| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.986899563318777, | |
| "eval_steps": 500, | |
| "global_step": 171, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.6939600110054016, | |
| "learning_rate": 4.9894597202472696e-05, | |
| "loss": 2.9255, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.8492265343666077, | |
| "learning_rate": 4.957927758986888e-05, | |
| "loss": 2.8178, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.8294272422790527, | |
| "learning_rate": 4.905670000773126e-05, | |
| "loss": 2.7191, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.6167703866958618, | |
| "learning_rate": 4.833127094718643e-05, | |
| "loss": 2.6713, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.49007371068000793, | |
| "learning_rate": 4.7409107388425504e-05, | |
| "loss": 2.6395, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.4197269678115845, | |
| "learning_rate": 4.629798522095818e-05, | |
| "loss": 2.5998, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.35962584614753723, | |
| "learning_rate": 4.5007273675572104e-05, | |
| "loss": 2.5696, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.31397658586502075, | |
| "learning_rate": 4.3547856320882044e-05, | |
| "loss": 2.5568, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.2711264193058014, | |
| "learning_rate": 4.193203929064353e-05, | |
| "loss": 2.4964, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.29537662863731384, | |
| "learning_rate": 4.0173447515678916e-05, | |
| "loss": 2.5512, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.28811562061309814, | |
| "learning_rate": 3.82869098354114e-05, | |
| "loss": 2.5155, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 0.2557935118675232, | |
| "learning_rate": 3.628833395777224e-05, | |
| "loss": 2.5253, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 0.243013396859169, | |
| "learning_rate": 3.4194572321847336e-05, | |
| "loss": 2.5395, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.26634490489959717, | |
| "learning_rate": 3.202327999433924e-05, | |
| "loss": 2.4728, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 0.288352906703949, | |
| "learning_rate": 2.9792765798093465e-05, | |
| "loss": 2.5059, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.25032252073287964, | |
| "learning_rate": 2.752183792800671e-05, | |
| "loss": 2.5368, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.24827222526073456, | |
| "learning_rate": 2.5229645356118163e-05, | |
| "loss": 2.4811, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 0.2552015781402588, | |
| "learning_rate": 2.2935516363191693e-05, | |
| "loss": 2.5105, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 0.2699149549007416, | |
| "learning_rate": 2.0658795558326743e-05, | |
| "loss": 2.5252, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 0.2718549966812134, | |
| "learning_rate": 1.8418680760885027e-05, | |
| "loss": 2.4448, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 0.2571737468242645, | |
| "learning_rate": 1.6234061120181142e-05, | |
| "loss": 2.438, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.25729846954345703, | |
| "learning_rate": 1.4123357837948175e-05, | |
| "loss": 2.49, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 0.29407253861427307, | |
| "learning_rate": 1.2104368836641908e-05, | |
| "loss": 2.4363, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.2614673376083374, | |
| "learning_rate": 1.0194118683375503e-05, | |
| "loss": 2.4613, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.24367694556713104, | |
| "learning_rate": 8.40871503495947e-06, | |
| "loss": 2.4758, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.26960763335227966, | |
| "learning_rate": 6.763212814534484e-06, | |
| "loss": 2.4591, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.24732153117656708, | |
| "learning_rate": 5.271487265090163e-06, | |
| "loss": 2.4655, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 0.2650165855884552, | |
| "learning_rate": 3.94611695031086e-06, | |
| "loss": 2.4813, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "grad_norm": 0.26127758622169495, | |
| "learning_rate": 2.7982776893115627e-06, | |
| "loss": 2.4453, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.2666429877281189, | |
| "learning_rate": 1.837648319629956e-06, | |
| "loss": 2.4941, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 0.25974979996681213, | |
| "learning_rate": 1.0723290831021471e-06, | |
| "loss": 2.4701, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "grad_norm": 0.2709636092185974, | |
| "learning_rate": 5.087733228106517e-07, | |
| "loss": 2.4543, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.24128825962543488, | |
| "learning_rate": 1.517330670512629e-07, | |
| "loss": 2.5142, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "grad_norm": 0.26422107219696045, | |
| "learning_rate": 4.218959166932268e-09, | |
| "loss": 2.4938, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "step": 171, | |
| "total_flos": 2.565405626125517e+16, | |
| "train_loss": 2.5343901483636153, | |
| "train_runtime": 1859.0377, | |
| "train_samples_per_second": 1.477, | |
| "train_steps_per_second": 0.092 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 171, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "total_flos": 2.565405626125517e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |