| { | |
| "best_global_step": 20000, | |
| "best_metric": 6.807223796844482, | |
| "best_model_checkpoint": "bert_base_km_50_v1/checkpoint-20000", | |
| "epoch": 25.0, | |
| "eval_steps": 10000, | |
| "global_step": 59550, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.20990764063811923, | |
| "grad_norm": 2.9680073261260986, | |
| "learning_rate": 5e-06, | |
| "loss": 13.1255, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 4.950173854827881, | |
| "learning_rate": 1e-05, | |
| "loss": 11.1725, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 21.20996856689453, | |
| "learning_rate": 1.5e-05, | |
| "loss": 9.4837, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 9.111143112182617, | |
| "learning_rate": 2e-05, | |
| "loss": 8.7334, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": 34.0069694519043, | |
| "learning_rate": 2.5e-05, | |
| "loss": 8.3231, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": 21.335039138793945, | |
| "learning_rate": 3e-05, | |
| "loss": 8.0483, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": 10.246840476989746, | |
| "learning_rate": 3.5e-05, | |
| "loss": 7.8945, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": 8.814268112182617, | |
| "learning_rate": 4e-05, | |
| "loss": 7.7738, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": 12.415926933288574, | |
| "learning_rate": 4.5e-05, | |
| "loss": 7.6923, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": 18.443321228027344, | |
| "learning_rate": 5e-05, | |
| "loss": 7.5486, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": 8.008631706237793, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 7.4581, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": 13.306659698486328, | |
| "learning_rate": 6e-05, | |
| "loss": 7.4253, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.72879932829555, | |
| "grad_norm": 7.610831260681152, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 7.3955, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.938706968933669, | |
| "grad_norm": 4.364519119262695, | |
| "learning_rate": 7e-05, | |
| "loss": 7.329, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.1486146095717884, | |
| "grad_norm": 4.721391201019287, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 7.2307, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.3585222502099077, | |
| "grad_norm": 4.419271945953369, | |
| "learning_rate": 8e-05, | |
| "loss": 7.2015, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.568429890848027, | |
| "grad_norm": 4.827095031738281, | |
| "learning_rate": 8.5e-05, | |
| "loss": 7.1881, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 3.7783375314861463, | |
| "grad_norm": 4.566966533660889, | |
| "learning_rate": 9e-05, | |
| "loss": 7.1595, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.988245172124265, | |
| "grad_norm": 6.984473705291748, | |
| "learning_rate": 9.5e-05, | |
| "loss": 7.1469, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "grad_norm": 4.0145697593688965, | |
| "learning_rate": 0.0001, | |
| "loss": 7.0285, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.198152812762385, | |
| "eval_accuracy": 0.15003199866440356, | |
| "eval_loss": 7.077643394470215, | |
| "eval_runtime": 1.0526, | |
| "eval_samples_per_second": 455.059, | |
| "eval_steps_per_second": 4.75, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.408060453400504, | |
| "grad_norm": 5.073930263519287, | |
| "learning_rate": 9.899091826437942e-05, | |
| "loss": 7.0406, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 4.617968094038623, | |
| "grad_norm": 5.878479480743408, | |
| "learning_rate": 9.798183652875883e-05, | |
| "loss": 7.0228, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.827875734676742, | |
| "grad_norm": 5.614640712738037, | |
| "learning_rate": 9.697275479313825e-05, | |
| "loss": 7.0053, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.037783375314861, | |
| "grad_norm": 5.407928466796875, | |
| "learning_rate": 9.596367305751767e-05, | |
| "loss": 6.9612, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.247691015952981, | |
| "grad_norm": 4.4640302658081055, | |
| "learning_rate": 9.495459132189707e-05, | |
| "loss": 6.8617, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 5.4575986565911, | |
| "grad_norm": 5.564060688018799, | |
| "learning_rate": 9.394550958627649e-05, | |
| "loss": 6.8604, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.667506297229219, | |
| "grad_norm": 4.915012359619141, | |
| "learning_rate": 9.293642785065591e-05, | |
| "loss": 6.8643, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 5.877413937867338, | |
| "grad_norm": 3.3723433017730713, | |
| "learning_rate": 9.192734611503532e-05, | |
| "loss": 6.8497, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.087321578505458, | |
| "grad_norm": 4.637331008911133, | |
| "learning_rate": 9.091826437941474e-05, | |
| "loss": 6.7895, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.297229219143577, | |
| "grad_norm": 4.291222095489502, | |
| "learning_rate": 8.990918264379415e-05, | |
| "loss": 6.7339, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.507136859781696, | |
| "grad_norm": 3.7567830085754395, | |
| "learning_rate": 8.890010090817356e-05, | |
| "loss": 6.7376, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 6.717044500419815, | |
| "grad_norm": 3.9562442302703857, | |
| "learning_rate": 8.789101917255298e-05, | |
| "loss": 6.736, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.926952141057934, | |
| "grad_norm": 3.4675660133361816, | |
| "learning_rate": 8.68819374369324e-05, | |
| "loss": 6.7293, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.136859781696054, | |
| "grad_norm": 3.6558496952056885, | |
| "learning_rate": 8.58728557013118e-05, | |
| "loss": 6.649, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.346767422334173, | |
| "grad_norm": 4.217350959777832, | |
| "learning_rate": 8.486377396569122e-05, | |
| "loss": 6.6156, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 7.5566750629722925, | |
| "grad_norm": 4.8325276374816895, | |
| "learning_rate": 8.385469223007064e-05, | |
| "loss": 6.6248, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.766582703610411, | |
| "grad_norm": 5.328977108001709, | |
| "learning_rate": 8.284561049445006e-05, | |
| "loss": 6.6405, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 7.97649034424853, | |
| "grad_norm": 3.9508869647979736, | |
| "learning_rate": 8.183652875882947e-05, | |
| "loss": 6.6366, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.18639798488665, | |
| "grad_norm": 2.976325273513794, | |
| "learning_rate": 8.082744702320889e-05, | |
| "loss": 6.5198, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "grad_norm": 3.8047006130218506, | |
| "learning_rate": 7.98183652875883e-05, | |
| "loss": 6.5312, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.39630562552477, | |
| "eval_accuracy": 0.1518280535368557, | |
| "eval_loss": 6.807223796844482, | |
| "eval_runtime": 1.0263, | |
| "eval_samples_per_second": 466.745, | |
| "eval_steps_per_second": 4.872, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.606213266162888, | |
| "grad_norm": 5.03962516784668, | |
| "learning_rate": 7.880928355196771e-05, | |
| "loss": 6.5353, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 8.816120906801007, | |
| "grad_norm": 3.950474262237549, | |
| "learning_rate": 7.780020181634713e-05, | |
| "loss": 6.5435, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 9.026028547439127, | |
| "grad_norm": 2.6887052059173584, | |
| "learning_rate": 7.679112008072655e-05, | |
| "loss": 6.526, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 9.235936188077247, | |
| "grad_norm": 4.982834339141846, | |
| "learning_rate": 7.578203834510595e-05, | |
| "loss": 6.4232, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 9.445843828715365, | |
| "grad_norm": 3.9568865299224854, | |
| "learning_rate": 7.477295660948537e-05, | |
| "loss": 6.4393, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 9.655751469353484, | |
| "grad_norm": 4.367034435272217, | |
| "learning_rate": 7.376387487386479e-05, | |
| "loss": 6.4488, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 9.865659109991604, | |
| "grad_norm": 4.204073905944824, | |
| "learning_rate": 7.27547931382442e-05, | |
| "loss": 6.4469, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 10.075566750629722, | |
| "grad_norm": 3.058058261871338, | |
| "learning_rate": 7.174571140262362e-05, | |
| "loss": 6.408, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 10.285474391267842, | |
| "grad_norm": 3.253079414367676, | |
| "learning_rate": 7.073662966700304e-05, | |
| "loss": 6.3531, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 10.495382031905962, | |
| "grad_norm": 2.9021670818328857, | |
| "learning_rate": 6.972754793138244e-05, | |
| "loss": 6.3593, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 10.705289672544081, | |
| "grad_norm": 3.753322124481201, | |
| "learning_rate": 6.871846619576186e-05, | |
| "loss": 6.3708, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 10.9151973131822, | |
| "grad_norm": 4.464712619781494, | |
| "learning_rate": 6.770938446014128e-05, | |
| "loss": 6.374, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 11.125104953820319, | |
| "grad_norm": 5.466015338897705, | |
| "learning_rate": 6.670030272452068e-05, | |
| "loss": 6.313, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 11.335012594458439, | |
| "grad_norm": 3.907973527908325, | |
| "learning_rate": 6.56912209889001e-05, | |
| "loss": 6.2821, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 11.544920235096557, | |
| "grad_norm": 3.5810680389404297, | |
| "learning_rate": 6.468213925327952e-05, | |
| "loss": 6.288, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 11.754827875734676, | |
| "grad_norm": 4.42948579788208, | |
| "learning_rate": 6.367305751765893e-05, | |
| "loss": 6.2918, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 11.964735516372796, | |
| "grad_norm": 4.894264221191406, | |
| "learning_rate": 6.266397578203835e-05, | |
| "loss": 6.294, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 12.174643157010916, | |
| "grad_norm": 4.274178504943848, | |
| "learning_rate": 6.165489404641777e-05, | |
| "loss": 6.2289, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 12.384550797649034, | |
| "grad_norm": 3.1867642402648926, | |
| "learning_rate": 6.064581231079718e-05, | |
| "loss": 6.2203, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "grad_norm": 3.951848268508911, | |
| "learning_rate": 5.963673057517659e-05, | |
| "loss": 6.2368, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.594458438287154, | |
| "eval_accuracy": 0.15351036740632956, | |
| "eval_loss": 6.913123607635498, | |
| "eval_runtime": 1.0392, | |
| "eval_samples_per_second": 460.911, | |
| "eval_steps_per_second": 4.811, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 12.804366078925273, | |
| "grad_norm": 3.7859292030334473, | |
| "learning_rate": 5.8627648839556e-05, | |
| "loss": 6.237, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 13.014273719563391, | |
| "grad_norm": 3.712524175643921, | |
| "learning_rate": 5.7618567103935416e-05, | |
| "loss": 6.2313, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 13.224181360201511, | |
| "grad_norm": 5.6810688972473145, | |
| "learning_rate": 5.660948536831484e-05, | |
| "loss": 6.1725, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 13.43408900083963, | |
| "grad_norm": 5.267341136932373, | |
| "learning_rate": 5.5600403632694254e-05, | |
| "loss": 6.1773, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 13.64399664147775, | |
| "grad_norm": 3.7381560802459717, | |
| "learning_rate": 5.4591321897073666e-05, | |
| "loss": 6.1867, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 13.853904282115868, | |
| "grad_norm": 4.101844310760498, | |
| "learning_rate": 5.358224016145308e-05, | |
| "loss": 6.1797, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 14.063811922753988, | |
| "grad_norm": 3.276782751083374, | |
| "learning_rate": 5.257315842583249e-05, | |
| "loss": 6.1641, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 14.273719563392108, | |
| "grad_norm": 3.653540849685669, | |
| "learning_rate": 5.15640766902119e-05, | |
| "loss": 6.1294, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 14.483627204030226, | |
| "grad_norm": 4.354523181915283, | |
| "learning_rate": 5.055499495459133e-05, | |
| "loss": 6.1368, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 14.693534844668346, | |
| "grad_norm": 4.55731725692749, | |
| "learning_rate": 4.954591321897074e-05, | |
| "loss": 6.1391, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 14.903442485306465, | |
| "grad_norm": 3.481832981109619, | |
| "learning_rate": 4.853683148335015e-05, | |
| "loss": 6.139, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 15.113350125944585, | |
| "grad_norm": 3.067518711090088, | |
| "learning_rate": 4.7527749747729566e-05, | |
| "loss": 6.1122, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 15.323257766582703, | |
| "grad_norm": 4.280571937561035, | |
| "learning_rate": 4.6518668012108985e-05, | |
| "loss": 6.1001, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 15.533165407220823, | |
| "grad_norm": 4.327350616455078, | |
| "learning_rate": 4.55095862764884e-05, | |
| "loss": 6.0982, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 15.743073047858942, | |
| "grad_norm": 3.9131927490234375, | |
| "learning_rate": 4.450050454086781e-05, | |
| "loss": 6.0966, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 15.95298068849706, | |
| "grad_norm": 5.128420352935791, | |
| "learning_rate": 4.349142280524723e-05, | |
| "loss": 6.0987, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 16.162888329135182, | |
| "grad_norm": 4.03704833984375, | |
| "learning_rate": 4.248234106962664e-05, | |
| "loss": 6.0742, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 16.3727959697733, | |
| "grad_norm": 4.2524638175964355, | |
| "learning_rate": 4.147325933400605e-05, | |
| "loss": 6.0675, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 16.582703610411418, | |
| "grad_norm": 3.9702067375183105, | |
| "learning_rate": 4.046417759838547e-05, | |
| "loss": 6.0751, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "grad_norm": 3.5109400749206543, | |
| "learning_rate": 3.9455095862764884e-05, | |
| "loss": 6.0623, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 16.79261125104954, | |
| "eval_accuracy": 0.15153639116294718, | |
| "eval_loss": 7.103849411010742, | |
| "eval_runtime": 1.0214, | |
| "eval_samples_per_second": 468.983, | |
| "eval_steps_per_second": 4.895, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 17.002518891687657, | |
| "grad_norm": 4.129485607147217, | |
| "learning_rate": 3.84460141271443e-05, | |
| "loss": 6.0667, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 17.212426532325775, | |
| "grad_norm": 3.6876776218414307, | |
| "learning_rate": 3.7436932391523715e-05, | |
| "loss": 6.0384, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 17.422334172963897, | |
| "grad_norm": 3.0906782150268555, | |
| "learning_rate": 3.642785065590313e-05, | |
| "loss": 6.0332, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 17.632241813602015, | |
| "grad_norm": 1.826346516609192, | |
| "learning_rate": 3.541876892028255e-05, | |
| "loss": 6.0371, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 17.842149454240133, | |
| "grad_norm": 2.580146551132202, | |
| "learning_rate": 3.440968718466196e-05, | |
| "loss": 6.0362, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 18.052057094878254, | |
| "grad_norm": 3.399287700653076, | |
| "learning_rate": 3.340060544904137e-05, | |
| "loss": 6.0285, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 18.261964735516372, | |
| "grad_norm": 3.7760932445526123, | |
| "learning_rate": 3.239152371342079e-05, | |
| "loss": 6.0123, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 18.471872376154494, | |
| "grad_norm": 4.151535987854004, | |
| "learning_rate": 3.13824419778002e-05, | |
| "loss": 6.0135, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 18.68178001679261, | |
| "grad_norm": 2.828458786010742, | |
| "learning_rate": 3.0373360242179615e-05, | |
| "loss": 6.0122, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 18.89168765743073, | |
| "grad_norm": 4.07775354385376, | |
| "learning_rate": 2.9364278506559034e-05, | |
| "loss": 6.0172, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 19.10159529806885, | |
| "grad_norm": 4.520580291748047, | |
| "learning_rate": 2.8355196770938446e-05, | |
| "loss": 6.0055, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 19.31150293870697, | |
| "grad_norm": 4.222710132598877, | |
| "learning_rate": 2.734611503531786e-05, | |
| "loss": 5.9863, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 19.521410579345087, | |
| "grad_norm": 2.6324989795684814, | |
| "learning_rate": 2.6337033299697278e-05, | |
| "loss": 5.9887, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 19.73131821998321, | |
| "grad_norm": 5.462238788604736, | |
| "learning_rate": 2.532795156407669e-05, | |
| "loss": 5.9909, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 19.941225860621326, | |
| "grad_norm": 4.209476947784424, | |
| "learning_rate": 2.4318869828456106e-05, | |
| "loss": 5.9855, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 20.151133501259444, | |
| "grad_norm": 3.8022096157073975, | |
| "learning_rate": 2.330978809283552e-05, | |
| "loss": 5.9773, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 20.361041141897566, | |
| "grad_norm": 2.194345474243164, | |
| "learning_rate": 2.2300706357214934e-05, | |
| "loss": 5.9739, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 20.570948782535684, | |
| "grad_norm": 2.65496826171875, | |
| "learning_rate": 2.129162462159435e-05, | |
| "loss": 5.9669, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 20.7808564231738, | |
| "grad_norm": 3.6479873657226562, | |
| "learning_rate": 2.0282542885973765e-05, | |
| "loss": 5.966, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "grad_norm": 3.799025058746338, | |
| "learning_rate": 1.927346115035318e-05, | |
| "loss": 5.966, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 20.990764063811923, | |
| "eval_accuracy": 0.1519331595031822, | |
| "eval_loss": 7.330878257751465, | |
| "eval_runtime": 1.0203, | |
| "eval_samples_per_second": 469.472, | |
| "eval_steps_per_second": 4.901, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 21.20067170445004, | |
| "grad_norm": 4.723677158355713, | |
| "learning_rate": 1.8264379414732593e-05, | |
| "loss": 5.954, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 21.410579345088163, | |
| "grad_norm": 4.613157749176025, | |
| "learning_rate": 1.725529767911201e-05, | |
| "loss": 5.9568, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 21.62048698572628, | |
| "grad_norm": 1.8885151147842407, | |
| "learning_rate": 1.6246215943491424e-05, | |
| "loss": 5.9514, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 21.8303946263644, | |
| "grad_norm": 4.836048126220703, | |
| "learning_rate": 1.5237134207870837e-05, | |
| "loss": 5.9492, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 22.04030226700252, | |
| "grad_norm": 2.0830018520355225, | |
| "learning_rate": 1.4228052472250252e-05, | |
| "loss": 5.9477, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 22.250209907640638, | |
| "grad_norm": 3.134883403778076, | |
| "learning_rate": 1.3218970736629668e-05, | |
| "loss": 5.9321, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 22.460117548278756, | |
| "grad_norm": 3.0352067947387695, | |
| "learning_rate": 1.2209889001009082e-05, | |
| "loss": 5.9357, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 22.670025188916878, | |
| "grad_norm": 1.1481927633285522, | |
| "learning_rate": 1.1200807265388496e-05, | |
| "loss": 5.9369, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 22.879932829554996, | |
| "grad_norm": 1.1260007619857788, | |
| "learning_rate": 1.0191725529767912e-05, | |
| "loss": 5.9316, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 23.089840470193113, | |
| "grad_norm": 1.700620412826538, | |
| "learning_rate": 9.182643794147326e-06, | |
| "loss": 5.9283, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 23.299748110831235, | |
| "grad_norm": 2.43479323387146, | |
| "learning_rate": 8.173562058526741e-06, | |
| "loss": 5.9257, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 23.509655751469353, | |
| "grad_norm": 4.065369129180908, | |
| "learning_rate": 7.164480322906156e-06, | |
| "loss": 5.9245, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 23.719563392107474, | |
| "grad_norm": 6.196600437164307, | |
| "learning_rate": 6.155398587285571e-06, | |
| "loss": 5.9192, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 23.929471032745592, | |
| "grad_norm": 3.683784008026123, | |
| "learning_rate": 5.146316851664986e-06, | |
| "loss": 5.9214, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 24.13937867338371, | |
| "grad_norm": 1.2258681058883667, | |
| "learning_rate": 4.1372351160444e-06, | |
| "loss": 5.9138, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 24.349286314021832, | |
| "grad_norm": 0.9102705717086792, | |
| "learning_rate": 3.1281533804238145e-06, | |
| "loss": 5.914, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 24.55919395465995, | |
| "grad_norm": 3.7073111534118652, | |
| "learning_rate": 2.1190716448032293e-06, | |
| "loss": 5.9116, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 24.769101595298068, | |
| "grad_norm": 1.2236994504928589, | |
| "learning_rate": 1.109989909182644e-06, | |
| "loss": 5.9085, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 24.97900923593619, | |
| "grad_norm": 3.0394434928894043, | |
| "learning_rate": 1.0090817356205853e-07, | |
| "loss": 5.9091, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "step": 59550, | |
| "total_flos": 1.5155177772432384e+18, | |
| "train_loss": 6.5591482797617875, | |
| "train_runtime": 24977.0462, | |
| "train_samples_per_second": 228.849, | |
| "train_steps_per_second": 2.384 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 59550, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.5155177772432384e+18, | |
| "train_batch_size": 96, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |