| "grad_norm": 0.8631778955459595, |
| "learning_rate": 1.8709190819954744e-05, |
| "loss": 3.143, |
| "step": 90050 |
| }, |
| { |
| "epoch": 9.697556775374018, |
| "grad_norm": 0.8524464964866638, |
| "learning_rate": 1.8385949789893328e-05, |
| "loss": 3.1402, |
| "step": 90100 |
| }, |
| { |
| "epoch": 9.702938327413626, |
| "grad_norm": 0.8782649040222168, |
| "learning_rate": 1.8062708759831913e-05, |
| "loss": 3.1413, |
| "step": 90150 |
| }, |
| { |
| "epoch": 9.708319879453235, |
| "grad_norm": 0.8450905084609985, |
| "learning_rate": 1.7739467729770497e-05, |
| "loss": 3.1314, |
| "step": 90200 |
| }, |
| { |
| "epoch": 9.713701431492842, |
| "grad_norm": 0.8762593269348145, |
| "learning_rate": 1.741622669970908e-05, |
| "loss": 3.1401, |
| "step": 90250 |
| }, |
| { |
| "epoch": 9.71908298353245, |
| "grad_norm": 0.8233435750007629, |
| "learning_rate": 1.7092985669647666e-05, |
| "loss": 3.1364, |
| "step": 90300 |
| }, |
| { |
| "epoch": 9.724464535572059, |
| "grad_norm": 0.8614866733551025, |
| "learning_rate": 1.676974463958625e-05, |
| "loss": 3.1457, |
| "step": 90350 |
| }, |
| { |
| "epoch": 9.729846087611667, |
| "grad_norm": 0.8564322590827942, |
| "learning_rate": 1.6446503609524835e-05, |
| "loss": 3.1397, |
| "step": 90400 |
| }, |
| { |
| "epoch": 9.735227639651276, |
| "grad_norm": 0.850202202796936, |
| "learning_rate": 1.612326257946342e-05, |
| "loss": 3.1339, |
| "step": 90450 |
| }, |
| { |
| "epoch": 9.740609191690883, |
| "grad_norm": 0.8573711514472961, |
| "learning_rate": 1.5800021549402004e-05, |
| "loss": 3.1491, |
| "step": 90500 |
| }, |
| { |
| "epoch": 9.745990743730491, |
| "grad_norm": 0.8609964847564697, |
| "learning_rate": 1.5476780519340585e-05, |
| "loss": 3.1383, |
| "step": 90550 |
| }, |
| { |
| "epoch": 9.7513722957701, |
| "grad_norm": 0.8481546640396118, |
| "learning_rate": 1.5153539489279171e-05, |
| "loss": 3.1178, |
| "step": 90600 |
| }, |
| { |
| "epoch": 9.756753847809708, |
| "grad_norm": 0.8250866532325745, |
| "learning_rate": 1.4830298459217756e-05, |
| "loss": 3.1341, |
| "step": 90650 |
| }, |
| { |
| "epoch": 9.762135399849317, |
| "grad_norm": 0.8713231086730957, |
| "learning_rate": 1.4507057429156339e-05, |
| "loss": 3.1359, |
| "step": 90700 |
| }, |
| { |
| "epoch": 9.767516951888926, |
| "grad_norm": 0.8582589030265808, |
| "learning_rate": 1.4183816399094925e-05, |
| "loss": 3.1271, |
| "step": 90750 |
| }, |
| { |
| "epoch": 9.772898503928532, |
| "grad_norm": 0.8594558238983154, |
| "learning_rate": 1.3860575369033508e-05, |
| "loss": 3.141, |
| "step": 90800 |
| }, |
| { |
| "epoch": 9.77828005596814, |
| "grad_norm": 0.832633912563324, |
| "learning_rate": 1.3537334338972092e-05, |
| "loss": 3.1495, |
| "step": 90850 |
| }, |
| { |
| "epoch": 9.78366160800775, |
| "grad_norm": 0.8840243220329285, |
| "learning_rate": 1.3214093308910678e-05, |
| "loss": 3.1235, |
| "step": 90900 |
| }, |
| { |
| "epoch": 9.789043160047358, |
| "grad_norm": 0.8464884161949158, |
| "learning_rate": 1.2890852278849261e-05, |
| "loss": 3.1395, |
| "step": 90950 |
| }, |
| { |
| "epoch": 9.794424712086967, |
| "grad_norm": 0.8295110464096069, |
| "learning_rate": 1.2567611248787846e-05, |
| "loss": 3.14, |
| "step": 91000 |
| }, |
| { |
| "epoch": 9.794424712086967, |
| "eval_accuracy": 0.39398591054542553, |
| "eval_loss": 3.3052074909210205, |
| "eval_runtime": 184.568, |
| "eval_samples_per_second": 97.585, |
| "eval_steps_per_second": 6.101, |
| "step": 91000 |
| }, |
| { |
| "epoch": 9.799806264126573, |
| "grad_norm": 0.8717403411865234, |
| "learning_rate": 1.2244370218726428e-05, |
| "loss": 3.1482, |
| "step": 91050 |
| }, |
| { |
| "epoch": 9.805187816166182, |
| "grad_norm": 0.878490149974823, |
| "learning_rate": 1.1921129188665013e-05, |
| "loss": 3.1342, |
| "step": 91100 |
| }, |
| { |
| "epoch": 9.81056936820579, |
| "grad_norm": 0.8271324634552002, |
| "learning_rate": 1.1597888158603599e-05, |
| "loss": 3.1386, |
| "step": 91150 |
| }, |
| { |
| "epoch": 9.815950920245399, |
| "grad_norm": 0.8197286128997803, |
| "learning_rate": 1.1274647128542182e-05, |
| "loss": 3.1268, |
| "step": 91200 |
| }, |
| { |
| "epoch": 9.821332472285007, |
| "grad_norm": 0.8513312935829163, |
| "learning_rate": 1.0951406098480766e-05, |
| "loss": 3.1306, |
| "step": 91250 |
| }, |
| { |
| "epoch": 9.826714024324616, |
| "grad_norm": 0.8409889340400696, |
| "learning_rate": 1.0628165068419349e-05, |
| "loss": 3.1581, |
| "step": 91300 |
| }, |
| { |
| "epoch": 9.832095576364223, |
| "grad_norm": 0.8312180638313293, |
| "learning_rate": 1.0304924038357935e-05, |
| "loss": 3.132, |
| "step": 91350 |
| }, |
| { |
| "epoch": 9.837477128403831, |
| "grad_norm": 0.8290335536003113, |
| "learning_rate": 9.98168300829652e-06, |
| "loss": 3.1575, |
| "step": 91400 |
| }, |
| { |
| "epoch": 9.84285868044344, |
| "grad_norm": 0.887336015701294, |
| "learning_rate": 9.658441978235103e-06, |
| "loss": 3.1239, |
| "step": 91450 |
| }, |
| { |
| "epoch": 9.848240232483048, |
| "grad_norm": 0.8567714095115662, |
| "learning_rate": 9.335200948173687e-06, |
| "loss": 3.145, |
| "step": 91500 |
| }, |
| { |
| "epoch": 9.853621784522657, |
| "grad_norm": 0.8222310543060303, |
| "learning_rate": 9.011959918112272e-06, |
| "loss": 3.1271, |
| "step": 91550 |
| }, |
| { |
| "epoch": 9.859003336562264, |
| "grad_norm": 0.8363861441612244, |
| "learning_rate": 8.688718888050856e-06, |
| "loss": 3.1345, |
| "step": 91600 |
| }, |
| { |
| "epoch": 9.864384888601872, |
| "grad_norm": 0.8393396735191345, |
| "learning_rate": 8.36547785798944e-06, |
| "loss": 3.1483, |
| "step": 91650 |
| }, |
| { |
| "epoch": 9.869766440641481, |
| "grad_norm": 0.8519042134284973, |
| "learning_rate": 8.042236827928023e-06, |
| "loss": 3.1135, |
| "step": 91700 |
| }, |
| { |
| "epoch": 9.87514799268109, |
| "grad_norm": 0.8198532462120056, |
| "learning_rate": 7.71899579786661e-06, |
| "loss": 3.1369, |
| "step": 91750 |
| }, |
| { |
| "epoch": 9.880529544720698, |
| "grad_norm": 0.8474845290184021, |
| "learning_rate": 7.395754767805193e-06, |
| "loss": 3.1332, |
| "step": 91800 |
| }, |
| { |
| "epoch": 9.885911096760307, |
| "grad_norm": 0.843837559223175, |
| "learning_rate": 7.072513737743777e-06, |
| "loss": 3.1413, |
| "step": 91850 |
| }, |
| { |
| "epoch": 9.891292648799913, |
| "grad_norm": 0.8688272833824158, |
| "learning_rate": 6.749272707682361e-06, |
| "loss": 3.1395, |
| "step": 91900 |
| }, |
| { |
| "epoch": 9.896674200839522, |
| "grad_norm": 0.834823489189148, |
| "learning_rate": 6.426031677620945e-06, |
| "loss": 3.1182, |
| "step": 91950 |
| }, |
| { |
| "epoch": 9.90205575287913, |
| "grad_norm": 0.8303846120834351, |
| "learning_rate": 6.10279064755953e-06, |
| "loss": 3.1462, |
| "step": 92000 |
| }, |
| { |
| "epoch": 9.90205575287913, |
| "eval_accuracy": 0.39428514059846664, |
| "eval_loss": 3.3026177883148193, |
| "eval_runtime": 184.592, |
| "eval_samples_per_second": 97.572, |
| "eval_steps_per_second": 6.1, |
| "step": 92000 |
| }, |
| { |
| "epoch": 9.907437304918739, |
| "grad_norm": 0.8718098402023315, |
| "learning_rate": 5.779549617498114e-06, |
| "loss": 3.1523, |
| "step": 92050 |
| }, |
| { |
| "epoch": 9.912818856958348, |
| "grad_norm": 0.8723344802856445, |
| "learning_rate": 5.4563085874366985e-06, |
| "loss": 3.1466, |
| "step": 92100 |
| }, |
| { |
| "epoch": 9.918200408997954, |
| "grad_norm": 0.8737227320671082, |
| "learning_rate": 5.133067557375282e-06, |
| "loss": 3.152, |
| "step": 92150 |
| }, |
| { |
| "epoch": 9.923581961037563, |
| "grad_norm": 0.8384645581245422, |
| "learning_rate": 4.809826527313866e-06, |
| "loss": 3.1369, |
| "step": 92200 |
| }, |
| { |
| "epoch": 9.928963513077171, |
| "grad_norm": 0.8230083584785461, |
| "learning_rate": 4.48658549725245e-06, |
| "loss": 3.1199, |
| "step": 92250 |
| }, |
| { |
| "epoch": 9.93434506511678, |
| "grad_norm": 0.8617830872535706, |
| "learning_rate": 4.163344467191035e-06, |
| "loss": 3.1382, |
| "step": 92300 |
| }, |
| { |
| "epoch": 9.939726617156388, |
| "grad_norm": 0.8417300581932068, |
| "learning_rate": 3.840103437129619e-06, |
| "loss": 3.1341, |
| "step": 92350 |
| }, |
| { |
| "epoch": 9.945108169195997, |
| "grad_norm": 0.8135554194450378, |
| "learning_rate": 3.5168624070682038e-06, |
| "loss": 3.136, |
| "step": 92400 |
| }, |
| { |
| "epoch": 9.950489721235604, |
| "grad_norm": 0.832378089427948, |
| "learning_rate": 3.193621377006788e-06, |
| "loss": 3.1483, |
| "step": 92450 |
| }, |
| { |
| "epoch": 9.955871273275212, |
| "grad_norm": 0.8610823750495911, |
| "learning_rate": 2.870380346945372e-06, |
| "loss": 3.1366, |
| "step": 92500 |
| }, |
| { |
| "epoch": 9.961252825314821, |
| "grad_norm": 0.84225994348526, |
| "learning_rate": 2.5471393168839564e-06, |
| "loss": 3.1237, |
| "step": 92550 |
| }, |
| { |
| "epoch": 9.96663437735443, |
| "grad_norm": 0.8493366241455078, |
| "learning_rate": 2.2238982868225405e-06, |
| "loss": 3.1305, |
| "step": 92600 |
| }, |
| { |
| "epoch": 9.972015929394038, |
| "grad_norm": 0.8750933408737183, |
| "learning_rate": 1.9006572567611246e-06, |
| "loss": 3.1303, |
| "step": 92650 |
| }, |
| { |
| "epoch": 9.977397481433645, |
| "grad_norm": 0.8312073349952698, |
| "learning_rate": 1.5774162266997088e-06, |
| "loss": 3.1224, |
| "step": 92700 |
| }, |
| { |
| "epoch": 9.982779033473253, |
| "grad_norm": 0.8882536888122559, |
| "learning_rate": 1.2541751966382931e-06, |
| "loss": 3.1166, |
| "step": 92750 |
| }, |
| { |
| "epoch": 9.988160585512862, |
| "grad_norm": 0.8545377254486084, |
| "learning_rate": 9.309341665768773e-07, |
| "loss": 3.1243, |
| "step": 92800 |
| }, |
| { |
| "epoch": 9.99354213755247, |
| "grad_norm": 0.874809980392456, |
| "learning_rate": 6.076931365154616e-07, |
| "loss": 3.1167, |
| "step": 92850 |
| }, |
| { |
| "epoch": 9.998923689592079, |
| "grad_norm": 0.8542647361755371, |
| "learning_rate": 2.8445210645404587e-07, |
| "loss": 3.1182, |
| "step": 92900 |
| }, |
| { |
| "epoch": 10.0, |
| "step": 92910, |
| "total_flos": 7.7681075945472e+17, |
| "train_loss": 3.4603127508396456, |
| "train_runtime": 79796.1096, |
| "train_samples_per_second": 37.257, |
| "train_steps_per_second": 1.164 |
| } |
| ], |
| "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": 7.7681075945472e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
