| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 30.0, | |
| "eval_steps": 500, | |
| "global_step": 37620, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.39872408293460926, | |
| "grad_norm": 0.7819421887397766, | |
| "learning_rate": 4.800637958532696e-05, | |
| "loss": 8.5954, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7974481658692185, | |
| "grad_norm": 1.4368996620178223, | |
| "learning_rate": 4.6012759170653905e-05, | |
| "loss": 7.6895, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1961722488038278, | |
| "grad_norm": 1.0771726369857788, | |
| "learning_rate": 4.401913875598087e-05, | |
| "loss": 7.5338, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.594896331738437, | |
| "grad_norm": 1.4085040092468262, | |
| "learning_rate": 4.2025518341307815e-05, | |
| "loss": 7.4062, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.9936204146730463, | |
| "grad_norm": 1.2863935232162476, | |
| "learning_rate": 4.003189792663477e-05, | |
| "loss": 7.3026, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.3923444976076556, | |
| "grad_norm": 1.343314528465271, | |
| "learning_rate": 3.8038277511961725e-05, | |
| "loss": 7.2067, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.7910685805422646, | |
| "grad_norm": 1.8292430639266968, | |
| "learning_rate": 3.604465709728867e-05, | |
| "loss": 7.1446, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.189792663476874, | |
| "grad_norm": 1.3264261484146118, | |
| "learning_rate": 3.4051036682615634e-05, | |
| "loss": 7.075, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.588516746411483, | |
| "grad_norm": 1.511952519416809, | |
| "learning_rate": 3.205741626794259e-05, | |
| "loss": 7.0362, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.9872408293460926, | |
| "grad_norm": 1.9909762144088745, | |
| "learning_rate": 3.0063795853269537e-05, | |
| "loss": 6.9639, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.385964912280702, | |
| "grad_norm": 1.8488012552261353, | |
| "learning_rate": 2.8070175438596492e-05, | |
| "loss": 6.9264, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 4.784688995215311, | |
| "grad_norm": 1.4738715887069702, | |
| "learning_rate": 2.6076555023923443e-05, | |
| "loss": 6.9039, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 5.18341307814992, | |
| "grad_norm": 1.7991809844970703, | |
| "learning_rate": 2.4082934609250398e-05, | |
| "loss": 6.8603, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 5.582137161084529, | |
| "grad_norm": 1.811631441116333, | |
| "learning_rate": 2.2089314194577353e-05, | |
| "loss": 6.8354, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 5.980861244019139, | |
| "grad_norm": 1.732555627822876, | |
| "learning_rate": 2.0095693779904308e-05, | |
| "loss": 6.8048, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 6.379585326953748, | |
| "grad_norm": 1.7755110263824463, | |
| "learning_rate": 1.8102073365231263e-05, | |
| "loss": 6.7829, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 6.778309409888357, | |
| "grad_norm": 2.225064277648926, | |
| "learning_rate": 1.611244019138756e-05, | |
| "loss": 6.7659, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 7.177033492822966, | |
| "grad_norm": 1.872410774230957, | |
| "learning_rate": 1.4118819776714515e-05, | |
| "loss": 6.7374, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 7.575757575757576, | |
| "grad_norm": 1.7238253355026245, | |
| "learning_rate": 1.2125199362041468e-05, | |
| "loss": 6.7349, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 7.974481658692185, | |
| "grad_norm": 1.8047776222229004, | |
| "learning_rate": 1.0131578947368421e-05, | |
| "loss": 6.7204, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 8.373205741626794, | |
| "grad_norm": 2.0625622272491455, | |
| "learning_rate": 8.145933014354067e-06, | |
| "loss": 6.6957, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 8.771929824561404, | |
| "grad_norm": 1.8150039911270142, | |
| "learning_rate": 6.152312599681021e-06, | |
| "loss": 6.7014, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 9.170653907496012, | |
| "grad_norm": 2.004660129547119, | |
| "learning_rate": 4.158692185007974e-06, | |
| "loss": 6.6855, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 9.569377990430622, | |
| "grad_norm": 2.1859395503997803, | |
| "learning_rate": 2.1650717703349284e-06, | |
| "loss": 6.6813, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 9.96810207336523, | |
| "grad_norm": 2.1167070865631104, | |
| "learning_rate": 1.7145135566188198e-07, | |
| "loss": 6.6833, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 12540, | |
| "total_flos": 5910924298141440.0, | |
| "train_loss": 7.017849675472083, | |
| "train_runtime": 3142.2268, | |
| "train_samples_per_second": 510.596, | |
| "train_steps_per_second": 3.991 | |
| }, | |
| { | |
| "epoch": 10.36682615629984, | |
| "grad_norm": 2.343043327331543, | |
| "learning_rate": 4.87772461456672e-05, | |
| "loss": 6.6945, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 10.76555023923445, | |
| "grad_norm": 2.0774621963500977, | |
| "learning_rate": 4.7448165869218504e-05, | |
| "loss": 6.6465, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 11.164274322169058, | |
| "grad_norm": 2.5551674365997314, | |
| "learning_rate": 4.611908559276981e-05, | |
| "loss": 6.6086, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 11.562998405103668, | |
| "grad_norm": 2.335573196411133, | |
| "learning_rate": 4.479000531632111e-05, | |
| "loss": 6.5677, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 11.961722488038278, | |
| "grad_norm": 2.7488040924072266, | |
| "learning_rate": 4.346092503987241e-05, | |
| "loss": 6.5439, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 12.360446570972886, | |
| "grad_norm": 2.1827917098999023, | |
| "learning_rate": 4.213184476342371e-05, | |
| "loss": 6.4975, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 12.759170653907496, | |
| "grad_norm": 2.1983468532562256, | |
| "learning_rate": 4.080276448697501e-05, | |
| "loss": 6.4953, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 13.157894736842104, | |
| "grad_norm": 2.4948813915252686, | |
| "learning_rate": 3.9473684210526316e-05, | |
| "loss": 6.4546, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 13.556618819776714, | |
| "grad_norm": 2.0236194133758545, | |
| "learning_rate": 3.814460393407762e-05, | |
| "loss": 6.432, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 13.955342902711324, | |
| "grad_norm": 2.3203213214874268, | |
| "learning_rate": 3.681552365762892e-05, | |
| "loss": 6.4038, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 14.354066985645932, | |
| "grad_norm": 2.347102165222168, | |
| "learning_rate": 3.5486443381180226e-05, | |
| "loss": 6.3872, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 14.752791068580542, | |
| "grad_norm": 2.5316317081451416, | |
| "learning_rate": 3.415736310473153e-05, | |
| "loss": 6.3687, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 18810, | |
| "total_flos": 8866386447212160.0, | |
| "train_loss": 2.1665261722384206, | |
| "train_runtime": 1599.9695, | |
| "train_samples_per_second": 1504.163, | |
| "train_steps_per_second": 11.756 | |
| }, | |
| { | |
| "epoch": 15.151515151515152, | |
| "grad_norm": 2.2735719680786133, | |
| "learning_rate": 4.974747474747475e-05, | |
| "loss": 6.3524, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 15.55023923444976, | |
| "grad_norm": 3.0628674030303955, | |
| "learning_rate": 4.9082934609250406e-05, | |
| "loss": 6.3525, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 15.94896331738437, | |
| "grad_norm": 2.572157382965088, | |
| "learning_rate": 4.841839447102605e-05, | |
| "loss": 6.3283, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 16.34768740031898, | |
| "grad_norm": 2.8556103706359863, | |
| "learning_rate": 4.775385433280171e-05, | |
| "loss": 6.3006, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 16.74641148325359, | |
| "grad_norm": 2.7908451557159424, | |
| "learning_rate": 4.7089314194577354e-05, | |
| "loss": 6.2868, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 17.1451355661882, | |
| "grad_norm": 2.637225389480591, | |
| "learning_rate": 4.6424774056353006e-05, | |
| "loss": 6.2645, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 17.54385964912281, | |
| "grad_norm": 2.9248273372650146, | |
| "learning_rate": 4.576023391812866e-05, | |
| "loss": 6.24, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 17.942583732057415, | |
| "grad_norm": 2.293976306915283, | |
| "learning_rate": 4.509569377990431e-05, | |
| "loss": 6.2331, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 18.341307814992025, | |
| "grad_norm": 2.7412922382354736, | |
| "learning_rate": 4.443115364167996e-05, | |
| "loss": 6.1999, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 18.740031897926634, | |
| "grad_norm": 3.194049119949341, | |
| "learning_rate": 4.376661350345561e-05, | |
| "loss": 6.1953, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 19.138755980861244, | |
| "grad_norm": 2.366976022720337, | |
| "learning_rate": 4.310207336523126e-05, | |
| "loss": 6.1713, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 19.537480063795854, | |
| "grad_norm": 2.459841728210449, | |
| "learning_rate": 4.2437533227006915e-05, | |
| "loss": 6.1535, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 19.93620414673046, | |
| "grad_norm": 2.5043203830718994, | |
| "learning_rate": 4.177432216905902e-05, | |
| "loss": 6.1481, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 20.33492822966507, | |
| "grad_norm": 2.6015625, | |
| "learning_rate": 4.1109782030834664e-05, | |
| "loss": 6.1228, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 20.73365231259968, | |
| "grad_norm": 3.39577579498291, | |
| "learning_rate": 4.0445241892610315e-05, | |
| "loss": 6.1113, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 21.13237639553429, | |
| "grad_norm": 2.7234418392181396, | |
| "learning_rate": 3.978070175438597e-05, | |
| "loss": 6.1074, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 21.5311004784689, | |
| "grad_norm": 2.6176741123199463, | |
| "learning_rate": 3.911749069643807e-05, | |
| "loss": 6.0859, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 21.92982456140351, | |
| "grad_norm": 2.5121538639068604, | |
| "learning_rate": 3.845427963849017e-05, | |
| "loss": 6.0804, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 22.328548644338117, | |
| "grad_norm": 2.72497296333313, | |
| "learning_rate": 3.778973950026582e-05, | |
| "loss": 6.0623, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 22.727272727272727, | |
| "grad_norm": 2.9392683506011963, | |
| "learning_rate": 3.712519936204147e-05, | |
| "loss": 6.0536, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 23.125996810207337, | |
| "grad_norm": 2.5510990619659424, | |
| "learning_rate": 3.646065922381712e-05, | |
| "loss": 6.0508, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 23.524720893141946, | |
| "grad_norm": 2.9767627716064453, | |
| "learning_rate": 3.5796119085592774e-05, | |
| "loss": 6.029, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 23.923444976076556, | |
| "grad_norm": 2.644033908843994, | |
| "learning_rate": 3.513157894736842e-05, | |
| "loss": 6.0329, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 24.322169059011163, | |
| "grad_norm": 2.623633861541748, | |
| "learning_rate": 3.446836788942052e-05, | |
| "loss": 6.003, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 24.720893141945773, | |
| "grad_norm": 2.6472909450531006, | |
| "learning_rate": 3.3803827751196174e-05, | |
| "loss": 6.0, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 25.119617224880383, | |
| "grad_norm": 3.0670645236968994, | |
| "learning_rate": 3.3139287612971825e-05, | |
| "loss": 5.9948, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 25.518341307814993, | |
| "grad_norm": 2.6866748332977295, | |
| "learning_rate": 3.247474747474748e-05, | |
| "loss": 5.9883, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 25.917065390749602, | |
| "grad_norm": 2.8670027256011963, | |
| "learning_rate": 3.181020733652313e-05, | |
| "loss": 5.9801, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 26.31578947368421, | |
| "grad_norm": 2.797853708267212, | |
| "learning_rate": 3.114566719829878e-05, | |
| "loss": 5.9665, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 26.71451355661882, | |
| "grad_norm": 3.076382637023926, | |
| "learning_rate": 3.0481127060074432e-05, | |
| "loss": 5.9655, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 27.11323763955343, | |
| "grad_norm": 2.7438435554504395, | |
| "learning_rate": 2.981791600212653e-05, | |
| "loss": 5.9445, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 27.51196172248804, | |
| "grad_norm": 3.119704484939575, | |
| "learning_rate": 2.9153375863902184e-05, | |
| "loss": 5.9319, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 27.91068580542265, | |
| "grad_norm": 2.594749927520752, | |
| "learning_rate": 2.8488835725677832e-05, | |
| "loss": 5.9327, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 28.30940988835726, | |
| "grad_norm": 2.6697604656219482, | |
| "learning_rate": 2.7824295587453487e-05, | |
| "loss": 5.9261, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 28.708133971291865, | |
| "grad_norm": 3.4312615394592285, | |
| "learning_rate": 2.7159755449229135e-05, | |
| "loss": 5.9146, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 29.106858054226475, | |
| "grad_norm": 3.0306286811828613, | |
| "learning_rate": 2.6495215311004783e-05, | |
| "loss": 5.9243, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 29.505582137161085, | |
| "grad_norm": 2.841744899749756, | |
| "learning_rate": 2.583067517278044e-05, | |
| "loss": 5.9046, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 29.904306220095695, | |
| "grad_norm": 2.9730348587036133, | |
| "learning_rate": 2.5166135034556087e-05, | |
| "loss": 5.8976, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "step": 37620, | |
| "total_flos": 1.773277289442432e+16, | |
| "train_loss": 3.0397992164007466, | |
| "train_runtime": 4647.9256, | |
| "train_samples_per_second": 1035.565, | |
| "train_steps_per_second": 8.094 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 37620, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.773277289442432e+16, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |