| { | |
| "best_metric": 1.3278324604034424, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-400", | |
| "epoch": 0.026940108770689163, | |
| "eval_steps": 50, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 6.73502719267229e-05, | |
| "grad_norm": 2.030458927154541, | |
| "learning_rate": 1e-05, | |
| "loss": 2.2072, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 6.73502719267229e-05, | |
| "eval_loss": 3.1524617671966553, | |
| "eval_runtime": 1690.2881, | |
| "eval_samples_per_second": 14.795, | |
| "eval_steps_per_second": 3.699, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0001347005438534458, | |
| "grad_norm": 2.239384412765503, | |
| "learning_rate": 2e-05, | |
| "loss": 2.3997, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0002020508157801687, | |
| "grad_norm": 2.486010789871216, | |
| "learning_rate": 3e-05, | |
| "loss": 2.4882, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0002694010877068916, | |
| "grad_norm": 2.396327495574951, | |
| "learning_rate": 4e-05, | |
| "loss": 2.4251, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00033675135963361454, | |
| "grad_norm": 2.505659580230713, | |
| "learning_rate": 5e-05, | |
| "loss": 2.3752, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0004041016315603374, | |
| "grad_norm": 2.2823729515075684, | |
| "learning_rate": 6e-05, | |
| "loss": 2.3087, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00047145190348706035, | |
| "grad_norm": 2.2923173904418945, | |
| "learning_rate": 7e-05, | |
| "loss": 2.0596, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0005388021754137832, | |
| "grad_norm": 2.598090410232544, | |
| "learning_rate": 8e-05, | |
| "loss": 1.9093, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0006061524473405061, | |
| "grad_norm": 2.5075840950012207, | |
| "learning_rate": 9e-05, | |
| "loss": 1.5863, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0006735027192672291, | |
| "grad_norm": 1.9895573854446411, | |
| "learning_rate": 0.0001, | |
| "loss": 1.7257, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.000740852991193952, | |
| "grad_norm": 2.011659622192383, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 1.6529, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0008082032631206748, | |
| "grad_norm": 2.3391153812408447, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 1.5113, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0008755535350473977, | |
| "grad_norm": 1.9589104652404785, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 1.5592, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0009429038069741207, | |
| "grad_norm": 1.6105904579162598, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 1.6131, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0010102540789008435, | |
| "grad_norm": 1.6169345378875732, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 1.645, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0010776043508275665, | |
| "grad_norm": 1.6247849464416504, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 1.5378, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0011449546227542894, | |
| "grad_norm": 1.5837732553482056, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 1.6621, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0012123048946810122, | |
| "grad_norm": 1.5088438987731934, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 1.4937, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0012796551666077352, | |
| "grad_norm": 1.3942575454711914, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 1.4834, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0013470054385344582, | |
| "grad_norm": 1.6120097637176514, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 1.5793, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.001414355710461181, | |
| "grad_norm": 1.7541120052337646, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 1.4375, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.001481705982387904, | |
| "grad_norm": 1.7503563165664673, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 1.6468, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0015490562543146267, | |
| "grad_norm": 1.4936574697494507, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.3518, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0016164065262413497, | |
| "grad_norm": 1.5292739868164062, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 1.4441, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.0016837567981680727, | |
| "grad_norm": 1.5529022216796875, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 1.4522, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0017511070700947954, | |
| "grad_norm": 1.448577880859375, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 1.3866, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.0018184573420215184, | |
| "grad_norm": 1.4345710277557373, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 1.4611, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0018858076139482414, | |
| "grad_norm": 1.4968644380569458, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 1.485, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.001953157885874964, | |
| "grad_norm": 1.4573954343795776, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 1.7, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.002020508157801687, | |
| "grad_norm": 1.665392279624939, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 1.6596, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.00208785842972841, | |
| "grad_norm": 1.3012564182281494, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 1.2388, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.002155208701655133, | |
| "grad_norm": 1.714619755744934, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 1.5723, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0022225589735818557, | |
| "grad_norm": 1.408392310142517, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 1.5278, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.002289909245508579, | |
| "grad_norm": 1.3923598527908325, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 1.4961, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0023572595174353016, | |
| "grad_norm": 1.5535411834716797, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 1.5434, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0024246097893620244, | |
| "grad_norm": 1.6934750080108643, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.5558, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0024919600612887476, | |
| "grad_norm": 1.5046813488006592, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 1.3927, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0025593103332154704, | |
| "grad_norm": 1.781728744506836, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 1.6191, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.002626660605142193, | |
| "grad_norm": 1.8981614112854004, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 1.5438, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0026940108770689163, | |
| "grad_norm": 1.7245213985443115, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 1.7798, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.002761361148995639, | |
| "grad_norm": 1.7648022174835205, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 1.6861, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.002828711420922362, | |
| "grad_norm": 1.7356939315795898, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 1.736, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.002896061692849085, | |
| "grad_norm": 1.753619909286499, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 1.6496, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.002963411964775808, | |
| "grad_norm": 1.9407163858413696, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 1.7003, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0030307622367025306, | |
| "grad_norm": 1.6801741123199463, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 1.668, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0030981125086292534, | |
| "grad_norm": 1.9571483135223389, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 1.6165, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0031654627805559766, | |
| "grad_norm": 2.2739717960357666, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 1.6873, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.0032328130524826994, | |
| "grad_norm": 2.0475990772247314, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 1.5453, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.003300163324409422, | |
| "grad_norm": 2.1875641345977783, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.5151, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.0033675135963361453, | |
| "grad_norm": 3.4982423782348633, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 1.5134, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0033675135963361453, | |
| "eval_loss": 1.4206453561782837, | |
| "eval_runtime": 1693.3046, | |
| "eval_samples_per_second": 14.768, | |
| "eval_steps_per_second": 3.692, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.003434863868262868, | |
| "grad_norm": 2.413867950439453, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 1.3517, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.003502214140189591, | |
| "grad_norm": 1.2003623247146606, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 1.109, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.003569564412116314, | |
| "grad_norm": 1.154285192489624, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 1.1729, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.003636914684043037, | |
| "grad_norm": 1.2117410898208618, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 1.2341, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0037042649559697596, | |
| "grad_norm": 1.291092872619629, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 1.213, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.003771615227896483, | |
| "grad_norm": 1.0849498510360718, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 1.0593, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.0038389654998232056, | |
| "grad_norm": 1.084412693977356, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 1.1513, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.003906315771749928, | |
| "grad_norm": 1.2769420146942139, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 1.317, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.0039736660436766515, | |
| "grad_norm": 1.097766399383545, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 1.2195, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.004041016315603374, | |
| "grad_norm": 1.1258280277252197, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 1.2284, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.004108366587530097, | |
| "grad_norm": 1.2113773822784424, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 1.3336, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.00417571685945682, | |
| "grad_norm": 1.1779729127883911, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.2489, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.004243067131383543, | |
| "grad_norm": 1.2701348066329956, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 1.496, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.004310417403310266, | |
| "grad_norm": 1.2010540962219238, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 1.2049, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.004377767675236989, | |
| "grad_norm": 1.1773868799209595, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 1.2613, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.004445117947163711, | |
| "grad_norm": 1.1867501735687256, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 1.5148, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0045124682190904345, | |
| "grad_norm": 1.1402329206466675, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 1.1267, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.004579818491017158, | |
| "grad_norm": 1.2517718076705933, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 1.4253, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.00464716876294388, | |
| "grad_norm": 1.2350069284439087, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 1.3885, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.004714519034870603, | |
| "grad_norm": 1.1498900651931763, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 1.2204, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0047818693067973265, | |
| "grad_norm": 1.2308297157287598, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 1.519, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.004849219578724049, | |
| "grad_norm": 1.3052582740783691, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 1.3961, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.004916569850650772, | |
| "grad_norm": 1.3931699991226196, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 1.3947, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.004983920122577495, | |
| "grad_norm": 1.1981548070907593, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 1.1178, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.0050512703945042176, | |
| "grad_norm": 1.209280252456665, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 1.5509, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.005118620666430941, | |
| "grad_norm": 1.3320664167404175, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 1.3596, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.005185970938357664, | |
| "grad_norm": 1.3996878862380981, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 1.4188, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.005253321210284386, | |
| "grad_norm": 1.3204342126846313, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 1.3423, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0053206714822111095, | |
| "grad_norm": 1.2412264347076416, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 1.4367, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.005388021754137833, | |
| "grad_norm": 1.2513152360916138, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 1.3262, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.005455372026064555, | |
| "grad_norm": 1.1856772899627686, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 1.3741, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.005522722297991278, | |
| "grad_norm": 1.1955853700637817, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 1.4515, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.005590072569918001, | |
| "grad_norm": 1.2296124696731567, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 1.3032, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.005657422841844724, | |
| "grad_norm": 1.264333724975586, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 1.2194, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.005724773113771447, | |
| "grad_norm": 1.3912543058395386, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 1.4172, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.00579212338569817, | |
| "grad_norm": 1.4577277898788452, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 1.4781, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.0058594736576248925, | |
| "grad_norm": 1.327652096748352, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 1.3372, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.005926823929551616, | |
| "grad_norm": 1.5228052139282227, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.5345, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.005994174201478338, | |
| "grad_norm": 1.5462514162063599, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 1.8167, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.006061524473405061, | |
| "grad_norm": 1.3531618118286133, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 1.3477, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0061288747453317844, | |
| "grad_norm": 1.3700722455978394, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 1.3551, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.006196225017258507, | |
| "grad_norm": 1.5112624168395996, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 1.6612, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.00626357528918523, | |
| "grad_norm": 1.519962191581726, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 1.5554, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.006330925561111953, | |
| "grad_norm": 1.3995540142059326, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 1.2633, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0063982758330386755, | |
| "grad_norm": 1.8468939065933228, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 1.7552, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.006465626104965399, | |
| "grad_norm": 1.6963553428649902, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 1.5469, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.006532976376892122, | |
| "grad_norm": 1.7640057802200317, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 1.5224, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.006600326648818844, | |
| "grad_norm": 2.2663934230804443, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 1.6114, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.0066676769207455674, | |
| "grad_norm": 1.793444275856018, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 1.603, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.006735027192672291, | |
| "grad_norm": 2.0338592529296875, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 1.4991, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.006735027192672291, | |
| "eval_loss": 1.3707906007766724, | |
| "eval_runtime": 738.2821, | |
| "eval_samples_per_second": 33.872, | |
| "eval_steps_per_second": 8.468, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.006802377464599013, | |
| "grad_norm": 1.0350993871688843, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 1.0369, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.006869727736525736, | |
| "grad_norm": 1.2037113904953003, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 1.0181, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.006937078008452459, | |
| "grad_norm": 1.2135459184646606, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 1.2639, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.007004428280379182, | |
| "grad_norm": 1.301133155822754, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 0.9774, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.007071778552305905, | |
| "grad_norm": 1.127380132675171, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 1.1987, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.007139128824232628, | |
| "grad_norm": 1.1390761137008667, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 1.2403, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.0072064790961593505, | |
| "grad_norm": 1.0206562280654907, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 1.2218, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.007273829368086074, | |
| "grad_norm": 1.1791778802871704, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 1.195, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.007341179640012797, | |
| "grad_norm": 1.1491755247116089, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 1.2749, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.007408529911939519, | |
| "grad_norm": 1.105694055557251, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 1.3577, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.007475880183866242, | |
| "grad_norm": 1.261724591255188, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 1.3621, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.007543230455792966, | |
| "grad_norm": 1.0862749814987183, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 1.143, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.007610580727719688, | |
| "grad_norm": 0.9798551201820374, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 1.0633, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.007677930999646411, | |
| "grad_norm": 1.2961410284042358, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.3191, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.007745281271573134, | |
| "grad_norm": 1.138549566268921, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 1.3754, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.007812631543499857, | |
| "grad_norm": 1.0177663564682007, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 1.2438, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.00787998181542658, | |
| "grad_norm": 1.2038869857788086, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 1.195, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.007947332087353303, | |
| "grad_norm": 1.0598257780075073, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 1.2098, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.008014682359280026, | |
| "grad_norm": 1.2124276161193848, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 1.3533, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.008082032631206748, | |
| "grad_norm": 1.3013999462127686, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 1.3665, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.008149382903133471, | |
| "grad_norm": 2.011751651763916, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 1.3127, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.008216733175060194, | |
| "grad_norm": 1.0220164060592651, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 1.1425, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.008284083446986917, | |
| "grad_norm": 1.2821173667907715, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 1.4458, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.00835143371891364, | |
| "grad_norm": 1.3284802436828613, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 1.5041, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.008418783990840364, | |
| "grad_norm": 1.151969313621521, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 1.338, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.008486134262767085, | |
| "grad_norm": 1.336993932723999, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 1.2922, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.008553484534693808, | |
| "grad_norm": 1.0862728357315063, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.3108, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.008620834806620532, | |
| "grad_norm": 1.2815523147583008, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 1.3628, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.008688185078547255, | |
| "grad_norm": 1.260210394859314, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 1.3585, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.008755535350473978, | |
| "grad_norm": 1.4499257802963257, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 1.3755, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.008822885622400701, | |
| "grad_norm": 1.2586874961853027, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 1.3574, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.008890235894327423, | |
| "grad_norm": 1.2724554538726807, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 1.4911, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.008957586166254146, | |
| "grad_norm": 1.361464500427246, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 1.5767, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.009024936438180869, | |
| "grad_norm": 1.5993369817733765, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 1.4852, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.009092286710107592, | |
| "grad_norm": 1.4950506687164307, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 1.4919, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.009159636982034315, | |
| "grad_norm": 1.5922907590866089, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 1.6347, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.009226987253961039, | |
| "grad_norm": 1.2615551948547363, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 1.1959, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.00929433752588776, | |
| "grad_norm": 1.3236908912658691, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 1.4719, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.009361687797814483, | |
| "grad_norm": 1.2798570394515991, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 1.309, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.009429038069741207, | |
| "grad_norm": 1.588212013244629, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.5616, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.00949638834166793, | |
| "grad_norm": 1.4299564361572266, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 1.5298, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.009563738613594653, | |
| "grad_norm": 1.6370961666107178, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 1.6629, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.009631088885521374, | |
| "grad_norm": 1.483093023300171, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 1.4603, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.009698439157448098, | |
| "grad_norm": 1.846943736076355, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 1.6674, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.00976578942937482, | |
| "grad_norm": 1.6040513515472412, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 1.6899, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.009833139701301544, | |
| "grad_norm": 1.330278992652893, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 1.3632, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.009900489973228267, | |
| "grad_norm": 1.5452008247375488, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 1.6167, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.00996784024515499, | |
| "grad_norm": 1.6564722061157227, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 1.6265, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.010035190517081712, | |
| "grad_norm": 1.940214991569519, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 1.599, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.010102540789008435, | |
| "grad_norm": 2.222428321838379, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 1.6995, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.010102540789008435, | |
| "eval_loss": 1.3560807704925537, | |
| "eval_runtime": 738.5582, | |
| "eval_samples_per_second": 33.859, | |
| "eval_steps_per_second": 8.465, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.010169891060935158, | |
| "grad_norm": 0.9370902180671692, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 1.1064, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.010237241332861882, | |
| "grad_norm": 0.9733120799064636, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 1.1346, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.010304591604788605, | |
| "grad_norm": 0.9286036491394043, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 1.0471, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.010371941876715328, | |
| "grad_norm": 1.061566948890686, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 1.1593, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.01043929214864205, | |
| "grad_norm": 0.988382875919342, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 1.2365, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.010506642420568773, | |
| "grad_norm": 0.9433530569076538, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 1.1843, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.010573992692495496, | |
| "grad_norm": 0.9518124461174011, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 1.1346, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.010641342964422219, | |
| "grad_norm": 1.0984117984771729, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 1.3145, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.010708693236348942, | |
| "grad_norm": 1.0021629333496094, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 1.1335, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.010776043508275665, | |
| "grad_norm": 1.0588817596435547, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 1.2448, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.010843393780202387, | |
| "grad_norm": 1.0959153175354004, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 1.148, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.01091074405212911, | |
| "grad_norm": 0.9979251623153687, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 1.0908, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.010978094324055833, | |
| "grad_norm": 1.0619982481002808, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 1.2502, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.011045444595982556, | |
| "grad_norm": 1.091411828994751, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 1.1467, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.01111279486790928, | |
| "grad_norm": 1.1791459321975708, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 1.3743, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.011180145139836003, | |
| "grad_norm": 1.069138765335083, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 1.2523, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.011247495411762724, | |
| "grad_norm": 1.167673945426941, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 1.4057, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.011314845683689448, | |
| "grad_norm": 1.1494766473770142, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 1.2518, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.01138219595561617, | |
| "grad_norm": 1.156091570854187, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 1.3678, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.011449546227542894, | |
| "grad_norm": 1.1573467254638672, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 1.4178, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.011516896499469617, | |
| "grad_norm": 1.1399964094161987, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 1.2234, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.01158424677139634, | |
| "grad_norm": 1.1700021028518677, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 1.3678, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.011651597043323062, | |
| "grad_norm": 1.1301724910736084, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 1.3965, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.011718947315249785, | |
| "grad_norm": 1.0706028938293457, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 1.3663, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.011786297587176508, | |
| "grad_norm": 1.0472501516342163, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 1.2252, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.011853647859103231, | |
| "grad_norm": 1.1286253929138184, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 1.3053, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.011920998131029955, | |
| "grad_norm": 1.138118863105774, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 1.34, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.011988348402956676, | |
| "grad_norm": 1.0691560506820679, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 1.3048, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.0120556986748834, | |
| "grad_norm": 1.226596713066101, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 1.3146, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.012123048946810122, | |
| "grad_norm": 1.2226495742797852, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 1.4645, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.012190399218736846, | |
| "grad_norm": 1.215620994567871, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 1.299, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.012257749490663569, | |
| "grad_norm": 1.2768441438674927, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 1.5455, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.012325099762590292, | |
| "grad_norm": 1.2758901119232178, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 1.4963, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.012392450034517014, | |
| "grad_norm": 1.3229836225509644, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 1.435, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.012459800306443737, | |
| "grad_norm": 1.2354806661605835, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 1.4156, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.01252715057837046, | |
| "grad_norm": 1.3381010293960571, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 1.4482, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.012594500850297183, | |
| "grad_norm": 1.2151193618774414, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 1.3234, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.012661851122223906, | |
| "grad_norm": 1.646081566810608, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 1.4757, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.01272920139415063, | |
| "grad_norm": 1.3965771198272705, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 1.5457, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.012796551666077351, | |
| "grad_norm": 1.2820106744766235, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 1.4958, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.012863901938004074, | |
| "grad_norm": 1.421890139579773, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 1.5091, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.012931252209930797, | |
| "grad_norm": 1.5250858068466187, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 1.5578, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.01299860248185752, | |
| "grad_norm": 1.6469945907592773, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 1.5527, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.013065952753784244, | |
| "grad_norm": 1.660823106765747, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 1.4799, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.013133303025710967, | |
| "grad_norm": 1.4221848249435425, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 1.5711, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.013200653297637688, | |
| "grad_norm": 1.3503718376159668, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 1.4474, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.013268003569564412, | |
| "grad_norm": 1.51528000831604, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 1.5336, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.013335353841491135, | |
| "grad_norm": 1.7951074838638306, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 1.2894, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.013402704113417858, | |
| "grad_norm": 1.6057062149047852, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 1.7802, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.013470054385344581, | |
| "grad_norm": 1.7107088565826416, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 1.2964, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.013470054385344581, | |
| "eval_loss": 1.3432029485702515, | |
| "eval_runtime": 738.1699, | |
| "eval_samples_per_second": 33.877, | |
| "eval_steps_per_second": 8.47, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.013537404657271305, | |
| "grad_norm": 1.1227874755859375, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 1.1946, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.013604754929198026, | |
| "grad_norm": 1.0087556838989258, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 1.2882, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.01367210520112475, | |
| "grad_norm": 1.0933849811553955, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 1.3311, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.013739455473051472, | |
| "grad_norm": 0.9209799766540527, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 1.1001, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.013806805744978196, | |
| "grad_norm": 0.9551730155944824, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1042, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.013874156016904919, | |
| "grad_norm": 0.8248175382614136, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 1.0215, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.013941506288831642, | |
| "grad_norm": 0.9202468395233154, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 1.0346, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.014008856560758363, | |
| "grad_norm": 0.9693543910980225, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 1.1723, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.014076206832685087, | |
| "grad_norm": 0.8910868763923645, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 1.0814, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.01414355710461181, | |
| "grad_norm": 1.1232092380523682, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 1.311, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.014210907376538533, | |
| "grad_norm": 1.0476187467575073, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 1.3065, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.014278257648465256, | |
| "grad_norm": 0.9502547979354858, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 1.0397, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.01434560792039198, | |
| "grad_norm": 1.0085866451263428, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 1.2223, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.014412958192318701, | |
| "grad_norm": 0.9929443001747131, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 1.0958, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.014480308464245424, | |
| "grad_norm": 1.5844645500183105, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 1.262, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.014547658736172147, | |
| "grad_norm": 1.1886199712753296, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 1.4208, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.01461500900809887, | |
| "grad_norm": 0.991338312625885, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 1.2287, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.014682359280025594, | |
| "grad_norm": 1.0494619607925415, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 1.2124, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.014749709551952315, | |
| "grad_norm": 1.0562416315078735, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 1.2285, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.014817059823879038, | |
| "grad_norm": 1.1794100999832153, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 1.2839, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.014884410095805762, | |
| "grad_norm": 1.2214128971099854, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 1.4051, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.014951760367732485, | |
| "grad_norm": 1.0522944927215576, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 1.135, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.015019110639659208, | |
| "grad_norm": 1.0426898002624512, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 1.2934, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.015086460911585931, | |
| "grad_norm": 1.1241427659988403, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 1.237, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.015153811183512653, | |
| "grad_norm": 1.202584147453308, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 1.1759, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.015221161455439376, | |
| "grad_norm": 1.0882707834243774, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 1.3437, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.015288511727366099, | |
| "grad_norm": 1.1346945762634277, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 1.2853, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.015355861999292822, | |
| "grad_norm": 1.1027241945266724, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 1.1455, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.015423212271219545, | |
| "grad_norm": 1.2200098037719727, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 1.4002, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.015490562543146269, | |
| "grad_norm": 1.1374239921569824, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 1.4173, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.01555791281507299, | |
| "grad_norm": 1.2057877779006958, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 1.5388, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.015625263086999713, | |
| "grad_norm": 1.1581593751907349, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 1.452, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.01569261335892644, | |
| "grad_norm": 1.4164345264434814, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 1.5977, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.01575996363085316, | |
| "grad_norm": 1.3473681211471558, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 1.6104, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.01582731390277988, | |
| "grad_norm": 1.365702748298645, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 1.455, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.015894664174706606, | |
| "grad_norm": 1.410869836807251, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 1.5434, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.015962014446633328, | |
| "grad_norm": 1.152729868888855, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 1.3922, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.016029364718560053, | |
| "grad_norm": 1.2944568395614624, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 1.328, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.016096714990486774, | |
| "grad_norm": 1.3709889650344849, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 1.3385, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.016164065262413495, | |
| "grad_norm": 1.472702980041504, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 1.7066, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.01623141553434022, | |
| "grad_norm": 1.4194689989089966, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 1.5297, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.016298765806266942, | |
| "grad_norm": 1.4132965803146362, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 1.4771, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.016366116078193667, | |
| "grad_norm": 1.3434481620788574, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 1.3609, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.01643346635012039, | |
| "grad_norm": 1.3477119207382202, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.4376, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.016500816622047113, | |
| "grad_norm": 1.482155442237854, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 1.5748, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.016568166893973835, | |
| "grad_norm": 1.5454543828964233, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 1.3573, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.016635517165900556, | |
| "grad_norm": 1.6062289476394653, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 1.7007, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.01670286743782728, | |
| "grad_norm": 1.4669798612594604, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 1.5439, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.016770217709754003, | |
| "grad_norm": 1.748022198677063, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 1.562, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.016837567981680727, | |
| "grad_norm": 2.035632848739624, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 1.6668, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.016837567981680727, | |
| "eval_loss": 1.3342212438583374, | |
| "eval_runtime": 739.2168, | |
| "eval_samples_per_second": 33.829, | |
| "eval_steps_per_second": 8.458, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.01690491825360745, | |
| "grad_norm": 0.8713572025299072, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 0.9818, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.01697226852553417, | |
| "grad_norm": 0.9513303637504578, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 1.1104, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.017039618797460895, | |
| "grad_norm": 1.0168979167938232, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 1.1361, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.017106969069387617, | |
| "grad_norm": 1.0765156745910645, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 1.2839, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.017174319341314342, | |
| "grad_norm": 0.9411038160324097, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 1.1857, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.017241669613241063, | |
| "grad_norm": 1.0114924907684326, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 1.1497, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.017309019885167785, | |
| "grad_norm": 0.9451395869255066, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 1.1326, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.01737637015709451, | |
| "grad_norm": 1.0250037908554077, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 1.2372, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.01744372042902123, | |
| "grad_norm": 0.9830464124679565, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 1.1286, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.017511070700947956, | |
| "grad_norm": 1.1806437969207764, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 1.1044, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.017578420972874678, | |
| "grad_norm": 0.910880446434021, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 1.0903, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.017645771244801402, | |
| "grad_norm": 1.0307596921920776, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 1.2891, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.017713121516728124, | |
| "grad_norm": 1.1910873651504517, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 1.3077, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.017780471788654845, | |
| "grad_norm": 1.0718063116073608, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 1.3975, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.01784782206058157, | |
| "grad_norm": 0.9779868721961975, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 1.3028, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.017915172332508292, | |
| "grad_norm": 1.050061821937561, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 1.2114, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.017982522604435017, | |
| "grad_norm": 1.144936203956604, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 1.4258, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.018049872876361738, | |
| "grad_norm": 1.0306668281555176, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 1.2216, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.01811722314828846, | |
| "grad_norm": 1.282397985458374, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 1.2653, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.018184573420215185, | |
| "grad_norm": 1.2951265573501587, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.5976, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.018251923692141906, | |
| "grad_norm": 1.2171090841293335, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 1.39, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.01831927396406863, | |
| "grad_norm": 1.021344542503357, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 1.178, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.018386624235995352, | |
| "grad_norm": 1.0217483043670654, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 1.2201, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.018453974507922077, | |
| "grad_norm": 1.1176892518997192, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 1.234, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.0185213247798488, | |
| "grad_norm": 1.0938963890075684, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 1.359, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.01858867505177552, | |
| "grad_norm": 1.1674708127975464, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 1.4127, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.018656025323702245, | |
| "grad_norm": 1.1793794631958008, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 1.2394, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.018723375595628967, | |
| "grad_norm": 1.1035133600234985, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 1.2598, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.01879072586755569, | |
| "grad_norm": 1.2473664283752441, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 1.3994, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.018858076139482413, | |
| "grad_norm": 1.1195029020309448, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 1.2052, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.018925426411409135, | |
| "grad_norm": 1.2758724689483643, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 1.5162, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.01899277668333586, | |
| "grad_norm": 1.1609270572662354, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 1.3269, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.01906012695526258, | |
| "grad_norm": 1.2329775094985962, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.4856, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.019127477227189306, | |
| "grad_norm": 1.1754076480865479, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 1.3518, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.019194827499116027, | |
| "grad_norm": 1.2435510158538818, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 1.3986, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.01926217777104275, | |
| "grad_norm": 1.1661019325256348, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 1.3242, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.019329528042969474, | |
| "grad_norm": 1.2058144807815552, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 1.3828, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.019396878314896195, | |
| "grad_norm": 1.216773271560669, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 1.3686, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.01946422858682292, | |
| "grad_norm": 1.413074254989624, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 1.4573, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.01953157885874964, | |
| "grad_norm": 1.2821346521377563, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 1.3789, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.019598929130676367, | |
| "grad_norm": 1.2425432205200195, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 1.5083, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.019666279402603088, | |
| "grad_norm": 1.3399276733398438, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 1.4766, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.01973362967452981, | |
| "grad_norm": 1.17010498046875, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 1.2784, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.019800979946456534, | |
| "grad_norm": 1.728461503982544, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 1.3018, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.019868330218383256, | |
| "grad_norm": 1.5740691423416138, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 1.4088, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.01993568049030998, | |
| "grad_norm": 1.5061848163604736, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 1.6017, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.020003030762236702, | |
| "grad_norm": 1.895371437072754, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 1.5244, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.020070381034163424, | |
| "grad_norm": 1.5589200258255005, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 1.742, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.02013773130609015, | |
| "grad_norm": 1.8027925491333008, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 1.7559, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.02020508157801687, | |
| "grad_norm": 1.7761001586914062, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 1.5844, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.02020508157801687, | |
| "eval_loss": 1.3293969631195068, | |
| "eval_runtime": 738.6384, | |
| "eval_samples_per_second": 33.856, | |
| "eval_steps_per_second": 8.464, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.020272431849943595, | |
| "grad_norm": 0.9483279585838318, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 1.1289, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.020339782121870317, | |
| "grad_norm": 0.880562961101532, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 1.0546, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.02040713239379704, | |
| "grad_norm": 0.9176260232925415, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 1.0747, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.020474482665723763, | |
| "grad_norm": 0.941221296787262, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 0.9572, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.020541832937650484, | |
| "grad_norm": 0.9974879622459412, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 1.1104, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.02060918320957721, | |
| "grad_norm": 1.0085439682006836, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 1.1802, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.02067653348150393, | |
| "grad_norm": 1.0595355033874512, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 1.1322, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.020743883753430656, | |
| "grad_norm": 1.025657057762146, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 1.1411, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.020811234025357377, | |
| "grad_norm": 1.0750242471694946, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 1.3605, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.0208785842972841, | |
| "grad_norm": 1.1114938259124756, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 1.2302, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.020945934569210824, | |
| "grad_norm": 0.9650045037269592, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 1.1415, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.021013284841137545, | |
| "grad_norm": 0.9643846154212952, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 1.1501, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.02108063511306427, | |
| "grad_norm": 0.9403222799301147, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 1.1286, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.02114798538499099, | |
| "grad_norm": 1.1337164640426636, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 1.2084, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.021215335656917717, | |
| "grad_norm": 0.9642013311386108, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 1.1753, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.021282685928844438, | |
| "grad_norm": 1.0559114217758179, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 1.2275, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.02135003620077116, | |
| "grad_norm": 1.1197682619094849, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 1.1159, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.021417386472697884, | |
| "grad_norm": 1.038716435432434, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 1.4075, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.021484736744624606, | |
| "grad_norm": 0.96999591588974, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 1.1907, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.02155208701655133, | |
| "grad_norm": 1.1059964895248413, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 1.2472, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.021619437288478052, | |
| "grad_norm": 1.0419682264328003, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 1.2452, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.021686787560404774, | |
| "grad_norm": 1.1106568574905396, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.1885, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.0217541378323315, | |
| "grad_norm": 1.0976035594940186, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 1.3643, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.02182148810425822, | |
| "grad_norm": 1.1701173782348633, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 1.3457, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.021888838376184945, | |
| "grad_norm": 1.2230879068374634, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 1.5155, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.021956188648111667, | |
| "grad_norm": 1.1395500898361206, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 1.3319, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.022023538920038388, | |
| "grad_norm": 1.1425197124481201, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 1.3877, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.022090889191965113, | |
| "grad_norm": 1.0961964130401611, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 1.2585, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.022158239463891834, | |
| "grad_norm": 1.2439264059066772, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 1.491, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.02222558973581856, | |
| "grad_norm": 1.177886962890625, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 1.415, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.02229294000774528, | |
| "grad_norm": 1.4000835418701172, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 1.4878, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.022360290279672006, | |
| "grad_norm": 1.2325903177261353, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 1.2558, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.022427640551598727, | |
| "grad_norm": 1.2314106225967407, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 1.3935, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.02249499082352545, | |
| "grad_norm": 1.2869185209274292, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 1.51, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.022562341095452174, | |
| "grad_norm": 1.3866029977798462, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 1.4728, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.022629691367378895, | |
| "grad_norm": 1.3128485679626465, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 1.3169, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.02269704163930562, | |
| "grad_norm": 1.304387092590332, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 1.4541, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.02276439191123234, | |
| "grad_norm": 1.4286638498306274, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 1.4538, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.022831742183159063, | |
| "grad_norm": 1.4982728958129883, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 1.5706, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.022899092455085788, | |
| "grad_norm": 1.199910283088684, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 1.4591, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.02296644272701251, | |
| "grad_norm": 1.2605010271072388, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 1.3878, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.023033792998939234, | |
| "grad_norm": 1.4372059106826782, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 1.5722, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.023101143270865956, | |
| "grad_norm": 2.0687360763549805, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 1.3645, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.02316849354279268, | |
| "grad_norm": 1.444715976715088, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 1.3622, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.023235843814719402, | |
| "grad_norm": 1.7077946662902832, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 1.5487, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.023303194086646124, | |
| "grad_norm": 1.355229377746582, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 1.5458, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.02337054435857285, | |
| "grad_norm": 1.549574851989746, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 1.3594, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.02343789463049957, | |
| "grad_norm": 1.6200557947158813, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 1.6635, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.023505244902426295, | |
| "grad_norm": 1.8985865116119385, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 1.4596, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.023572595174353016, | |
| "grad_norm": 1.9179009199142456, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 1.5636, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.023572595174353016, | |
| "eval_loss": 1.3278803825378418, | |
| "eval_runtime": 738.5748, | |
| "eval_samples_per_second": 33.858, | |
| "eval_steps_per_second": 8.465, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.023639945446279738, | |
| "grad_norm": 1.4624953269958496, | |
| "learning_rate": 3.844650207332562e-06, | |
| "loss": 1.0904, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.023707295718206463, | |
| "grad_norm": 1.0814251899719238, | |
| "learning_rate": 3.691267552111183e-06, | |
| "loss": 1.0723, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.023774645990133184, | |
| "grad_norm": 0.9951095581054688, | |
| "learning_rate": 3.54088980417534e-06, | |
| "loss": 1.1652, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.02384199626205991, | |
| "grad_norm": 0.9580594897270203, | |
| "learning_rate": 3.393526721321616e-06, | |
| "loss": 1.1686, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.02390934653398663, | |
| "grad_norm": 1.0866096019744873, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 1.1205, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.023976696805913352, | |
| "grad_norm": 0.9641712307929993, | |
| "learning_rate": 3.1078826033397843e-06, | |
| "loss": 1.1141, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.024044047077840077, | |
| "grad_norm": 0.8843631744384766, | |
| "learning_rate": 2.9696201032491434e-06, | |
| "loss": 1.117, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.0241113973497668, | |
| "grad_norm": 0.8816883563995361, | |
| "learning_rate": 2.8344093371128424e-06, | |
| "loss": 1.0141, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.024178747621693523, | |
| "grad_norm": 0.9398783445358276, | |
| "learning_rate": 2.70225907856374e-06, | |
| "loss": 1.1264, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.024246097893620245, | |
| "grad_norm": 1.1492022275924683, | |
| "learning_rate": 2.573177902642726e-06, | |
| "loss": 1.3692, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.02431344816554697, | |
| "grad_norm": 0.9554076790809631, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 1.0897, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.02438079843747369, | |
| "grad_norm": 1.031767725944519, | |
| "learning_rate": 2.324256102563188e-06, | |
| "loss": 1.2614, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.024448148709400413, | |
| "grad_norm": 1.0227885246276855, | |
| "learning_rate": 2.204431630583548e-06, | |
| "loss": 1.1484, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.024515498981327138, | |
| "grad_norm": 0.943084716796875, | |
| "learning_rate": 2.087708544541689e-06, | |
| "loss": 1.1367, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.02458284925325386, | |
| "grad_norm": 0.9876199960708618, | |
| "learning_rate": 1.974094418431388e-06, | |
| "loss": 1.3024, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.024650199525180584, | |
| "grad_norm": 1.0107133388519287, | |
| "learning_rate": 1.8635966245104664e-06, | |
| "loss": 1.152, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.024717549797107306, | |
| "grad_norm": 0.987256646156311, | |
| "learning_rate": 1.7562223328224325e-06, | |
| "loss": 1.2102, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.024784900069034027, | |
| "grad_norm": 1.1262116432189941, | |
| "learning_rate": 1.6519785107311891e-06, | |
| "loss": 1.2629, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.024852250340960752, | |
| "grad_norm": 1.0786629915237427, | |
| "learning_rate": 1.5508719224689717e-06, | |
| "loss": 1.3168, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.024919600612887473, | |
| "grad_norm": 1.2885220050811768, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 1.3615, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.0249869508848142, | |
| "grad_norm": 1.0082440376281738, | |
| "learning_rate": 1.358096486081778e-06, | |
| "loss": 1.2357, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.02505430115674092, | |
| "grad_norm": 1.1223047971725464, | |
| "learning_rate": 1.2664401468786114e-06, | |
| "loss": 1.3303, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.025121651428667645, | |
| "grad_norm": 0.9703611731529236, | |
| "learning_rate": 1.1779460585363944e-06, | |
| "loss": 0.9298, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.025189001700594366, | |
| "grad_norm": 1.0855671167373657, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 1.1574, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.025256351972521088, | |
| "grad_norm": 1.0677353143692017, | |
| "learning_rate": 1.0104673978866164e-06, | |
| "loss": 1.3944, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.025323702244447813, | |
| "grad_norm": 1.1818300485610962, | |
| "learning_rate": 9.314936930293283e-07, | |
| "loss": 1.4814, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.025391052516374534, | |
| "grad_norm": 1.0969845056533813, | |
| "learning_rate": 8.557039732283944e-07, | |
| "loss": 1.3965, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.02545840278830126, | |
| "grad_norm": 1.1741873025894165, | |
| "learning_rate": 7.83103156370113e-07, | |
| "loss": 1.4252, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.02552575306022798, | |
| "grad_norm": 1.0685746669769287, | |
| "learning_rate": 7.136959534174592e-07, | |
| "loss": 1.2656, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.025593103332154702, | |
| "grad_norm": 1.052201747894287, | |
| "learning_rate": 6.474868681043578e-07, | |
| "loss": 1.2646, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.025660453604081427, | |
| "grad_norm": 1.2119255065917969, | |
| "learning_rate": 5.844801966434832e-07, | |
| "loss": 1.4586, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.02572780387600815, | |
| "grad_norm": 1.235493540763855, | |
| "learning_rate": 5.246800274474439e-07, | |
| "loss": 1.3549, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.025795154147934873, | |
| "grad_norm": 1.270772099494934, | |
| "learning_rate": 4.680902408635335e-07, | |
| "loss": 1.389, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.025862504419861595, | |
| "grad_norm": 1.2856876850128174, | |
| "learning_rate": 4.1471450892189846e-07, | |
| "loss": 1.4744, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.02592985469178832, | |
| "grad_norm": 1.3383358716964722, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 1.5902, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.02599720496371504, | |
| "grad_norm": 1.2960913181304932, | |
| "learning_rate": 3.1761885408435054e-07, | |
| "loss": 1.4461, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.026064555235641763, | |
| "grad_norm": 1.5527251958847046, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 1.5167, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.026131905507568488, | |
| "grad_norm": 1.2937921285629272, | |
| "learning_rate": 2.334182641175686e-07, | |
| "loss": 1.2337, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.02619925577949521, | |
| "grad_norm": 1.2841928005218506, | |
| "learning_rate": 1.9616057881935436e-07, | |
| "loss": 1.2654, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.026266606051421934, | |
| "grad_norm": 1.3135737180709839, | |
| "learning_rate": 1.6213459328950352e-07, | |
| "loss": 1.433, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.026333956323348656, | |
| "grad_norm": 1.3532600402832031, | |
| "learning_rate": 1.3134251542544774e-07, | |
| "loss": 1.3902, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.026401306595275377, | |
| "grad_norm": 1.3329143524169922, | |
| "learning_rate": 1.0378634328099269e-07, | |
| "loss": 1.4056, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.026468656867202102, | |
| "grad_norm": 1.5213643312454224, | |
| "learning_rate": 7.946786493666647e-08, | |
| "loss": 1.5343, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.026536007139128823, | |
| "grad_norm": 1.5116889476776123, | |
| "learning_rate": 5.838865838366792e-08, | |
| "loss": 1.6032, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.02660335741105555, | |
| "grad_norm": 1.2199372053146362, | |
| "learning_rate": 4.055009142152067e-08, | |
| "loss": 1.3873, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.02667070768298227, | |
| "grad_norm": 1.539748191833496, | |
| "learning_rate": 2.595332156925534e-08, | |
| "loss": 1.4327, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.02673805795490899, | |
| "grad_norm": 1.466637134552002, | |
| "learning_rate": 1.4599295990352924e-08, | |
| "loss": 1.5486, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.026805408226835716, | |
| "grad_norm": 1.5383957624435425, | |
| "learning_rate": 6.488751431266149e-09, | |
| "loss": 1.2895, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.026872758498762438, | |
| "grad_norm": 1.4582974910736084, | |
| "learning_rate": 1.622214173602199e-09, | |
| "loss": 1.3618, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.026940108770689163, | |
| "grad_norm": 1.8420403003692627, | |
| "learning_rate": 0.0, | |
| "loss": 1.5286, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.026940108770689163, | |
| "eval_loss": 1.3278324604034424, | |
| "eval_runtime": 738.8324, | |
| "eval_samples_per_second": 33.847, | |
| "eval_steps_per_second": 8.462, | |
| "step": 400 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0933414555287552e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |