| { | |
| "best_global_step": 3750, | |
| "best_metric": 0.5079798102378845, | |
| "best_model_checkpoint": "./results/checkpoint-3500", | |
| "epoch": 4.997752808988764, | |
| "eval_steps": 250, | |
| "global_step": 9455, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.026437541308658295, | |
| "grad_norm": 0.2475886046886444, | |
| "learning_rate": 0.00034625958983852136, | |
| "loss": 0.8762, | |
| "mean_token_accuracy": 0.8128470236063003, | |
| "num_tokens": 1638400.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05287508261731659, | |
| "grad_norm": 0.25265493988990784, | |
| "learning_rate": 0.000407611186724682, | |
| "loss": 0.6461, | |
| "mean_token_accuracy": 0.8552572166919709, | |
| "num_tokens": 3276800.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07931262392597488, | |
| "grad_norm": 0.21018819510936737, | |
| "learning_rate": 0.0004434995702624468, | |
| "loss": 0.635, | |
| "mean_token_accuracy": 0.8568304070830345, | |
| "num_tokens": 4915200.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.10575016523463318, | |
| "grad_norm": 0.23957596719264984, | |
| "learning_rate": 0.0004689627836108426, | |
| "loss": 0.6235, | |
| "mean_token_accuracy": 0.8593035244941711, | |
| "num_tokens": 6553600.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13218770654329148, | |
| "grad_norm": 0.192116379737854, | |
| "learning_rate": 0.0004887135863147016, | |
| "loss": 0.6217, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.13218770654329148, | |
| "eval_loss": 0.630805492401123, | |
| "eval_mean_token_accuracy": 0.8574222893971973, | |
| "eval_num_tokens": 8192000.0, | |
| "eval_runtime": 1444.1199, | |
| "eval_samples_per_second": 5.238, | |
| "eval_steps_per_second": 0.655, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.15862524785194976, | |
| "grad_norm": 0.2664429545402527, | |
| "learning_rate": 0.0004991822047759241, | |
| "loss": 0.6239, | |
| "mean_token_accuracy": 0.859162701368332, | |
| "num_tokens": 9830400.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.18506278916060806, | |
| "grad_norm": 0.5154497027397156, | |
| "learning_rate": 0.0004964562206956711, | |
| "loss": 0.6973, | |
| "mean_token_accuracy": 0.8490096586942673, | |
| "num_tokens": 11468800.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.21150033046926636, | |
| "grad_norm": 0.2535574436187744, | |
| "learning_rate": 0.0004937302366154182, | |
| "loss": 0.6079, | |
| "mean_token_accuracy": 0.8621737584471703, | |
| "num_tokens": 13107200.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23793787177792466, | |
| "grad_norm": 0.16894356906414032, | |
| "learning_rate": 0.0004910042525351653, | |
| "loss": 0.6082, | |
| "mean_token_accuracy": 0.8616935542225838, | |
| "num_tokens": 14745600.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.26437541308658297, | |
| "grad_norm": 0.1515253186225891, | |
| "learning_rate": 0.00048827826845491225, | |
| "loss": 0.6043, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.26437541308658297, | |
| "eval_loss": 0.6044251918792725, | |
| "eval_mean_token_accuracy": 0.8618915110637472, | |
| "eval_num_tokens": 16384000.0, | |
| "eval_runtime": 1444.8043, | |
| "eval_samples_per_second": 5.236, | |
| "eval_steps_per_second": 0.655, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.29081295439524124, | |
| "grad_norm": 0.15358857810497284, | |
| "learning_rate": 0.0004855522843746593, | |
| "loss": 0.6137, | |
| "mean_token_accuracy": 0.8617021070420742, | |
| "num_tokens": 18022400.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3172504957038995, | |
| "grad_norm": 0.34029558300971985, | |
| "learning_rate": 0.00048282630029440626, | |
| "loss": 0.5973, | |
| "mean_token_accuracy": 0.8634732463955879, | |
| "num_tokens": 19660800.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.34368803701255785, | |
| "grad_norm": 0.22880446910858154, | |
| "learning_rate": 0.00048010031621415335, | |
| "loss": 0.5984, | |
| "mean_token_accuracy": 0.8633871012926102, | |
| "num_tokens": 21299200.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.3701255783212161, | |
| "grad_norm": 0.17699383199214935, | |
| "learning_rate": 0.0004773743321339004, | |
| "loss": 0.5871, | |
| "mean_token_accuracy": 0.8653806248307228, | |
| "num_tokens": 22937600.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.3965631196298744, | |
| "grad_norm": 0.18763592839241028, | |
| "learning_rate": 0.00047464834805364736, | |
| "loss": 0.5889, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.3965631196298744, | |
| "eval_loss": 0.5848008394241333, | |
| "eval_mean_token_accuracy": 0.865356788007704, | |
| "eval_num_tokens": 24576000.0, | |
| "eval_runtime": 1443.8073, | |
| "eval_samples_per_second": 5.24, | |
| "eval_steps_per_second": 0.655, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4230006609385327, | |
| "grad_norm": 0.1546117663383484, | |
| "learning_rate": 0.0004719223639733944, | |
| "loss": 0.5821, | |
| "mean_token_accuracy": 0.8651845099031925, | |
| "num_tokens": 26214400.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.449438202247191, | |
| "grad_norm": 0.13760164380073547, | |
| "learning_rate": 0.0004691963798931415, | |
| "loss": 0.5824, | |
| "mean_token_accuracy": 0.8658082884550095, | |
| "num_tokens": 27852800.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.47587574355584933, | |
| "grad_norm": 0.1396537572145462, | |
| "learning_rate": 0.00046647039581288846, | |
| "loss": 0.5707, | |
| "mean_token_accuracy": 0.8684304764866829, | |
| "num_tokens": 29491200.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5023132848645075, | |
| "grad_norm": 0.13558807969093323, | |
| "learning_rate": 0.0004637444117326355, | |
| "loss": 0.58, | |
| "mean_token_accuracy": 0.8662952136993408, | |
| "num_tokens": 31129600.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.5287508261731659, | |
| "grad_norm": 0.12448519468307495, | |
| "learning_rate": 0.0004610184276523825, | |
| "loss": 0.5769, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5287508261731659, | |
| "eval_loss": 0.5692248940467834, | |
| "eval_mean_token_accuracy": 0.8682624975546218, | |
| "eval_num_tokens": 32768000.0, | |
| "eval_runtime": 1443.7648, | |
| "eval_samples_per_second": 5.24, | |
| "eval_steps_per_second": 0.655, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5551883674818242, | |
| "grad_norm": 0.12668457627296448, | |
| "learning_rate": 0.00045829244357212956, | |
| "loss": 0.5683, | |
| "mean_token_accuracy": 0.8675177201628685, | |
| "num_tokens": 34406400.0, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.5816259087904825, | |
| "grad_norm": 0.1283566802740097, | |
| "learning_rate": 0.0004555664594918766, | |
| "loss": 0.5686, | |
| "mean_token_accuracy": 0.8686253690719604, | |
| "num_tokens": 36044800.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.6080634500991408, | |
| "grad_norm": 0.15960244834423065, | |
| "learning_rate": 0.0004528404754116236, | |
| "loss": 0.5644, | |
| "mean_token_accuracy": 0.8688807463645936, | |
| "num_tokens": 37683200.0, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.634500991407799, | |
| "grad_norm": 0.11330343037843704, | |
| "learning_rate": 0.0004501144913313706, | |
| "loss": 0.5649, | |
| "mean_token_accuracy": 0.8686217039823532, | |
| "num_tokens": 39321600.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.6609385327164574, | |
| "grad_norm": 0.12131626158952713, | |
| "learning_rate": 0.0004473885072511177, | |
| "loss": 0.5638, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6609385327164574, | |
| "eval_loss": 0.5589728951454163, | |
| "eval_mean_token_accuracy": 0.8698779833493223, | |
| "eval_num_tokens": 40960000.0, | |
| "eval_runtime": 1443.9339, | |
| "eval_samples_per_second": 5.239, | |
| "eval_steps_per_second": 0.655, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.6873760740251157, | |
| "grad_norm": 0.135975182056427, | |
| "learning_rate": 0.0004446625231708647, | |
| "loss": 0.5606, | |
| "mean_token_accuracy": 0.8690710559487342, | |
| "num_tokens": 42598400.0, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.713813615333774, | |
| "grad_norm": 0.15456999838352203, | |
| "learning_rate": 0.0004419365390906117, | |
| "loss": 0.5495, | |
| "mean_token_accuracy": 0.8715310388803482, | |
| "num_tokens": 44236800.0, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.7402511566424322, | |
| "grad_norm": 0.17465010285377502, | |
| "learning_rate": 0.00043921055501035873, | |
| "loss": 0.5424, | |
| "mean_token_accuracy": 0.873205646276474, | |
| "num_tokens": 45875200.0, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7666886979510905, | |
| "grad_norm": 0.11691431701183319, | |
| "learning_rate": 0.0004364845709301058, | |
| "loss": 0.5407, | |
| "mean_token_accuracy": 0.8731714335083961, | |
| "num_tokens": 47513600.0, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.7931262392597488, | |
| "grad_norm": 0.1413394808769226, | |
| "learning_rate": 0.0004337585868498528, | |
| "loss": 0.5477, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.7931262392597488, | |
| "eval_loss": 0.5501642227172852, | |
| "eval_mean_token_accuracy": 0.8714489584619348, | |
| "eval_num_tokens": 49152000.0, | |
| "eval_runtime": 1443.5916, | |
| "eval_samples_per_second": 5.24, | |
| "eval_steps_per_second": 0.655, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8195637805684072, | |
| "grad_norm": 0.10578064620494843, | |
| "learning_rate": 0.00043103260276959983, | |
| "loss": 0.5508, | |
| "mean_token_accuracy": 0.8713572224974633, | |
| "num_tokens": 50790400.0, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.8460013218770654, | |
| "grad_norm": 0.10736281424760818, | |
| "learning_rate": 0.0004283066186893468, | |
| "loss": 0.5481, | |
| "mean_token_accuracy": 0.8717375376820564, | |
| "num_tokens": 52428800.0, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.8724388631857237, | |
| "grad_norm": 0.12125309556722641, | |
| "learning_rate": 0.0004255806346090939, | |
| "loss": 0.5537, | |
| "mean_token_accuracy": 0.86990652769804, | |
| "num_tokens": 54067200.0, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.898876404494382, | |
| "grad_norm": 0.10367817431688309, | |
| "learning_rate": 0.00042285465052884093, | |
| "loss": 0.5446, | |
| "mean_token_accuracy": 0.8718084079027176, | |
| "num_tokens": 55705600.0, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.9253139458030403, | |
| "grad_norm": 0.19373266398906708, | |
| "learning_rate": 0.0004201286664485879, | |
| "loss": 0.5486, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9253139458030403, | |
| "eval_loss": 0.5393346548080444, | |
| "eval_mean_token_accuracy": 0.8730530525236755, | |
| "eval_num_tokens": 57344000.0, | |
| "eval_runtime": 1444.4931, | |
| "eval_samples_per_second": 5.237, | |
| "eval_steps_per_second": 0.655, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.9517514871116987, | |
| "grad_norm": 0.10371921956539154, | |
| "learning_rate": 0.00041740268236833495, | |
| "loss": 0.5381, | |
| "mean_token_accuracy": 0.8720436228811741, | |
| "num_tokens": 58982400.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.9781890284203569, | |
| "grad_norm": 0.09864738583564758, | |
| "learning_rate": 0.00041467669828808203, | |
| "loss": 0.5321, | |
| "mean_token_accuracy": 0.8747250729799271, | |
| "num_tokens": 60620800.0, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.0042300066093852, | |
| "grad_norm": 0.12164577841758728, | |
| "learning_rate": 0.00041200523388943414, | |
| "loss": 0.519, | |
| "mean_token_accuracy": 0.8759235533966025, | |
| "num_tokens": 62234624.0, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.0306675479180436, | |
| "grad_norm": 0.16176024079322815, | |
| "learning_rate": 0.0004092792498091811, | |
| "loss": 0.4182, | |
| "mean_token_accuracy": 0.8925769734382629, | |
| "num_tokens": 63873024.0, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.057105089226702, | |
| "grad_norm": 0.09831411391496658, | |
| "learning_rate": 0.00040655326572892816, | |
| "loss": 0.426, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.057105089226702, | |
| "eval_loss": 0.5495159029960632, | |
| "eval_mean_token_accuracy": 0.8728643428321621, | |
| "eval_num_tokens": 65511424.0, | |
| "eval_runtime": 1443.7886, | |
| "eval_samples_per_second": 5.24, | |
| "eval_steps_per_second": 0.655, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0835426305353602, | |
| "grad_norm": 0.11064837127923965, | |
| "learning_rate": 0.00040382728164867513, | |
| "loss": 0.4203, | |
| "mean_token_accuracy": 0.891697817593813, | |
| "num_tokens": 67149824.0, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.1099801718440185, | |
| "grad_norm": 0.12444479018449783, | |
| "learning_rate": 0.0004011012975684222, | |
| "loss": 0.4324, | |
| "mean_token_accuracy": 0.8894703006744384, | |
| "num_tokens": 68788224.0, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.1364177131526767, | |
| "grad_norm": 0.10909954458475113, | |
| "learning_rate": 0.00039837531348816925, | |
| "loss": 0.4243, | |
| "mean_token_accuracy": 0.8913825675845146, | |
| "num_tokens": 70426624.0, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.162855254461335, | |
| "grad_norm": 0.09556654095649719, | |
| "learning_rate": 0.00039564932940791623, | |
| "loss": 0.426, | |
| "mean_token_accuracy": 0.8912683191895485, | |
| "num_tokens": 72065024.0, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.1892927957699935, | |
| "grad_norm": 0.11516989767551422, | |
| "learning_rate": 0.00039292334532766327, | |
| "loss": 0.4313, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.1892927957699935, | |
| "eval_loss": 0.5482946634292603, | |
| "eval_mean_token_accuracy": 0.873437058408971, | |
| "eval_num_tokens": 73703424.0, | |
| "eval_runtime": 1445.2243, | |
| "eval_samples_per_second": 5.234, | |
| "eval_steps_per_second": 0.655, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.2157303370786516, | |
| "grad_norm": 0.11027960479259491, | |
| "learning_rate": 0.00039019736124741035, | |
| "loss": 0.4393, | |
| "mean_token_accuracy": 0.889146803766489, | |
| "num_tokens": 75341824.0, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.24216787838731, | |
| "grad_norm": 0.118904247879982, | |
| "learning_rate": 0.00038747137716715733, | |
| "loss": 0.4408, | |
| "mean_token_accuracy": 0.8877504804730415, | |
| "num_tokens": 76980224.0, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.2686054196959682, | |
| "grad_norm": 0.10859134048223495, | |
| "learning_rate": 0.00038474539308690437, | |
| "loss": 0.4349, | |
| "mean_token_accuracy": 0.8895466673374176, | |
| "num_tokens": 78618624.0, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.2950429610046266, | |
| "grad_norm": 0.09974760562181473, | |
| "learning_rate": 0.0003820194090066514, | |
| "loss": 0.4375, | |
| "mean_token_accuracy": 0.8887909257411957, | |
| "num_tokens": 80257024.0, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 1.321480502313285, | |
| "grad_norm": 0.10627176612615585, | |
| "learning_rate": 0.00037929342492639843, | |
| "loss": 0.4354, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.321480502313285, | |
| "eval_loss": 0.5409240126609802, | |
| "eval_mean_token_accuracy": 0.8742006784639701, | |
| "eval_num_tokens": 81895424.0, | |
| "eval_runtime": 1443.9561, | |
| "eval_samples_per_second": 5.239, | |
| "eval_steps_per_second": 0.655, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.3479180436219431, | |
| "grad_norm": 0.10294341295957565, | |
| "learning_rate": 0.00037656744084614547, | |
| "loss": 0.4372, | |
| "mean_token_accuracy": 0.8892436383664608, | |
| "num_tokens": 83533824.0, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 1.3743555849306015, | |
| "grad_norm": 0.12393230944871902, | |
| "learning_rate": 0.0003738414567658925, | |
| "loss": 0.4386, | |
| "mean_token_accuracy": 0.8886467415094376, | |
| "num_tokens": 85172224.0, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.4007931262392597, | |
| "grad_norm": 0.09495224803686142, | |
| "learning_rate": 0.0003711154726856395, | |
| "loss": 0.4302, | |
| "mean_token_accuracy": 0.8906915840506554, | |
| "num_tokens": 86810624.0, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 1.427230667547918, | |
| "grad_norm": 0.10494447499513626, | |
| "learning_rate": 0.00036838948860538656, | |
| "loss": 0.4312, | |
| "mean_token_accuracy": 0.8900836914777756, | |
| "num_tokens": 88449024.0, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.4536682088565764, | |
| "grad_norm": 0.12757033109664917, | |
| "learning_rate": 0.0003656635045251336, | |
| "loss": 0.437, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.4536682088565764, | |
| "eval_loss": 0.53493332862854, | |
| "eval_mean_token_accuracy": 0.8759168862162382, | |
| "eval_num_tokens": 90087424.0, | |
| "eval_runtime": 1443.9928, | |
| "eval_samples_per_second": 5.239, | |
| "eval_steps_per_second": 0.655, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 1.4801057501652346, | |
| "grad_norm": 0.12719060480594635, | |
| "learning_rate": 0.0003629375204448806, | |
| "loss": 0.4384, | |
| "mean_token_accuracy": 0.8888401076197624, | |
| "num_tokens": 91725824.0, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.5065432914738928, | |
| "grad_norm": 0.1001153439283371, | |
| "learning_rate": 0.0003602115363646276, | |
| "loss": 0.443, | |
| "mean_token_accuracy": 0.8874511170387268, | |
| "num_tokens": 93364224.0, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 1.5329808327825512, | |
| "grad_norm": 0.10144038498401642, | |
| "learning_rate": 0.0003574855522843747, | |
| "loss": 0.4439, | |
| "mean_token_accuracy": 0.8875879681110382, | |
| "num_tokens": 95002624.0, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.5594183740912095, | |
| "grad_norm": 0.10192844271659851, | |
| "learning_rate": 0.0003547595682041217, | |
| "loss": 0.4345, | |
| "mean_token_accuracy": 0.8896059301495552, | |
| "num_tokens": 96641024.0, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 1.585855915399868, | |
| "grad_norm": 0.11319652944803238, | |
| "learning_rate": 0.0003520335841238687, | |
| "loss": 0.4378, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.585855915399868, | |
| "eval_loss": 0.5288907289505005, | |
| "eval_mean_token_accuracy": 0.8770441555951665, | |
| "eval_num_tokens": 98279424.0, | |
| "eval_runtime": 1443.9606, | |
| "eval_samples_per_second": 5.239, | |
| "eval_steps_per_second": 0.655, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.612293456708526, | |
| "grad_norm": 0.10589715093374252, | |
| "learning_rate": 0.00034930760004361574, | |
| "loss": 0.4334, | |
| "mean_token_accuracy": 0.8895524734258652, | |
| "num_tokens": 99917824.0, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 1.6387309980171842, | |
| "grad_norm": 0.09860669076442719, | |
| "learning_rate": 0.0003465816159633628, | |
| "loss": 0.4377, | |
| "mean_token_accuracy": 0.8887499925494194, | |
| "num_tokens": 101556224.0, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.6651685393258426, | |
| "grad_norm": 0.11345923691987991, | |
| "learning_rate": 0.0003438556318831098, | |
| "loss": 0.4413, | |
| "mean_token_accuracy": 0.8883473780751229, | |
| "num_tokens": 103194624.0, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 1.691606080634501, | |
| "grad_norm": 0.10729037970304489, | |
| "learning_rate": 0.00034112964780285684, | |
| "loss": 0.4365, | |
| "mean_token_accuracy": 0.8893462771177292, | |
| "num_tokens": 104833024.0, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.7180436219431594, | |
| "grad_norm": 0.0929078608751297, | |
| "learning_rate": 0.0003384036637226039, | |
| "loss": 0.4339, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 1.7180436219431594, | |
| "eval_loss": 0.522659182548523, | |
| "eval_mean_token_accuracy": 0.8778351331740051, | |
| "eval_num_tokens": 106471424.0, | |
| "eval_runtime": 1444.1286, | |
| "eval_samples_per_second": 5.238, | |
| "eval_steps_per_second": 0.655, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 1.7444811632518176, | |
| "grad_norm": 0.09583564847707748, | |
| "learning_rate": 0.0003356776796423509, | |
| "loss": 0.4292, | |
| "mean_token_accuracy": 0.8906918911635876, | |
| "num_tokens": 108109824.0, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.7709187045604757, | |
| "grad_norm": 0.09396171569824219, | |
| "learning_rate": 0.00033295169556209794, | |
| "loss": 0.4259, | |
| "mean_token_accuracy": 0.8916892632842064, | |
| "num_tokens": 109748224.0, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 1.7973562458691341, | |
| "grad_norm": 0.08548162132501602, | |
| "learning_rate": 0.0003302257114818449, | |
| "loss": 0.4327, | |
| "mean_token_accuracy": 0.8902602550387383, | |
| "num_tokens": 111386624.0, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.8237937871777925, | |
| "grad_norm": 0.1594037264585495, | |
| "learning_rate": 0.000327499727401592, | |
| "loss": 0.4279, | |
| "mean_token_accuracy": 0.8916428324580192, | |
| "num_tokens": 113025024.0, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 1.8502313284864509, | |
| "grad_norm": 0.09137943387031555, | |
| "learning_rate": 0.00032477374332133904, | |
| "loss": 0.4259, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.8502313284864509, | |
| "eval_loss": 0.5174182653427124, | |
| "eval_mean_token_accuracy": 0.878870385834833, | |
| "eval_num_tokens": 114663424.0, | |
| "eval_runtime": 1444.5837, | |
| "eval_samples_per_second": 5.237, | |
| "eval_steps_per_second": 0.655, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.876668869795109, | |
| "grad_norm": 0.13611243665218353, | |
| "learning_rate": 0.000322047759241086, | |
| "loss": 0.4334, | |
| "mean_token_accuracy": 0.8908357685804367, | |
| "num_tokens": 116301824.0, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 1.9031064111037672, | |
| "grad_norm": 0.10409994423389435, | |
| "learning_rate": 0.00031932177516083305, | |
| "loss": 0.4396, | |
| "mean_token_accuracy": 0.8884219127893448, | |
| "num_tokens": 117940224.0, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.9295439524124256, | |
| "grad_norm": 0.08799432218074799, | |
| "learning_rate": 0.00031659579108058014, | |
| "loss": 0.4238, | |
| "mean_token_accuracy": 0.892235452234745, | |
| "num_tokens": 119578624.0, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 1.955981493721084, | |
| "grad_norm": 0.0893888846039772, | |
| "learning_rate": 0.0003138698070003271, | |
| "loss": 0.4292, | |
| "mean_token_accuracy": 0.8910862573981285, | |
| "num_tokens": 121217024.0, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.9824190350297424, | |
| "grad_norm": 0.08708135783672333, | |
| "learning_rate": 0.00031114382292007415, | |
| "loss": 0.4319, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 1.9824190350297424, | |
| "eval_loss": 0.5079798102378845, | |
| "eval_mean_token_accuracy": 0.880161566348711, | |
| "eval_num_tokens": 122855424.0, | |
| "eval_runtime": 1444.3228, | |
| "eval_samples_per_second": 5.238, | |
| "eval_steps_per_second": 0.655, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 2.0084600132187704, | |
| "grad_norm": 0.10443489998579025, | |
| "learning_rate": 0.00030847235852142626, | |
| "loss": 0.3692, | |
| "mean_token_accuracy": 0.89680361027081, | |
| "num_tokens": 124469248.0, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.034897554527429, | |
| "grad_norm": 0.09454915672540665, | |
| "learning_rate": 0.00030574637444117324, | |
| "loss": 0.2389, | |
| "mean_token_accuracy": 0.9304979285597801, | |
| "num_tokens": 126107648.0, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 2.061335095836087, | |
| "grad_norm": 0.10566703230142593, | |
| "learning_rate": 0.0003030203903609203, | |
| "loss": 0.2421, | |
| "mean_token_accuracy": 0.9295185798406601, | |
| "num_tokens": 127746048.0, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.0877726371447456, | |
| "grad_norm": 0.10303833335638046, | |
| "learning_rate": 0.00030029440628066736, | |
| "loss": 0.2398, | |
| "mean_token_accuracy": 0.9298442155122757, | |
| "num_tokens": 129384448.0, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 2.114210178453404, | |
| "grad_norm": 0.0934186801314354, | |
| "learning_rate": 0.00029756842220041434, | |
| "loss": 0.2444, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.114210178453404, | |
| "eval_loss": 0.5857627391815186, | |
| "eval_mean_token_accuracy": 0.8763663001342757, | |
| "eval_num_tokens": 131022848.0, | |
| "eval_runtime": 1444.4937, | |
| "eval_samples_per_second": 5.237, | |
| "eval_steps_per_second": 0.655, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.140647719762062, | |
| "grad_norm": 0.11742925643920898, | |
| "learning_rate": 0.0002948424381201614, | |
| "loss": 0.2433, | |
| "mean_token_accuracy": 0.9287463417649269, | |
| "num_tokens": 132661248.0, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 2.1670852610707203, | |
| "grad_norm": 0.09310949593782425, | |
| "learning_rate": 0.0002921164540399084, | |
| "loss": 0.2502, | |
| "mean_token_accuracy": 0.9270796775817871, | |
| "num_tokens": 134299648.0, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.1935228023793787, | |
| "grad_norm": 0.10311378538608551, | |
| "learning_rate": 0.00028939046995965544, | |
| "loss": 0.2487, | |
| "mean_token_accuracy": 0.9274065324664116, | |
| "num_tokens": 135938048.0, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 2.219960343688037, | |
| "grad_norm": 0.10651703178882599, | |
| "learning_rate": 0.00028666448587940247, | |
| "loss": 0.2481, | |
| "mean_token_accuracy": 0.9273692658543586, | |
| "num_tokens": 137576448.0, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.2463978849966955, | |
| "grad_norm": 0.11140276491641998, | |
| "learning_rate": 0.0002839385017991495, | |
| "loss": 0.2513, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 2.2463978849966955, | |
| "eval_loss": 0.5914037823677063, | |
| "eval_mean_token_accuracy": 0.8769333329815693, | |
| "eval_num_tokens": 139214848.0, | |
| "eval_runtime": 1445.0132, | |
| "eval_samples_per_second": 5.235, | |
| "eval_steps_per_second": 0.655, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 2.2728354263053534, | |
| "grad_norm": 0.10259566456079483, | |
| "learning_rate": 0.00028121251771889654, | |
| "loss": 0.253, | |
| "mean_token_accuracy": 0.9264763659238815, | |
| "num_tokens": 140853248.0, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.299272967614012, | |
| "grad_norm": 0.10440666228532791, | |
| "learning_rate": 0.00027848653363864357, | |
| "loss": 0.25, | |
| "mean_token_accuracy": 0.9270870080590248, | |
| "num_tokens": 142491648.0, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 2.32571050892267, | |
| "grad_norm": 0.11085245013237, | |
| "learning_rate": 0.0002757605495583906, | |
| "loss": 0.2507, | |
| "mean_token_accuracy": 0.9270876184105873, | |
| "num_tokens": 144130048.0, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.3521480502313286, | |
| "grad_norm": 0.11133650690317154, | |
| "learning_rate": 0.0002730345654781376, | |
| "loss": 0.2509, | |
| "mean_token_accuracy": 0.9270026966929436, | |
| "num_tokens": 145768448.0, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 2.378585591539987, | |
| "grad_norm": 0.09678333252668381, | |
| "learning_rate": 0.00027030858139788467, | |
| "loss": 0.252, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.378585591539987, | |
| "eval_loss": 0.588071882724762, | |
| "eval_mean_token_accuracy": 0.8775507902242164, | |
| "eval_num_tokens": 147406848.0, | |
| "eval_runtime": 1444.9835, | |
| "eval_samples_per_second": 5.235, | |
| "eval_steps_per_second": 0.655, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.405023132848645, | |
| "grad_norm": 0.09919407218694687, | |
| "learning_rate": 0.0002675825973176317, | |
| "loss": 0.2597, | |
| "mean_token_accuracy": 0.9257050442695618, | |
| "num_tokens": 149045248.0, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 2.4314606741573033, | |
| "grad_norm": 0.09900713711977005, | |
| "learning_rate": 0.0002648566132373787, | |
| "loss": 0.2594, | |
| "mean_token_accuracy": 0.9249798482656479, | |
| "num_tokens": 150683648.0, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.4578982154659617, | |
| "grad_norm": 0.11208795011043549, | |
| "learning_rate": 0.0002621306291571257, | |
| "loss": 0.2512, | |
| "mean_token_accuracy": 0.926915942132473, | |
| "num_tokens": 152322048.0, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 2.48433575677462, | |
| "grad_norm": 0.1047317236661911, | |
| "learning_rate": 0.0002594046450768728, | |
| "loss": 0.2575, | |
| "mean_token_accuracy": 0.9254233974218369, | |
| "num_tokens": 153960448.0, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 2.5107732980832784, | |
| "grad_norm": 0.10790830850601196, | |
| "learning_rate": 0.0002566786609966198, | |
| "loss": 0.2597, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 2.5107732980832784, | |
| "eval_loss": 0.5828577876091003, | |
| "eval_mean_token_accuracy": 0.8786447106153717, | |
| "eval_num_tokens": 155598848.0, | |
| "eval_runtime": 1445.085, | |
| "eval_samples_per_second": 5.235, | |
| "eval_steps_per_second": 0.655, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 2.5372108393919364, | |
| "grad_norm": 0.1063418760895729, | |
| "learning_rate": 0.0002539526769163668, | |
| "loss": 0.2532, | |
| "mean_token_accuracy": 0.9256470040977001, | |
| "num_tokens": 157237248.0, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.5636483807005948, | |
| "grad_norm": 0.10517441481351852, | |
| "learning_rate": 0.00025122669283611385, | |
| "loss": 0.257, | |
| "mean_token_accuracy": 0.92532870054245, | |
| "num_tokens": 158875648.0, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 2.590085922009253, | |
| "grad_norm": 0.10175232589244843, | |
| "learning_rate": 0.0002485007087558609, | |
| "loss": 0.2578, | |
| "mean_token_accuracy": 0.9254624971747398, | |
| "num_tokens": 160514048.0, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 2.6165234633179115, | |
| "grad_norm": 0.10122060030698776, | |
| "learning_rate": 0.0002457747246756079, | |
| "loss": 0.2585, | |
| "mean_token_accuracy": 0.925160077214241, | |
| "num_tokens": 162152448.0, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 2.64296100462657, | |
| "grad_norm": 0.17676186561584473, | |
| "learning_rate": 0.00024304874059535492, | |
| "loss": 0.2549, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.64296100462657, | |
| "eval_loss": 0.5834685564041138, | |
| "eval_mean_token_accuracy": 0.8788562039618261, | |
| "eval_num_tokens": 163790848.0, | |
| "eval_runtime": 1444.3995, | |
| "eval_samples_per_second": 5.237, | |
| "eval_steps_per_second": 0.655, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.669398545935228, | |
| "grad_norm": 0.09294166415929794, | |
| "learning_rate": 0.00024032275651510195, | |
| "loss": 0.2578, | |
| "mean_token_accuracy": 0.9256011824309825, | |
| "num_tokens": 165429248.0, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 2.6958360872438862, | |
| "grad_norm": 0.10612129420042038, | |
| "learning_rate": 0.000237596772434849, | |
| "loss": 0.2525, | |
| "mean_token_accuracy": 0.9266947796940803, | |
| "num_tokens": 167067648.0, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 2.7222736285525446, | |
| "grad_norm": 0.09773480892181396, | |
| "learning_rate": 0.00023487078835459602, | |
| "loss": 0.2574, | |
| "mean_token_accuracy": 0.9255730789899826, | |
| "num_tokens": 168706048.0, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 2.748711169861203, | |
| "grad_norm": 0.10722821205854416, | |
| "learning_rate": 0.00023214480427434303, | |
| "loss": 0.255, | |
| "mean_token_accuracy": 0.9261156010627747, | |
| "num_tokens": 170344448.0, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.7751487111698614, | |
| "grad_norm": 0.10466789454221725, | |
| "learning_rate": 0.00022941882019409009, | |
| "loss": 0.2579, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 2.7751487111698614, | |
| "eval_loss": 0.5760974287986755, | |
| "eval_mean_token_accuracy": 0.8801544366628876, | |
| "eval_num_tokens": 171982848.0, | |
| "eval_runtime": 1445.5403, | |
| "eval_samples_per_second": 5.233, | |
| "eval_steps_per_second": 0.654, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 2.8015862524785193, | |
| "grad_norm": 0.1087755486369133, | |
| "learning_rate": 0.0002266928361138371, | |
| "loss": 0.2609, | |
| "mean_token_accuracy": 0.9251267817616463, | |
| "num_tokens": 173621248.0, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 2.8280237937871777, | |
| "grad_norm": 0.09775590896606445, | |
| "learning_rate": 0.00022396685203358413, | |
| "loss": 0.2557, | |
| "mean_token_accuracy": 0.9260593932867051, | |
| "num_tokens": 175259648.0, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 2.854461335095836, | |
| "grad_norm": 0.09694032371044159, | |
| "learning_rate": 0.00022124086795333116, | |
| "loss": 0.2575, | |
| "mean_token_accuracy": 0.9255865186452865, | |
| "num_tokens": 176898048.0, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.8808988764044945, | |
| "grad_norm": 0.09860485792160034, | |
| "learning_rate": 0.0002185148838730782, | |
| "loss": 0.2567, | |
| "mean_token_accuracy": 0.9257099312543869, | |
| "num_tokens": 178536448.0, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 2.907336417713153, | |
| "grad_norm": 0.10467279702425003, | |
| "learning_rate": 0.0002157888997928252, | |
| "loss": 0.2537, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.907336417713153, | |
| "eval_loss": 0.5691426396369934, | |
| "eval_mean_token_accuracy": 0.8808775775402343, | |
| "eval_num_tokens": 180174848.0, | |
| "eval_runtime": 1445.1515, | |
| "eval_samples_per_second": 5.235, | |
| "eval_steps_per_second": 0.655, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.933773959021811, | |
| "grad_norm": 0.13350491225719452, | |
| "learning_rate": 0.00021306291571257226, | |
| "loss": 0.2548, | |
| "mean_token_accuracy": 0.9265344049036502, | |
| "num_tokens": 181813248.0, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 2.960211500330469, | |
| "grad_norm": 0.10815941542387009, | |
| "learning_rate": 0.00021033693163231926, | |
| "loss": 0.2529, | |
| "mean_token_accuracy": 0.9266226869821549, | |
| "num_tokens": 183451648.0, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.9866490416391276, | |
| "grad_norm": 0.09711950272321701, | |
| "learning_rate": 0.0002076109475520663, | |
| "loss": 0.2477, | |
| "mean_token_accuracy": 0.928189767897129, | |
| "num_tokens": 185090048.0, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 3.012690019828156, | |
| "grad_norm": 0.07202441990375519, | |
| "learning_rate": 0.0002049394831534184, | |
| "loss": 0.1737, | |
| "mean_token_accuracy": 0.9494730455016122, | |
| "num_tokens": 186703872.0, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 3.0391275611368145, | |
| "grad_norm": 0.09105918556451797, | |
| "learning_rate": 0.0002022134990731654, | |
| "loss": 0.0857, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 3.0391275611368145, | |
| "eval_loss": 0.7057962417602539, | |
| "eval_mean_token_accuracy": 0.8797050480736739, | |
| "eval_num_tokens": 188342272.0, | |
| "eval_runtime": 1445.2714, | |
| "eval_samples_per_second": 5.234, | |
| "eval_steps_per_second": 0.655, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 3.0655651024454724, | |
| "grad_norm": 0.08339951187372208, | |
| "learning_rate": 0.00019948751499291245, | |
| "loss": 0.0836, | |
| "mean_token_accuracy": 0.9740020345151424, | |
| "num_tokens": 189980672.0, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 3.092002643754131, | |
| "grad_norm": 0.07610207796096802, | |
| "learning_rate": 0.00019676153091265948, | |
| "loss": 0.0835, | |
| "mean_token_accuracy": 0.9742461076378822, | |
| "num_tokens": 191619072.0, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 3.118440185062789, | |
| "grad_norm": 0.08562301844358444, | |
| "learning_rate": 0.0001940355468324065, | |
| "loss": 0.0841, | |
| "mean_token_accuracy": 0.9739308601617813, | |
| "num_tokens": 193257472.0, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 3.1448777263714476, | |
| "grad_norm": 0.08178895711898804, | |
| "learning_rate": 0.00019130956275215352, | |
| "loss": 0.0824, | |
| "mean_token_accuracy": 0.9746059566736222, | |
| "num_tokens": 194895872.0, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 3.1713152676801055, | |
| "grad_norm": 0.08008107542991638, | |
| "learning_rate": 0.00018858357867190058, | |
| "loss": 0.0826, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.1713152676801055, | |
| "eval_loss": 0.718048095703125, | |
| "eval_mean_token_accuracy": 0.8796352478842907, | |
| "eval_num_tokens": 196534272.0, | |
| "eval_runtime": 1444.6019, | |
| "eval_samples_per_second": 5.237, | |
| "eval_steps_per_second": 0.655, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.197752808988764, | |
| "grad_norm": 0.08383160084486008, | |
| "learning_rate": 0.00018585759459164758, | |
| "loss": 0.0833, | |
| "mean_token_accuracy": 0.9743710477650166, | |
| "num_tokens": 198172672.0, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 3.2241903502974223, | |
| "grad_norm": 0.08407159149646759, | |
| "learning_rate": 0.00018313161051139462, | |
| "loss": 0.0843, | |
| "mean_token_accuracy": 0.9739204734563828, | |
| "num_tokens": 199811072.0, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 3.2506278916060807, | |
| "grad_norm": 0.0758429542183876, | |
| "learning_rate": 0.00018040562643114165, | |
| "loss": 0.0825, | |
| "mean_token_accuracy": 0.9741593536734581, | |
| "num_tokens": 201449472.0, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 3.277065432914739, | |
| "grad_norm": 0.08588337153196335, | |
| "learning_rate": 0.00017767964235088868, | |
| "loss": 0.0841, | |
| "mean_token_accuracy": 0.9740646561980247, | |
| "num_tokens": 203087872.0, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 3.303502974223397, | |
| "grad_norm": 0.08413270860910416, | |
| "learning_rate": 0.0001749536582706357, | |
| "loss": 0.0859, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 3.303502974223397, | |
| "eval_loss": 0.7220771908760071, | |
| "eval_mean_token_accuracy": 0.8795976611700925, | |
| "eval_num_tokens": 204726272.0, | |
| "eval_runtime": 1445.7274, | |
| "eval_samples_per_second": 5.233, | |
| "eval_steps_per_second": 0.654, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 3.3299405155320554, | |
| "grad_norm": 0.0801534354686737, | |
| "learning_rate": 0.00017222767419038275, | |
| "loss": 0.0856, | |
| "mean_token_accuracy": 0.9733324344456196, | |
| "num_tokens": 206364672.0, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 3.3563780568407138, | |
| "grad_norm": 0.08935518562793732, | |
| "learning_rate": 0.00016950169011012976, | |
| "loss": 0.0851, | |
| "mean_token_accuracy": 0.9735490143299103, | |
| "num_tokens": 208003072.0, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 3.382815598149372, | |
| "grad_norm": 0.08593747019767761, | |
| "learning_rate": 0.0001667757060298768, | |
| "loss": 0.0898, | |
| "mean_token_accuracy": 0.9727987706661224, | |
| "num_tokens": 209641472.0, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.4092531394580305, | |
| "grad_norm": 0.08515259623527527, | |
| "learning_rate": 0.00016404972194962382, | |
| "loss": 0.0862, | |
| "mean_token_accuracy": 0.9732246008515358, | |
| "num_tokens": 211279872.0, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 3.4356906807666885, | |
| "grad_norm": 0.08400440216064453, | |
| "learning_rate": 0.00016132373786937086, | |
| "loss": 0.0848, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.4356906807666885, | |
| "eval_loss": 0.7208439111709595, | |
| "eval_mean_token_accuracy": 0.8801698074154319, | |
| "eval_num_tokens": 212918272.0, | |
| "eval_runtime": 1445.2121, | |
| "eval_samples_per_second": 5.235, | |
| "eval_steps_per_second": 0.655, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.462128222075347, | |
| "grad_norm": 0.09465966373682022, | |
| "learning_rate": 0.00015859775378911786, | |
| "loss": 0.0845, | |
| "mean_token_accuracy": 0.9737170253694057, | |
| "num_tokens": 214556672.0, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 3.4885657633840053, | |
| "grad_norm": 0.08350127190351486, | |
| "learning_rate": 0.0001558717697088649, | |
| "loss": 0.0845, | |
| "mean_token_accuracy": 0.9738056141138077, | |
| "num_tokens": 216195072.0, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 3.5150033046926636, | |
| "grad_norm": 0.09258560091257095, | |
| "learning_rate": 0.00015314578562861193, | |
| "loss": 0.0867, | |
| "mean_token_accuracy": 0.973144568502903, | |
| "num_tokens": 217833472.0, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 3.541440846001322, | |
| "grad_norm": 0.08615857362747192, | |
| "learning_rate": 0.00015041980154835896, | |
| "loss": 0.0836, | |
| "mean_token_accuracy": 0.9740035626292228, | |
| "num_tokens": 219471872.0, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 3.56787838730998, | |
| "grad_norm": 0.08243568986654282, | |
| "learning_rate": 0.00014769381746810597, | |
| "loss": 0.0847, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 3.56787838730998, | |
| "eval_loss": 0.7248117923736572, | |
| "eval_mean_token_accuracy": 0.8803664739489807, | |
| "eval_num_tokens": 221110272.0, | |
| "eval_runtime": 1445.5737, | |
| "eval_samples_per_second": 5.233, | |
| "eval_steps_per_second": 0.654, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 3.5943159286186384, | |
| "grad_norm": 0.08806952089071274, | |
| "learning_rate": 0.00014496783338785303, | |
| "loss": 0.0861, | |
| "mean_token_accuracy": 0.9733803936839104, | |
| "num_tokens": 222748672.0, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.6207534699272967, | |
| "grad_norm": 0.08953431993722916, | |
| "learning_rate": 0.00014224184930760003, | |
| "loss": 0.0856, | |
| "mean_token_accuracy": 0.9733651205897331, | |
| "num_tokens": 224387072.0, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 3.647191011235955, | |
| "grad_norm": 0.08528242260217667, | |
| "learning_rate": 0.00013951586522734707, | |
| "loss": 0.0852, | |
| "mean_token_accuracy": 0.9733651199936867, | |
| "num_tokens": 226025472.0, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 3.6736285525446135, | |
| "grad_norm": 0.08363524079322815, | |
| "learning_rate": 0.0001367898811470941, | |
| "loss": 0.0845, | |
| "mean_token_accuracy": 0.9736650982499122, | |
| "num_tokens": 227663872.0, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 3.7000660938532715, | |
| "grad_norm": 0.07602488994598389, | |
| "learning_rate": 0.00013406389706684113, | |
| "loss": 0.0845, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.7000660938532715, | |
| "eval_loss": 0.7213518023490906, | |
| "eval_mean_token_accuracy": 0.8806469935695415, | |
| "eval_num_tokens": 229302272.0, | |
| "eval_runtime": 1444.415, | |
| "eval_samples_per_second": 5.237, | |
| "eval_steps_per_second": 0.655, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.72650363516193, | |
| "grad_norm": 0.08517055958509445, | |
| "learning_rate": 0.00013133791298658814, | |
| "loss": 0.0847, | |
| "mean_token_accuracy": 0.9737151943147182, | |
| "num_tokens": 230940672.0, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 3.7529411764705882, | |
| "grad_norm": 0.0767749696969986, | |
| "learning_rate": 0.0001286119289063352, | |
| "loss": 0.0828, | |
| "mean_token_accuracy": 0.9742509955167771, | |
| "num_tokens": 232579072.0, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 3.7793787177792466, | |
| "grad_norm": 0.08359658718109131, | |
| "learning_rate": 0.0001258859448260822, | |
| "loss": 0.0849, | |
| "mean_token_accuracy": 0.973515414595604, | |
| "num_tokens": 234217472.0, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 3.805816259087905, | |
| "grad_norm": 0.07397591322660446, | |
| "learning_rate": 0.00012315996074582924, | |
| "loss": 0.0835, | |
| "mean_token_accuracy": 0.9739962309598923, | |
| "num_tokens": 235855872.0, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.832253800396563, | |
| "grad_norm": 0.08941724896430969, | |
| "learning_rate": 0.00012043397666557627, | |
| "loss": 0.0821, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 3.832253800396563, | |
| "eval_loss": 0.727186381816864, | |
| "eval_mean_token_accuracy": 0.881018961017782, | |
| "eval_num_tokens": 237494272.0, | |
| "eval_runtime": 1445.0501, | |
| "eval_samples_per_second": 5.235, | |
| "eval_steps_per_second": 0.655, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 3.8586913417052213, | |
| "grad_norm": 0.08002132922410965, | |
| "learning_rate": 0.0001177079925853233, | |
| "loss": 0.0824, | |
| "mean_token_accuracy": 0.9744101479649544, | |
| "num_tokens": 239132672.0, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 3.8851288830138797, | |
| "grad_norm": 0.08109039068222046, | |
| "learning_rate": 0.00011498200850507034, | |
| "loss": 0.0842, | |
| "mean_token_accuracy": 0.9738893151283264, | |
| "num_tokens": 240771072.0, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 3.911566424322538, | |
| "grad_norm": 0.08812084048986435, | |
| "learning_rate": 0.00011225602442481736, | |
| "loss": 0.0808, | |
| "mean_token_accuracy": 0.9748161256313324, | |
| "num_tokens": 242409472.0, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.9380039656311965, | |
| "grad_norm": 0.08132945001125336, | |
| "learning_rate": 0.00010953004034456439, | |
| "loss": 0.0824, | |
| "mean_token_accuracy": 0.9742913186550141, | |
| "num_tokens": 244047872.0, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 3.9644415069398544, | |
| "grad_norm": 0.08432092517614365, | |
| "learning_rate": 0.00010680405626431142, | |
| "loss": 0.0815, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.9644415069398544, | |
| "eval_loss": 0.727695643901825, | |
| "eval_mean_token_accuracy": 0.8814008755109275, | |
| "eval_num_tokens": 245686272.0, | |
| "eval_runtime": 1444.3382, | |
| "eval_samples_per_second": 5.238, | |
| "eval_steps_per_second": 0.655, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.990879048248513, | |
| "grad_norm": 0.08465278148651123, | |
| "learning_rate": 0.00010407807218405844, | |
| "loss": 0.0823, | |
| "mean_token_accuracy": 0.9746123726665974, | |
| "num_tokens": 247324672.0, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 4.016920026437541, | |
| "grad_norm": 0.04739232733845711, | |
| "learning_rate": 0.00010135208810380548, | |
| "loss": 0.0478, | |
| "mean_token_accuracy": 0.9862868610372398, | |
| "num_tokens": 248938496.0, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 4.0433575677462, | |
| "grad_norm": 0.05006248131394386, | |
| "learning_rate": 9.86261040235525e-05, | |
| "loss": 0.0278, | |
| "mean_token_accuracy": 0.9926484674215317, | |
| "num_tokens": 250576896.0, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 4.069795109054858, | |
| "grad_norm": 0.048755794763565063, | |
| "learning_rate": 9.590011994329953e-05, | |
| "loss": 0.0275, | |
| "mean_token_accuracy": 0.9927639365196228, | |
| "num_tokens": 252215296.0, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 4.0962326503635165, | |
| "grad_norm": 0.04973224923014641, | |
| "learning_rate": 9.317413586304656e-05, | |
| "loss": 0.0269, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 4.0962326503635165, | |
| "eval_loss": 0.8192626237869263, | |
| "eval_mean_token_accuracy": 0.88220195217062, | |
| "eval_num_tokens": 253853696.0, | |
| "eval_runtime": 1444.8032, | |
| "eval_samples_per_second": 5.236, | |
| "eval_steps_per_second": 0.655, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 4.122670191672174, | |
| "grad_norm": 0.05098351836204529, | |
| "learning_rate": 9.044815178279358e-05, | |
| "loss": 0.0268, | |
| "mean_token_accuracy": 0.9928580224514008, | |
| "num_tokens": 255492096.0, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 4.149107732980832, | |
| "grad_norm": 0.04829578474164009, | |
| "learning_rate": 8.772216770254061e-05, | |
| "loss": 0.0272, | |
| "mean_token_accuracy": 0.9927712675929069, | |
| "num_tokens": 257130496.0, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 4.175545274289491, | |
| "grad_norm": 0.04151781275868416, | |
| "learning_rate": 8.499618362228765e-05, | |
| "loss": 0.0269, | |
| "mean_token_accuracy": 0.9928335842490196, | |
| "num_tokens": 258768896.0, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 4.201982815598149, | |
| "grad_norm": 0.05445162579417229, | |
| "learning_rate": 8.227019954203467e-05, | |
| "loss": 0.0265, | |
| "mean_token_accuracy": 0.9929741024971008, | |
| "num_tokens": 260407296.0, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 4.228420356906808, | |
| "grad_norm": 0.03888251632452011, | |
| "learning_rate": 7.95442154617817e-05, | |
| "loss": 0.0273, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.228420356906808, | |
| "eval_loss": 0.8266236782073975, | |
| "eval_mean_token_accuracy": 0.8823872005712659, | |
| "eval_num_tokens": 262045696.0, | |
| "eval_runtime": 1444.6622, | |
| "eval_samples_per_second": 5.237, | |
| "eval_steps_per_second": 0.655, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.254857898215466, | |
| "grad_norm": 0.049200285226106644, | |
| "learning_rate": 7.681823138152873e-05, | |
| "loss": 0.0265, | |
| "mean_token_accuracy": 0.9928293077647686, | |
| "num_tokens": 263684096.0, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 4.281295439524124, | |
| "grad_norm": 0.04635064676403999, | |
| "learning_rate": 7.409224730127575e-05, | |
| "loss": 0.0259, | |
| "mean_token_accuracy": 0.9931225627660751, | |
| "num_tokens": 265322496.0, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 4.307732980832783, | |
| "grad_norm": 0.047765497118234634, | |
| "learning_rate": 7.136626322102279e-05, | |
| "loss": 0.0269, | |
| "mean_token_accuracy": 0.992902010679245, | |
| "num_tokens": 266960896.0, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 4.334170522141441, | |
| "grad_norm": 0.040082089602947235, | |
| "learning_rate": 6.864027914076983e-05, | |
| "loss": 0.0327, | |
| "mean_token_accuracy": 0.9925085598230362, | |
| "num_tokens": 268599296.0, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 4.360608063450099, | |
| "grad_norm": 0.040357593446969986, | |
| "learning_rate": 6.591429506051685e-05, | |
| "loss": 0.0262, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 4.360608063450099, | |
| "eval_loss": 0.8343076109886169, | |
| "eval_mean_token_accuracy": 0.8827618814169983, | |
| "eval_num_tokens": 270237696.0, | |
| "eval_runtime": 1444.2593, | |
| "eval_samples_per_second": 5.238, | |
| "eval_steps_per_second": 0.655, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 4.387045604758757, | |
| "grad_norm": 0.04214048385620117, | |
| "learning_rate": 6.318831098026388e-05, | |
| "loss": 0.0258, | |
| "mean_token_accuracy": 0.9930578023195267, | |
| "num_tokens": 271876096.0, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 4.413483146067415, | |
| "grad_norm": 0.0495334193110466, | |
| "learning_rate": 6.0462326900010904e-05, | |
| "loss": 0.0259, | |
| "mean_token_accuracy": 0.9930883496999741, | |
| "num_tokens": 273514496.0, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 4.439920687376074, | |
| "grad_norm": 0.04878053814172745, | |
| "learning_rate": 5.773634281975793e-05, | |
| "loss": 0.0256, | |
| "mean_token_accuracy": 0.9932361990213394, | |
| "num_tokens": 275152896.0, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 4.466358228684732, | |
| "grad_norm": 0.04534803330898285, | |
| "learning_rate": 5.5010358739504963e-05, | |
| "loss": 0.0257, | |
| "mean_token_accuracy": 0.9931818246841431, | |
| "num_tokens": 276791296.0, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 4.492795769993391, | |
| "grad_norm": 0.04720517247915268, | |
| "learning_rate": 5.228437465925199e-05, | |
| "loss": 0.0255, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.492795769993391, | |
| "eval_loss": 0.8357202410697937, | |
| "eval_mean_token_accuracy": 0.8828653151338751, | |
| "eval_num_tokens": 278429696.0, | |
| "eval_runtime": 1445.1121, | |
| "eval_samples_per_second": 5.235, | |
| "eval_steps_per_second": 0.655, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.519233311302049, | |
| "grad_norm": 0.04248838871717453, | |
| "learning_rate": 4.9558390578999016e-05, | |
| "loss": 0.0252, | |
| "mean_token_accuracy": 0.9932594150304794, | |
| "num_tokens": 280068096.0, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 4.545670852610707, | |
| "grad_norm": 0.04879326745867729, | |
| "learning_rate": 4.683240649874604e-05, | |
| "loss": 0.0256, | |
| "mean_token_accuracy": 0.9932215362787247, | |
| "num_tokens": 281706496.0, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 4.572108393919366, | |
| "grad_norm": 0.04033661261200905, | |
| "learning_rate": 4.4106422418493076e-05, | |
| "loss": 0.0256, | |
| "mean_token_accuracy": 0.993228867650032, | |
| "num_tokens": 283344896.0, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 4.598545935228024, | |
| "grad_norm": 0.043814148753881454, | |
| "learning_rate": 4.138043833824011e-05, | |
| "loss": 0.0251, | |
| "mean_token_accuracy": 0.9933846592903137, | |
| "num_tokens": 284983296.0, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 4.624983476536682, | |
| "grad_norm": 0.04637971520423889, | |
| "learning_rate": 3.8654454257987135e-05, | |
| "loss": 0.0256, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 4.624983476536682, | |
| "eval_loss": 0.8386735916137695, | |
| "eval_mean_token_accuracy": 0.8830419348485878, | |
| "eval_num_tokens": 286621696.0, | |
| "eval_runtime": 1444.7416, | |
| "eval_samples_per_second": 5.236, | |
| "eval_steps_per_second": 0.655, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 4.65142101784534, | |
| "grad_norm": 0.04898098483681679, | |
| "learning_rate": 3.592847017773417e-05, | |
| "loss": 0.025, | |
| "mean_token_accuracy": 0.9933046251535416, | |
| "num_tokens": 288260096.0, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 4.677858559153998, | |
| "grad_norm": 0.04658009856939316, | |
| "learning_rate": 3.3202486097481194e-05, | |
| "loss": 0.025, | |
| "mean_token_accuracy": 0.9934017658233643, | |
| "num_tokens": 289898496.0, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 4.704296100462657, | |
| "grad_norm": 0.04260571300983429, | |
| "learning_rate": 3.0476502017228217e-05, | |
| "loss": 0.0246, | |
| "mean_token_accuracy": 0.9935636669397354, | |
| "num_tokens": 291536896.0, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 4.730733641771315, | |
| "grad_norm": 0.04221141338348389, | |
| "learning_rate": 2.775051793697525e-05, | |
| "loss": 0.0249, | |
| "mean_token_accuracy": 0.9934304803609848, | |
| "num_tokens": 293175296.0, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 4.757171183079974, | |
| "grad_norm": 0.046503596007823944, | |
| "learning_rate": 2.502453385672228e-05, | |
| "loss": 0.0247, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.757171183079974, | |
| "eval_loss": 0.8432579636573792, | |
| "eval_mean_token_accuracy": 0.8832672491008081, | |
| "eval_num_tokens": 294813696.0, | |
| "eval_runtime": 1444.873, | |
| "eval_samples_per_second": 5.236, | |
| "eval_steps_per_second": 0.655, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.783608724388632, | |
| "grad_norm": 0.049112509936094284, | |
| "learning_rate": 2.2298549776469306e-05, | |
| "loss": 0.0244, | |
| "mean_token_accuracy": 0.9935187622904778, | |
| "num_tokens": 296452096.0, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 4.81004626569729, | |
| "grad_norm": 0.049860984086990356, | |
| "learning_rate": 1.9572565696216336e-05, | |
| "loss": 0.0248, | |
| "mean_token_accuracy": 0.9935349524021149, | |
| "num_tokens": 298090496.0, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 4.836483807005949, | |
| "grad_norm": 0.041883405297994614, | |
| "learning_rate": 1.6846581615963362e-05, | |
| "loss": 0.0243, | |
| "mean_token_accuracy": 0.9936204850673676, | |
| "num_tokens": 299728896.0, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 4.8629213483146065, | |
| "grad_norm": 0.043799083679914474, | |
| "learning_rate": 1.4120597535710392e-05, | |
| "loss": 0.024, | |
| "mean_token_accuracy": 0.9937127381563187, | |
| "num_tokens": 301367296.0, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 4.889358889623265, | |
| "grad_norm": 0.04291240870952606, | |
| "learning_rate": 1.139461345545742e-05, | |
| "loss": 0.0238, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 4.889358889623265, | |
| "eval_loss": 0.8444077372550964, | |
| "eval_mean_token_accuracy": 0.88355980766248, | |
| "eval_num_tokens": 303005696.0, | |
| "eval_runtime": 1444.2008, | |
| "eval_samples_per_second": 5.238, | |
| "eval_steps_per_second": 0.655, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 4.915796430931923, | |
| "grad_norm": 0.04179864004254341, | |
| "learning_rate": 8.668629375204448e-06, | |
| "loss": 0.0237, | |
| "mean_token_accuracy": 0.9937710836529732, | |
| "num_tokens": 304644096.0, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 4.942233972240581, | |
| "grad_norm": 0.04021856561303139, | |
| "learning_rate": 5.942645294951477e-06, | |
| "loss": 0.0239, | |
| "mean_token_accuracy": 0.993884414434433, | |
| "num_tokens": 306282496.0, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 4.96867151354924, | |
| "grad_norm": 0.04214942827820778, | |
| "learning_rate": 3.2166612146985063e-06, | |
| "loss": 0.0243, | |
| "mean_token_accuracy": 0.9936669170856476, | |
| "num_tokens": 307920896.0, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 4.995109054857898, | |
| "grad_norm": 0.04387414827942848, | |
| "learning_rate": 4.906771344455349e-07, | |
| "loss": 0.0242, | |
| "mean_token_accuracy": 0.9936302602291107, | |
| "num_tokens": 309559296.0, | |
| "step": 9450 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 9455, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 464304822288384.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |