{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.187061797200861, "eval_steps": 187, "global_step": 1309, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 6.78125, "learning_rate": 1e-05, "loss": 1.644, "step": 1 }, { "epoch": 0.03, "grad_norm": 3.90625, "learning_rate": 1e-05, "loss": 1.3919, "step": 187 }, { "epoch": 0.03, "eval_validation_loss": 0.9903513789176941, "eval_validation_runtime": 194.2676, "eval_validation_samples_per_second": 1.457, "eval_validation_steps_per_second": 0.731, "step": 187 }, { "epoch": 0.03, "eval_validation_privacy_sources_loss": 0.9903513789176941, "eval_validation_privacy_sources_runtime": 193.9712, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 187 }, { "epoch": 0.03, "eval_validation_agenda_digitale_loss": 1.5542374849319458, "eval_validation_agenda_digitale_runtime": 175.0962, "eval_validation_agenda_digitale_samples_per_second": 1.451, "eval_validation_agenda_digitale_steps_per_second": 0.725, "step": 187 }, { "epoch": 0.03, "eval_validation_legal_articles_loss": 1.4528673887252808, "eval_validation_legal_articles_runtime": 282.6051, "eval_validation_legal_articles_samples_per_second": 1.447, "eval_validation_legal_articles_steps_per_second": 0.725, "step": 187 }, { "epoch": 0.03, "eval_validation_dolma_loss": 1.8500893115997314, "eval_validation_dolma_runtime": 111.0119, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.73, "step": 187 }, { "epoch": 0.03, "eval_validation_leggepertutti_loss": 1.314853549003601, "eval_validation_leggepertutti_runtime": 32.3401, "eval_validation_leggepertutti_samples_per_second": 1.453, "eval_validation_leggepertutti_steps_per_second": 0.742, "step": 187 }, { "epoch": 0.03, "eval_validation_stack_loss": 0.6712294220924377, "eval_validation_stack_runtime": 150.1286, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 187 }, { "epoch": 0.05, "grad_norm": 3.859375, "learning_rate": 1e-05, "loss": 1.347, "step": 374 }, { "epoch": 0.05, "eval_validation_loss": 0.9417576789855957, "eval_validation_runtime": 193.9636, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 374 }, { "epoch": 0.05, "eval_validation_privacy_sources_loss": 0.9417576789855957, "eval_validation_privacy_sources_runtime": 193.9776, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 374 }, { "epoch": 0.05, "eval_validation_agenda_digitale_loss": 1.5321811437606812, "eval_validation_agenda_digitale_runtime": 174.0643, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.73, "step": 374 }, { "epoch": 0.05, "eval_validation_legal_articles_loss": 1.4296847581863403, "eval_validation_legal_articles_runtime": 280.3604, "eval_validation_legal_articles_samples_per_second": 1.459, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 374 }, { "epoch": 0.05, "eval_validation_dolma_loss": 1.8486019372940063, "eval_validation_dolma_runtime": 111.0186, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.73, "step": 374 }, { "epoch": 0.05, "eval_validation_leggepertutti_loss": 1.2959588766098022, "eval_validation_leggepertutti_runtime": 32.2131, "eval_validation_leggepertutti_samples_per_second": 1.459, "eval_validation_leggepertutti_steps_per_second": 0.745, "step": 374 }, { "epoch": 0.05, "eval_validation_stack_loss": 0.6691831946372986, "eval_validation_stack_runtime": 150.1411, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 374 }, { "epoch": 0.08, "grad_norm": 4.125, "learning_rate": 1e-05, "loss": 1.3201, "step": 561 }, { "epoch": 0.08, "eval_validation_loss": 0.9132883548736572, "eval_validation_runtime": 193.9346, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 561 }, { "epoch": 0.08, "eval_validation_privacy_sources_loss": 0.9132883548736572, "eval_validation_privacy_sources_runtime": 193.9654, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 561 }, { "epoch": 0.08, "eval_validation_agenda_digitale_loss": 1.5173826217651367, "eval_validation_agenda_digitale_runtime": 174.0763, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.73, "step": 561 }, { "epoch": 0.08, "eval_validation_legal_articles_loss": 1.4147425889968872, "eval_validation_legal_articles_runtime": 280.3444, "eval_validation_legal_articles_samples_per_second": 1.459, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 561 }, { "epoch": 0.08, "eval_validation_dolma_loss": 1.8503817319869995, "eval_validation_dolma_runtime": 110.9998, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.73, "step": 561 }, { "epoch": 0.08, "eval_validation_leggepertutti_loss": 1.2830017805099487, "eval_validation_leggepertutti_runtime": 32.2309, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.745, "step": 561 }, { "epoch": 0.08, "eval_validation_stack_loss": 0.676902711391449, "eval_validation_stack_runtime": 150.1138, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 561 }, { "epoch": 0.11, "grad_norm": 3.65625, "learning_rate": 1e-05, "loss": 1.2986, "step": 748 }, { "epoch": 0.11, "eval_validation_loss": 0.8927286863327026, "eval_validation_runtime": 193.9716, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 748 }, { "epoch": 0.11, "eval_validation_privacy_sources_loss": 0.8927286863327026, "eval_validation_privacy_sources_runtime": 193.9644, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 748 }, { "epoch": 0.11, "eval_validation_agenda_digitale_loss": 1.505477786064148, "eval_validation_agenda_digitale_runtime": 174.1137, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 748 }, { "epoch": 0.11, "eval_validation_legal_articles_loss": 1.4018447399139404, "eval_validation_legal_articles_runtime": 280.3891, "eval_validation_legal_articles_samples_per_second": 1.459, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 748 }, { "epoch": 0.11, "eval_validation_dolma_loss": 1.850791096687317, "eval_validation_dolma_runtime": 111.0043, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.73, "step": 748 }, { "epoch": 0.11, "eval_validation_leggepertutti_loss": 1.272330403327942, "eval_validation_leggepertutti_runtime": 32.2119, "eval_validation_leggepertutti_samples_per_second": 1.459, "eval_validation_leggepertutti_steps_per_second": 0.745, "step": 748 }, { "epoch": 0.11, "eval_validation_stack_loss": 0.6671051979064941, "eval_validation_stack_runtime": 150.1118, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 748 }, { "epoch": 0.13, "grad_norm": 2.984375, "learning_rate": 1e-05, "loss": 1.2799, "step": 935 }, { "epoch": 0.13, "eval_validation_loss": 0.8765723705291748, "eval_validation_runtime": 193.9707, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 935 }, { "epoch": 0.13, "eval_validation_privacy_sources_loss": 0.8765723705291748, "eval_validation_privacy_sources_runtime": 193.9898, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 935 }, { "epoch": 0.13, "eval_validation_agenda_digitale_loss": 1.4963380098342896, "eval_validation_agenda_digitale_runtime": 174.0485, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.73, "step": 935 }, { "epoch": 0.13, "eval_validation_legal_articles_loss": 1.3923107385635376, "eval_validation_legal_articles_runtime": 280.3134, "eval_validation_legal_articles_samples_per_second": 1.459, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 935 }, { "epoch": 0.13, "eval_validation_dolma_loss": 1.8486608266830444, "eval_validation_dolma_runtime": 110.998, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.73, "step": 935 }, { "epoch": 0.13, "eval_validation_leggepertutti_loss": 1.264065146446228, "eval_validation_leggepertutti_runtime": 32.2304, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.745, "step": 935 }, { "epoch": 0.13, "eval_validation_stack_loss": 0.6646791100502014, "eval_validation_stack_runtime": 150.0945, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 935 }, { "epoch": 0.16, "grad_norm": 3.125, "learning_rate": 1e-05, "loss": 1.2685, "step": 1122 }, { "epoch": 0.16, "eval_validation_loss": 0.8656526803970337, "eval_validation_runtime": 193.9746, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 1122 }, { "epoch": 0.16, "eval_validation_privacy_sources_loss": 0.8656526803970337, "eval_validation_privacy_sources_runtime": 193.9453, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 1122 }, { "epoch": 0.16, "eval_validation_agenda_digitale_loss": 1.4903523921966553, "eval_validation_agenda_digitale_runtime": 173.9959, "eval_validation_agenda_digitale_samples_per_second": 1.46, "eval_validation_agenda_digitale_steps_per_second": 0.73, "step": 1122 }, { "epoch": 0.16, "eval_validation_legal_articles_loss": 1.385682463645935, "eval_validation_legal_articles_runtime": 280.251, "eval_validation_legal_articles_samples_per_second": 1.459, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 1122 }, { "epoch": 0.16, "eval_validation_dolma_loss": 1.8492717742919922, "eval_validation_dolma_runtime": 110.9751, "eval_validation_dolma_samples_per_second": 1.46, "eval_validation_dolma_steps_per_second": 0.73, "step": 1122 }, { "epoch": 0.16, "eval_validation_leggepertutti_loss": 1.2608650922775269, "eval_validation_leggepertutti_runtime": 32.2117, "eval_validation_leggepertutti_samples_per_second": 1.459, "eval_validation_leggepertutti_steps_per_second": 0.745, "step": 1122 }, { "epoch": 0.16, "eval_validation_stack_loss": 0.6671770215034485, "eval_validation_stack_runtime": 150.0329, "eval_validation_stack_samples_per_second": 1.46, "eval_validation_stack_steps_per_second": 0.733, "step": 1122 }, { "epoch": 0.19, "grad_norm": 2.90625, "learning_rate": 1e-05, "loss": 1.2516, "step": 1309 }, { "epoch": 0.19, "eval_validation_loss": 0.8514130711555481, "eval_validation_runtime": 194.0021, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 1309 }, { "epoch": 0.19, "eval_validation_privacy_sources_loss": 0.8514130711555481, "eval_validation_privacy_sources_runtime": 193.9274, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 1309 }, { "epoch": 0.19, "eval_validation_agenda_digitale_loss": 1.479090929031372, "eval_validation_agenda_digitale_runtime": 174.0783, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.73, "step": 1309 }, { "epoch": 0.19, "eval_validation_legal_articles_loss": 1.375425100326538, "eval_validation_legal_articles_runtime": 280.3589, "eval_validation_legal_articles_samples_per_second": 1.459, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 1309 }, { "epoch": 0.19, "eval_validation_dolma_loss": 1.8478556871414185, "eval_validation_dolma_runtime": 111.0439, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.729, "step": 1309 }, { "epoch": 0.19, "eval_validation_leggepertutti_loss": 1.2495064735412598, "eval_validation_leggepertutti_runtime": 32.2623, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 1309 }, { "epoch": 0.19, "eval_validation_stack_loss": 0.6663044095039368, "eval_validation_stack_runtime": 150.1438, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 1309 } ], "logging_steps": 187, "max_steps": 13994, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 187, "total_flos": 7.319994288568271e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }