| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.8850340738118417, | |
| "eval_steps": 500, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008850340738118417, | |
| "grad_norm": 1.517469048500061, | |
| "learning_rate": 1.9919999999999997e-06, | |
| "loss": 1.519, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.017700681476236834, | |
| "grad_norm": 1.6662976741790771, | |
| "learning_rate": 1.984e-06, | |
| "loss": 1.4776, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.026551022214355253, | |
| "grad_norm": 0.8021761775016785, | |
| "learning_rate": 1.976e-06, | |
| "loss": 1.4026, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03540136295247367, | |
| "grad_norm": 0.8772447109222412, | |
| "learning_rate": 1.968e-06, | |
| "loss": 1.3306, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.04425170369059209, | |
| "grad_norm": 0.4179963767528534, | |
| "learning_rate": 1.96e-06, | |
| "loss": 1.286, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.053102044428710506, | |
| "grad_norm": 0.4829399287700653, | |
| "learning_rate": 1.9519999999999997e-06, | |
| "loss": 1.2756, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.061952385166828924, | |
| "grad_norm": 0.5651261806488037, | |
| "learning_rate": 1.944e-06, | |
| "loss": 1.2721, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.07080272590494734, | |
| "grad_norm": 0.5160872936248779, | |
| "learning_rate": 1.9359999999999998e-06, | |
| "loss": 1.2579, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.07965306664306576, | |
| "grad_norm": 0.7501122355461121, | |
| "learning_rate": 1.928e-06, | |
| "loss": 1.2853, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.08850340738118417, | |
| "grad_norm": 0.591393232345581, | |
| "learning_rate": 1.92e-06, | |
| "loss": 1.2472, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.0973537481193026, | |
| "grad_norm": 0.5573005080223083, | |
| "learning_rate": 1.9119999999999997e-06, | |
| "loss": 1.2348, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.10620408885742101, | |
| "grad_norm": 0.607552707195282, | |
| "learning_rate": 1.904e-06, | |
| "loss": 1.2679, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.11505442959553942, | |
| "grad_norm": 0.6899151802062988, | |
| "learning_rate": 1.8959999999999997e-06, | |
| "loss": 1.2513, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.12390477033365785, | |
| "grad_norm": 0.5849478244781494, | |
| "learning_rate": 1.8879999999999998e-06, | |
| "loss": 1.226, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.13275511107177626, | |
| "grad_norm": 0.6027043461799622, | |
| "learning_rate": 1.8799999999999998e-06, | |
| "loss": 1.2163, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.14160545180989467, | |
| "grad_norm": 0.6785966157913208, | |
| "learning_rate": 1.872e-06, | |
| "loss": 1.2076, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.1504557925480131, | |
| "grad_norm": 0.5975379347801208, | |
| "learning_rate": 1.864e-06, | |
| "loss": 1.2146, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.15930613328613152, | |
| "grad_norm": 0.685453474521637, | |
| "learning_rate": 1.856e-06, | |
| "loss": 1.198, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.16815647402424994, | |
| "grad_norm": 1.1250847578048706, | |
| "learning_rate": 1.848e-06, | |
| "loss": 1.218, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.17700681476236835, | |
| "grad_norm": 0.6129019260406494, | |
| "learning_rate": 1.84e-06, | |
| "loss": 1.1894, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.18585715550048676, | |
| "grad_norm": 0.7388463616371155, | |
| "learning_rate": 1.832e-06, | |
| "loss": 1.1962, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.1947074962386052, | |
| "grad_norm": 0.5909475684165955, | |
| "learning_rate": 1.824e-06, | |
| "loss": 1.241, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2035578369767236, | |
| "grad_norm": 0.5264613032341003, | |
| "learning_rate": 1.816e-06, | |
| "loss": 1.1915, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.21240817771484202, | |
| "grad_norm": 0.6419306397438049, | |
| "learning_rate": 1.8079999999999999e-06, | |
| "loss": 1.1938, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.22125851845296043, | |
| "grad_norm": 0.6240788698196411, | |
| "learning_rate": 1.8e-06, | |
| "loss": 1.214, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.23010885919107885, | |
| "grad_norm": 0.7954173684120178, | |
| "learning_rate": 1.792e-06, | |
| "loss": 1.195, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.23895919992919729, | |
| "grad_norm": 0.7451553344726562, | |
| "learning_rate": 1.784e-06, | |
| "loss": 1.1793, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.2478095406673157, | |
| "grad_norm": 0.8471085429191589, | |
| "learning_rate": 1.776e-06, | |
| "loss": 1.1895, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.2566598814054341, | |
| "grad_norm": 1.0085513591766357, | |
| "learning_rate": 1.7679999999999998e-06, | |
| "loss": 1.2087, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.2655102221435525, | |
| "grad_norm": 0.7709102630615234, | |
| "learning_rate": 1.7599999999999999e-06, | |
| "loss": 1.1959, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.27436056288167093, | |
| "grad_norm": 0.7612162232398987, | |
| "learning_rate": 1.752e-06, | |
| "loss": 1.1803, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.28321090361978934, | |
| "grad_norm": 1.0669310092926025, | |
| "learning_rate": 1.744e-06, | |
| "loss": 1.1548, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.29206124435790776, | |
| "grad_norm": 0.7489383220672607, | |
| "learning_rate": 1.736e-06, | |
| "loss": 1.1654, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.3009115850960262, | |
| "grad_norm": 0.8011970520019531, | |
| "learning_rate": 1.7279999999999998e-06, | |
| "loss": 1.1789, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.30976192583414464, | |
| "grad_norm": 0.7374864816665649, | |
| "learning_rate": 1.7199999999999998e-06, | |
| "loss": 1.171, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.31861226657226305, | |
| "grad_norm": 0.8267546892166138, | |
| "learning_rate": 1.7119999999999999e-06, | |
| "loss": 1.1435, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.32746260731038146, | |
| "grad_norm": 0.8837570548057556, | |
| "learning_rate": 1.704e-06, | |
| "loss": 1.1674, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.33631294804849987, | |
| "grad_norm": 0.7216938138008118, | |
| "learning_rate": 1.696e-06, | |
| "loss": 1.1446, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.3451632887866183, | |
| "grad_norm": 0.8478527069091797, | |
| "learning_rate": 1.6879999999999998e-06, | |
| "loss": 1.1721, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.3540136295247367, | |
| "grad_norm": 0.8281861543655396, | |
| "learning_rate": 1.6799999999999998e-06, | |
| "loss": 1.1741, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.3628639702628551, | |
| "grad_norm": 0.6918274760246277, | |
| "learning_rate": 1.6719999999999998e-06, | |
| "loss": 1.1433, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.3717143110009735, | |
| "grad_norm": 0.6722283363342285, | |
| "learning_rate": 1.6639999999999999e-06, | |
| "loss": 1.189, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.38056465173909193, | |
| "grad_norm": 0.8337692022323608, | |
| "learning_rate": 1.656e-06, | |
| "loss": 1.1694, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.3894149924772104, | |
| "grad_norm": 0.7980256080627441, | |
| "learning_rate": 1.648e-06, | |
| "loss": 1.1794, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.3982653332153288, | |
| "grad_norm": 0.9903015494346619, | |
| "learning_rate": 1.6399999999999998e-06, | |
| "loss": 1.2027, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.4071156739534472, | |
| "grad_norm": 0.7119297385215759, | |
| "learning_rate": 1.6319999999999998e-06, | |
| "loss": 1.1638, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.41596601469156563, | |
| "grad_norm": 0.9170375466346741, | |
| "learning_rate": 1.624e-06, | |
| "loss": 1.1706, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.42481635542968404, | |
| "grad_norm": 0.7098402976989746, | |
| "learning_rate": 1.616e-06, | |
| "loss": 1.1738, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.43366669616780246, | |
| "grad_norm": 0.9326470494270325, | |
| "learning_rate": 1.608e-06, | |
| "loss": 1.1526, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.44251703690592087, | |
| "grad_norm": 0.7765191197395325, | |
| "learning_rate": 1.6e-06, | |
| "loss": 1.1687, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.4513673776440393, | |
| "grad_norm": 0.8953288197517395, | |
| "learning_rate": 1.592e-06, | |
| "loss": 1.1763, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.4602177183821577, | |
| "grad_norm": 1.0437840223312378, | |
| "learning_rate": 1.584e-06, | |
| "loss": 1.1883, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.4690680591202761, | |
| "grad_norm": 0.8432090282440186, | |
| "learning_rate": 1.576e-06, | |
| "loss": 1.1826, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.47791839985839457, | |
| "grad_norm": 0.7580952048301697, | |
| "learning_rate": 1.568e-06, | |
| "loss": 1.1813, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.486768740596513, | |
| "grad_norm": 0.9914817214012146, | |
| "learning_rate": 1.5599999999999999e-06, | |
| "loss": 1.1751, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.4956190813346314, | |
| "grad_norm": 0.7643041610717773, | |
| "learning_rate": 1.552e-06, | |
| "loss": 1.1615, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.5044694220727498, | |
| "grad_norm": 0.8275469541549683, | |
| "learning_rate": 1.544e-06, | |
| "loss": 1.188, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.5133197628108682, | |
| "grad_norm": 1.950104832649231, | |
| "learning_rate": 1.536e-06, | |
| "loss": 1.2018, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.5221701035489866, | |
| "grad_norm": 0.8191819787025452, | |
| "learning_rate": 1.528e-06, | |
| "loss": 1.1865, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.531020444287105, | |
| "grad_norm": 0.7783402800559998, | |
| "learning_rate": 1.5199999999999998e-06, | |
| "loss": 1.1687, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5398707850252235, | |
| "grad_norm": 1.0385206937789917, | |
| "learning_rate": 1.5119999999999999e-06, | |
| "loss": 1.1883, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.5487211257633419, | |
| "grad_norm": 0.9720994234085083, | |
| "learning_rate": 1.504e-06, | |
| "loss": 1.1824, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.5575714665014603, | |
| "grad_norm": 0.9594699740409851, | |
| "learning_rate": 1.496e-06, | |
| "loss": 1.1908, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.5664218072395787, | |
| "grad_norm": 0.9793533682823181, | |
| "learning_rate": 1.488e-06, | |
| "loss": 1.1385, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.5752721479776971, | |
| "grad_norm": 0.9187788367271423, | |
| "learning_rate": 1.48e-06, | |
| "loss": 1.1536, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.5841224887158155, | |
| "grad_norm": 0.9121326208114624, | |
| "learning_rate": 1.4719999999999998e-06, | |
| "loss": 1.1577, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.5929728294539339, | |
| "grad_norm": 1.3253475427627563, | |
| "learning_rate": 1.4639999999999999e-06, | |
| "loss": 1.1462, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.6018231701920524, | |
| "grad_norm": 0.9218689799308777, | |
| "learning_rate": 1.456e-06, | |
| "loss": 1.1745, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.6106735109301709, | |
| "grad_norm": 0.8204641938209534, | |
| "learning_rate": 1.448e-06, | |
| "loss": 1.1871, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.6195238516682893, | |
| "grad_norm": 0.8714615702629089, | |
| "learning_rate": 1.44e-06, | |
| "loss": 1.183, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.6283741924064077, | |
| "grad_norm": 1.2564398050308228, | |
| "learning_rate": 1.4319999999999998e-06, | |
| "loss": 1.1614, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.6372245331445261, | |
| "grad_norm": 0.8741295337677002, | |
| "learning_rate": 1.4239999999999998e-06, | |
| "loss": 1.1314, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.6460748738826445, | |
| "grad_norm": 0.8480414152145386, | |
| "learning_rate": 1.4159999999999999e-06, | |
| "loss": 1.1522, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.6549252146207629, | |
| "grad_norm": 1.0986804962158203, | |
| "learning_rate": 1.408e-06, | |
| "loss": 1.1885, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.6637755553588813, | |
| "grad_norm": 0.9965903759002686, | |
| "learning_rate": 1.4e-06, | |
| "loss": 1.1804, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.6726258960969997, | |
| "grad_norm": 1.079156756401062, | |
| "learning_rate": 1.3919999999999998e-06, | |
| "loss": 1.1797, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.6814762368351182, | |
| "grad_norm": 1.0092281103134155, | |
| "learning_rate": 1.3839999999999998e-06, | |
| "loss": 1.1613, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.6903265775732366, | |
| "grad_norm": 0.9988503456115723, | |
| "learning_rate": 1.3759999999999998e-06, | |
| "loss": 1.1526, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.699176918311355, | |
| "grad_norm": 1.0048474073410034, | |
| "learning_rate": 1.368e-06, | |
| "loss": 1.1804, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.7080272590494734, | |
| "grad_norm": 0.7530746459960938, | |
| "learning_rate": 1.3600000000000001e-06, | |
| "loss": 1.1465, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.7168775997875918, | |
| "grad_norm": 0.6247321963310242, | |
| "learning_rate": 1.352e-06, | |
| "loss": 1.1385, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.7257279405257102, | |
| "grad_norm": 0.7886361479759216, | |
| "learning_rate": 1.344e-06, | |
| "loss": 1.1857, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.7345782812638286, | |
| "grad_norm": 0.8461400270462036, | |
| "learning_rate": 1.336e-06, | |
| "loss": 1.1612, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.743428622001947, | |
| "grad_norm": 1.109844446182251, | |
| "learning_rate": 1.328e-06, | |
| "loss": 1.1406, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.7522789627400654, | |
| "grad_norm": 0.8707802891731262, | |
| "learning_rate": 1.32e-06, | |
| "loss": 1.1719, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.7611293034781839, | |
| "grad_norm": 0.8593458533287048, | |
| "learning_rate": 1.312e-06, | |
| "loss": 1.1696, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.7699796442163023, | |
| "grad_norm": 0.8872305750846863, | |
| "learning_rate": 1.304e-06, | |
| "loss": 1.1772, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.7788299849544208, | |
| "grad_norm": 0.9227966070175171, | |
| "learning_rate": 1.296e-06, | |
| "loss": 1.173, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.7876803256925392, | |
| "grad_norm": 0.8818754553794861, | |
| "learning_rate": 1.288e-06, | |
| "loss": 1.1889, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.7965306664306576, | |
| "grad_norm": 1.0658597946166992, | |
| "learning_rate": 1.28e-06, | |
| "loss": 1.1547, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.805381007168776, | |
| "grad_norm": 0.8553236126899719, | |
| "learning_rate": 1.272e-06, | |
| "loss": 1.1172, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.8142313479068944, | |
| "grad_norm": 1.120353102684021, | |
| "learning_rate": 1.2639999999999999e-06, | |
| "loss": 1.1649, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.8230816886450129, | |
| "grad_norm": 0.734362006187439, | |
| "learning_rate": 1.256e-06, | |
| "loss": 1.1938, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.8319320293831313, | |
| "grad_norm": 0.8664830923080444, | |
| "learning_rate": 1.248e-06, | |
| "loss": 1.1666, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.8407823701212497, | |
| "grad_norm": 2.742342948913574, | |
| "learning_rate": 1.24e-06, | |
| "loss": 1.1807, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.8496327108593681, | |
| "grad_norm": 1.302372694015503, | |
| "learning_rate": 1.232e-06, | |
| "loss": 1.1709, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.8584830515974865, | |
| "grad_norm": 0.9084349870681763, | |
| "learning_rate": 1.2239999999999998e-06, | |
| "loss": 1.1405, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.8673333923356049, | |
| "grad_norm": 0.8565220832824707, | |
| "learning_rate": 1.2159999999999999e-06, | |
| "loss": 1.1527, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.8761837330737233, | |
| "grad_norm": 0.7832551002502441, | |
| "learning_rate": 1.208e-06, | |
| "loss": 1.1787, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.8850340738118417, | |
| "grad_norm": 0.8366639018058777, | |
| "learning_rate": 1.2e-06, | |
| "loss": 1.1925, | |
| "step": 10000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 25000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.876647618221867e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |