{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6005253374988001, "eval_steps": 187, "global_step": 4301, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 6.75, "learning_rate": 1e-05, "loss": 1.5273, "step": 1 }, { "epoch": 0.03, "grad_norm": 4.84375, "learning_rate": 1e-05, "loss": 1.4233, "step": 187 }, { "epoch": 0.03, "eval_validation_loss": 0.9953275918960571, "eval_validation_runtime": 194.4686, "eval_validation_samples_per_second": 1.455, "eval_validation_steps_per_second": 0.73, "step": 187 }, { "epoch": 0.03, "eval_validation_privacy_sources_loss": 0.9953275918960571, "eval_validation_privacy_sources_runtime": 194.1974, "eval_validation_privacy_sources_samples_per_second": 1.457, "eval_validation_privacy_sources_steps_per_second": 0.731, "step": 187 }, { "epoch": 0.03, "eval_validation_agenda_digitale_loss": 1.5556422472000122, "eval_validation_agenda_digitale_runtime": 174.2863, "eval_validation_agenda_digitale_samples_per_second": 1.457, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 187 }, { "epoch": 0.03, "eval_validation_legal_articles_loss": 1.4555308818817139, "eval_validation_legal_articles_runtime": 280.5049, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 187 }, { "epoch": 0.03, "eval_validation_dolma_loss": 1.8503074645996094, "eval_validation_dolma_runtime": 111.0987, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 187 }, { "epoch": 0.03, "eval_validation_leggepertutti_loss": 1.3198411464691162, "eval_validation_leggepertutti_runtime": 32.2433, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 187 }, { "epoch": 0.03, "eval_validation_stack_loss": 0.7054847478866577, "eval_validation_stack_runtime": 150.1631, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.733, "step": 187 }, { "epoch": 0.05, "grad_norm": 3.21875, "learning_rate": 1e-05, "loss": 1.36, "step": 374 }, { "epoch": 0.05, "eval_validation_loss": 0.9383934140205383, "eval_validation_runtime": 194.1055, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 374 }, { "epoch": 0.05, "eval_validation_privacy_sources_loss": 0.9383934140205383, "eval_validation_privacy_sources_runtime": 194.049, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 374 }, { "epoch": 0.05, "eval_validation_agenda_digitale_loss": 1.5316377878189087, "eval_validation_agenda_digitale_runtime": 174.1261, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 374 }, { "epoch": 0.05, "eval_validation_legal_articles_loss": 1.4261502027511597, "eval_validation_legal_articles_runtime": 280.4336, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 374 }, { "epoch": 0.05, "eval_validation_dolma_loss": 1.8504927158355713, "eval_validation_dolma_runtime": 111.0511, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.729, "step": 374 }, { "epoch": 0.05, "eval_validation_leggepertutti_loss": 1.2899781465530396, "eval_validation_leggepertutti_runtime": 32.2352, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.745, "step": 374 }, { "epoch": 0.05, "eval_validation_stack_loss": 0.6623784303665161, "eval_validation_stack_runtime": 150.1338, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 374 }, { "epoch": 0.08, "grad_norm": 2.984375, "learning_rate": 1e-05, "loss": 1.335, "step": 561 }, { "epoch": 0.08, "eval_validation_loss": 0.9139429926872253, "eval_validation_runtime": 194.0822, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 561 }, { "epoch": 0.08, "eval_validation_privacy_sources_loss": 0.9139429926872253, "eval_validation_privacy_sources_runtime": 194.1531, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.731, "step": 561 }, { "epoch": 0.08, "eval_validation_agenda_digitale_loss": 1.5203362703323364, "eval_validation_agenda_digitale_runtime": 174.1785, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 561 }, { "epoch": 0.08, "eval_validation_legal_articles_loss": 1.4142487049102783, "eval_validation_legal_articles_runtime": 280.7608, "eval_validation_legal_articles_samples_per_second": 1.457, "eval_validation_legal_articles_steps_per_second": 0.73, "step": 561 }, { "epoch": 0.08, "eval_validation_dolma_loss": 1.8510735034942627, "eval_validation_dolma_runtime": 111.1069, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 561 }, { "epoch": 0.08, "eval_validation_leggepertutti_loss": 1.2792346477508545, "eval_validation_leggepertutti_runtime": 32.2763, "eval_validation_leggepertutti_samples_per_second": 1.456, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 561 }, { "epoch": 0.08, "eval_validation_stack_loss": 0.6663106083869934, "eval_validation_stack_runtime": 150.2829, "eval_validation_stack_samples_per_second": 1.457, "eval_validation_stack_steps_per_second": 0.732, "step": 561 }, { "epoch": 0.1, "grad_norm": 3.34375, "learning_rate": 1e-05, "loss": 1.3227, "step": 748 }, { "epoch": 0.1, "eval_validation_loss": 0.8906309008598328, "eval_validation_runtime": 193.9915, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 748 }, { "epoch": 0.1, "eval_validation_privacy_sources_loss": 0.8906309008598328, "eval_validation_privacy_sources_runtime": 193.9118, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 748 }, { "epoch": 0.1, "eval_validation_agenda_digitale_loss": 1.505711555480957, "eval_validation_agenda_digitale_runtime": 174.1465, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 748 }, { "epoch": 0.1, "eval_validation_legal_articles_loss": 1.4007651805877686, "eval_validation_legal_articles_runtime": 280.6207, "eval_validation_legal_articles_samples_per_second": 1.457, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 748 }, { "epoch": 0.1, "eval_validation_dolma_loss": 1.8508217334747314, "eval_validation_dolma_runtime": 111.0428, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.729, "step": 748 }, { "epoch": 0.1, "eval_validation_leggepertutti_loss": 1.2712979316711426, "eval_validation_leggepertutti_runtime": 32.243, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 748 }, { "epoch": 0.1, "eval_validation_stack_loss": 0.6624295711517334, "eval_validation_stack_runtime": 150.1684, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.733, "step": 748 }, { "epoch": 0.13, "grad_norm": 3.1875, "learning_rate": 1e-05, "loss": 1.2894, "step": 935 }, { "epoch": 0.13, "eval_validation_loss": 0.8783557415008545, "eval_validation_runtime": 194.0651, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 935 }, { "epoch": 0.13, "eval_validation_privacy_sources_loss": 0.8783557415008545, "eval_validation_privacy_sources_runtime": 194.2156, "eval_validation_privacy_sources_samples_per_second": 1.457, "eval_validation_privacy_sources_steps_per_second": 0.731, "step": 935 }, { "epoch": 0.13, "eval_validation_agenda_digitale_loss": 1.4970660209655762, "eval_validation_agenda_digitale_runtime": 174.2301, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 935 }, { "epoch": 0.13, "eval_validation_legal_articles_loss": 1.3921587467193604, "eval_validation_legal_articles_runtime": 280.4294, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 935 }, { "epoch": 0.13, "eval_validation_dolma_loss": 1.851251244544983, "eval_validation_dolma_runtime": 111.0768, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 935 }, { "epoch": 0.13, "eval_validation_leggepertutti_loss": 1.2646007537841797, "eval_validation_leggepertutti_runtime": 32.2511, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 935 }, { "epoch": 0.13, "eval_validation_stack_loss": 0.667582631111145, "eval_validation_stack_runtime": 150.215, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 935 }, { "epoch": 0.16, "grad_norm": 6.3125, "learning_rate": 1e-05, "loss": 1.2936, "step": 1122 }, { "epoch": 0.16, "eval_validation_loss": 0.8658438920974731, "eval_validation_runtime": 193.9819, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 1122 }, { "epoch": 0.16, "eval_validation_privacy_sources_loss": 0.8658438920974731, "eval_validation_privacy_sources_runtime": 193.9777, "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": 1.4888020753860474, "eval_validation_agenda_digitale_runtime": 174.0686, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.73, "step": 1122 }, { "epoch": 0.16, "eval_validation_legal_articles_loss": 1.384485125541687, "eval_validation_legal_articles_runtime": 280.3377, "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": 1.855146050453186, "eval_validation_dolma_runtime": 110.9957, "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": 1.2589794397354126, "eval_validation_leggepertutti_runtime": 32.2366, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 1122 }, { "epoch": 0.16, "eval_validation_stack_loss": 0.6629025340080261, "eval_validation_stack_runtime": 150.1177, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 1122 }, { "epoch": 0.18, "grad_norm": 3.25, "learning_rate": 1e-05, "loss": 1.2703, "step": 1309 }, { "epoch": 0.18, "eval_validation_loss": 0.8527970910072327, "eval_validation_runtime": 193.9817, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 1309 }, { "epoch": 0.18, "eval_validation_privacy_sources_loss": 0.8527970910072327, "eval_validation_privacy_sources_runtime": 194.001, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 1309 }, { "epoch": 0.18, "eval_validation_agenda_digitale_loss": 1.4821778535842896, "eval_validation_agenda_digitale_runtime": 174.0669, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.73, "step": 1309 }, { "epoch": 0.18, "eval_validation_legal_articles_loss": 1.3752673864364624, "eval_validation_legal_articles_runtime": 280.3145, "eval_validation_legal_articles_samples_per_second": 1.459, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 1309 }, { "epoch": 0.18, "eval_validation_dolma_loss": 1.851420283317566, "eval_validation_dolma_runtime": 111.0747, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 1309 }, { "epoch": 0.18, "eval_validation_leggepertutti_loss": 1.2481449842453003, "eval_validation_leggepertutti_runtime": 32.2435, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 1309 }, { "epoch": 0.18, "eval_validation_stack_loss": 0.663459062576294, "eval_validation_stack_runtime": 150.3253, "eval_validation_stack_samples_per_second": 1.457, "eval_validation_stack_steps_per_second": 0.732, "step": 1309 }, { "epoch": 0.21, "grad_norm": 3.3125, "learning_rate": 1e-05, "loss": 1.2603, "step": 1496 }, { "epoch": 0.21, "eval_validation_loss": 0.8457313776016235, "eval_validation_runtime": 194.1045, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 1496 }, { "epoch": 0.21, "eval_validation_privacy_sources_loss": 0.8457313776016235, "eval_validation_privacy_sources_runtime": 194.0987, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 1496 }, { "epoch": 0.21, "eval_validation_agenda_digitale_loss": 1.4775303602218628, "eval_validation_agenda_digitale_runtime": 174.2206, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 1496 }, { "epoch": 0.21, "eval_validation_legal_articles_loss": 1.3696094751358032, "eval_validation_legal_articles_runtime": 280.5342, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 1496 }, { "epoch": 0.21, "eval_validation_dolma_loss": 1.8503979444503784, "eval_validation_dolma_runtime": 111.072, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.729, "step": 1496 }, { "epoch": 0.21, "eval_validation_leggepertutti_loss": 1.2421447038650513, "eval_validation_leggepertutti_runtime": 32.2684, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 1496 }, { "epoch": 0.21, "eval_validation_stack_loss": 0.6622874140739441, "eval_validation_stack_runtime": 150.2083, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 1496 }, { "epoch": 0.23, "grad_norm": 2.78125, "learning_rate": 1e-05, "loss": 1.2556, "step": 1683 }, { "epoch": 0.23, "eval_validation_loss": 0.8370487689971924, "eval_validation_runtime": 194.2283, "eval_validation_samples_per_second": 1.457, "eval_validation_steps_per_second": 0.731, "step": 1683 }, { "epoch": 0.23, "eval_validation_privacy_sources_loss": 0.8370487689971924, "eval_validation_privacy_sources_runtime": 194.3184, "eval_validation_privacy_sources_samples_per_second": 1.456, "eval_validation_privacy_sources_steps_per_second": 0.731, "step": 1683 }, { "epoch": 0.23, "eval_validation_agenda_digitale_loss": 1.4706999063491821, "eval_validation_agenda_digitale_runtime": 174.3662, "eval_validation_agenda_digitale_samples_per_second": 1.457, "eval_validation_agenda_digitale_steps_per_second": 0.728, "step": 1683 }, { "epoch": 0.23, "eval_validation_legal_articles_loss": 1.363641381263733, "eval_validation_legal_articles_runtime": 280.656, "eval_validation_legal_articles_samples_per_second": 1.457, "eval_validation_legal_articles_steps_per_second": 0.73, "step": 1683 }, { "epoch": 0.23, "eval_validation_dolma_loss": 1.848027229309082, "eval_validation_dolma_runtime": 111.1824, "eval_validation_dolma_samples_per_second": 1.457, "eval_validation_dolma_steps_per_second": 0.729, "step": 1683 }, { "epoch": 0.23, "eval_validation_leggepertutti_loss": 1.240347146987915, "eval_validation_leggepertutti_runtime": 32.2674, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 1683 }, { "epoch": 0.23, "eval_validation_stack_loss": 0.6605746746063232, "eval_validation_stack_runtime": 150.2441, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 1683 }, { "epoch": 0.26, "grad_norm": 3.1875, "learning_rate": 1e-05, "loss": 1.25, "step": 1870 }, { "epoch": 0.26, "eval_validation_loss": 0.8285319805145264, "eval_validation_runtime": 194.0857, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 1870 }, { "epoch": 0.26, "eval_validation_privacy_sources_loss": 0.8285319805145264, "eval_validation_privacy_sources_runtime": 194.1082, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 1870 }, { "epoch": 0.26, "eval_validation_agenda_digitale_loss": 1.467832088470459, "eval_validation_agenda_digitale_runtime": 174.0948, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 1870 }, { "epoch": 0.26, "eval_validation_legal_articles_loss": 1.3606085777282715, "eval_validation_legal_articles_runtime": 280.4529, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 1870 }, { "epoch": 0.26, "eval_validation_dolma_loss": 1.8478673696517944, "eval_validation_dolma_runtime": 111.0459, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.729, "step": 1870 }, { "epoch": 0.26, "eval_validation_leggepertutti_loss": 1.2359322309494019, "eval_validation_leggepertutti_runtime": 32.2397, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 1870 }, { "epoch": 0.26, "eval_validation_stack_loss": 0.6594959497451782, "eval_validation_stack_runtime": 150.1577, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.733, "step": 1870 }, { "epoch": 0.29, "grad_norm": 4.96875, "learning_rate": 1e-05, "loss": 1.238, "step": 2057 }, { "epoch": 0.29, "eval_validation_loss": 0.8220862150192261, "eval_validation_runtime": 194.0761, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 2057 }, { "epoch": 0.29, "eval_validation_privacy_sources_loss": 0.8220862150192261, "eval_validation_privacy_sources_runtime": 194.0716, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 2057 }, { "epoch": 0.29, "eval_validation_agenda_digitale_loss": 1.4595245122909546, "eval_validation_agenda_digitale_runtime": 174.1501, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 2057 }, { "epoch": 0.29, "eval_validation_legal_articles_loss": 1.355904221534729, "eval_validation_legal_articles_runtime": 280.5038, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 2057 }, { "epoch": 0.29, "eval_validation_dolma_loss": 1.846190333366394, "eval_validation_dolma_runtime": 111.1135, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 2057 }, { "epoch": 0.29, "eval_validation_leggepertutti_loss": 1.2381559610366821, "eval_validation_leggepertutti_runtime": 32.2445, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 2057 }, { "epoch": 0.29, "eval_validation_stack_loss": 0.660571813583374, "eval_validation_stack_runtime": 150.2476, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 2057 }, { "epoch": 0.31, "grad_norm": 9.0, "learning_rate": 1e-05, "loss": 1.2384, "step": 2244 }, { "epoch": 0.31, "eval_validation_loss": 0.8178092241287231, "eval_validation_runtime": 194.1251, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.731, "step": 2244 }, { "epoch": 0.31, "eval_validation_privacy_sources_loss": 0.8178092241287231, "eval_validation_privacy_sources_runtime": 194.0771, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 2244 }, { "epoch": 0.31, "eval_validation_agenda_digitale_loss": 1.460057258605957, "eval_validation_agenda_digitale_runtime": 174.2114, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 2244 }, { "epoch": 0.31, "eval_validation_legal_articles_loss": 1.3545092344284058, "eval_validation_legal_articles_runtime": 280.7055, "eval_validation_legal_articles_samples_per_second": 1.457, "eval_validation_legal_articles_steps_per_second": 0.73, "step": 2244 }, { "epoch": 0.31, "eval_validation_dolma_loss": 1.8475868701934814, "eval_validation_dolma_runtime": 111.0986, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 2244 }, { "epoch": 0.31, "eval_validation_leggepertutti_loss": 1.2337557077407837, "eval_validation_leggepertutti_runtime": 32.2593, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 2244 }, { "epoch": 0.31, "eval_validation_stack_loss": 0.6590169072151184, "eval_validation_stack_runtime": 150.2183, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 2244 }, { "epoch": 0.34, "grad_norm": 2.96875, "learning_rate": 1e-05, "loss": 1.2267, "step": 2431 }, { "epoch": 0.34, "eval_validation_loss": 0.8118987083435059, "eval_validation_runtime": 194.1905, "eval_validation_samples_per_second": 1.457, "eval_validation_steps_per_second": 0.731, "step": 2431 }, { "epoch": 0.34, "eval_validation_privacy_sources_loss": 0.8118987083435059, "eval_validation_privacy_sources_runtime": 194.1644, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.731, "step": 2431 }, { "epoch": 0.34, "eval_validation_agenda_digitale_loss": 1.4531759023666382, "eval_validation_agenda_digitale_runtime": 174.2809, "eval_validation_agenda_digitale_samples_per_second": 1.457, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 2431 }, { "epoch": 0.34, "eval_validation_legal_articles_loss": 1.3473464250564575, "eval_validation_legal_articles_runtime": 280.6295, "eval_validation_legal_articles_samples_per_second": 1.457, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 2431 }, { "epoch": 0.34, "eval_validation_dolma_loss": 1.8490244150161743, "eval_validation_dolma_runtime": 111.164, "eval_validation_dolma_samples_per_second": 1.457, "eval_validation_dolma_steps_per_second": 0.729, "step": 2431 }, { "epoch": 0.34, "eval_validation_leggepertutti_loss": 1.2278121709823608, "eval_validation_leggepertutti_runtime": 32.2528, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 2431 }, { "epoch": 0.34, "eval_validation_stack_loss": 0.6574946641921997, "eval_validation_stack_runtime": 150.229, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 2431 }, { "epoch": 0.37, "grad_norm": 7.125, "learning_rate": 1e-05, "loss": 1.228, "step": 2618 }, { "epoch": 0.37, "eval_validation_loss": 0.8083770275115967, "eval_validation_runtime": 194.1727, "eval_validation_samples_per_second": 1.457, "eval_validation_steps_per_second": 0.731, "step": 2618 }, { "epoch": 0.37, "eval_validation_privacy_sources_loss": 0.8083770275115967, "eval_validation_privacy_sources_runtime": 194.109, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 2618 }, { "epoch": 0.37, "eval_validation_agenda_digitale_loss": 1.4517344236373901, "eval_validation_agenda_digitale_runtime": 174.2026, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 2618 }, { "epoch": 0.37, "eval_validation_legal_articles_loss": 1.3457854986190796, "eval_validation_legal_articles_runtime": 280.5077, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 2618 }, { "epoch": 0.37, "eval_validation_dolma_loss": 1.8457980155944824, "eval_validation_dolma_runtime": 111.0983, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 2618 }, { "epoch": 0.37, "eval_validation_leggepertutti_loss": 1.2257617712020874, "eval_validation_leggepertutti_runtime": 32.2598, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 2618 }, { "epoch": 0.37, "eval_validation_stack_loss": 0.6570916175842285, "eval_validation_stack_runtime": 150.252, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 2618 }, { "epoch": 0.39, "grad_norm": 2.953125, "learning_rate": 1e-05, "loss": 1.2256, "step": 2805 }, { "epoch": 0.39, "eval_validation_loss": 0.8021377325057983, "eval_validation_runtime": 194.2469, "eval_validation_samples_per_second": 1.457, "eval_validation_steps_per_second": 0.731, "step": 2805 }, { "epoch": 0.39, "eval_validation_privacy_sources_loss": 0.8021377325057983, "eval_validation_privacy_sources_runtime": 194.4649, "eval_validation_privacy_sources_samples_per_second": 1.455, "eval_validation_privacy_sources_steps_per_second": 0.73, "step": 2805 }, { "epoch": 0.39, "eval_validation_agenda_digitale_loss": 1.449271559715271, "eval_validation_agenda_digitale_runtime": 174.4445, "eval_validation_agenda_digitale_samples_per_second": 1.456, "eval_validation_agenda_digitale_steps_per_second": 0.728, "step": 2805 }, { "epoch": 0.39, "eval_validation_legal_articles_loss": 1.3434243202209473, "eval_validation_legal_articles_runtime": 281.0216, "eval_validation_legal_articles_samples_per_second": 1.455, "eval_validation_legal_articles_steps_per_second": 0.729, "step": 2805 }, { "epoch": 0.39, "eval_validation_dolma_loss": 1.8447160720825195, "eval_validation_dolma_runtime": 111.2369, "eval_validation_dolma_samples_per_second": 1.456, "eval_validation_dolma_steps_per_second": 0.728, "step": 2805 }, { "epoch": 0.39, "eval_validation_leggepertutti_loss": 1.225563883781433, "eval_validation_leggepertutti_runtime": 32.3051, "eval_validation_leggepertutti_samples_per_second": 1.455, "eval_validation_leggepertutti_steps_per_second": 0.743, "step": 2805 }, { "epoch": 0.39, "eval_validation_stack_loss": 0.6570674180984497, "eval_validation_stack_runtime": 150.4367, "eval_validation_stack_samples_per_second": 1.456, "eval_validation_stack_steps_per_second": 0.731, "step": 2805 }, { "epoch": 0.42, "grad_norm": 3.28125, "learning_rate": 1e-05, "loss": 1.2136, "step": 2992 }, { "epoch": 0.42, "eval_validation_loss": 0.8003722429275513, "eval_validation_runtime": 194.1597, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.731, "step": 2992 }, { "epoch": 0.42, "eval_validation_privacy_sources_loss": 0.8003722429275513, "eval_validation_privacy_sources_runtime": 194.2144, "eval_validation_privacy_sources_samples_per_second": 1.457, "eval_validation_privacy_sources_steps_per_second": 0.731, "step": 2992 }, { "epoch": 0.42, "eval_validation_agenda_digitale_loss": 1.445157766342163, "eval_validation_agenda_digitale_runtime": 174.3593, "eval_validation_agenda_digitale_samples_per_second": 1.457, "eval_validation_agenda_digitale_steps_per_second": 0.728, "step": 2992 }, { "epoch": 0.42, "eval_validation_legal_articles_loss": 1.340084433555603, "eval_validation_legal_articles_runtime": 280.628, "eval_validation_legal_articles_samples_per_second": 1.457, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 2992 }, { "epoch": 0.42, "eval_validation_dolma_loss": 1.8441739082336426, "eval_validation_dolma_runtime": 111.1155, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 2992 }, { "epoch": 0.42, "eval_validation_leggepertutti_loss": 1.2245877981185913, "eval_validation_leggepertutti_runtime": 32.2646, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 2992 }, { "epoch": 0.42, "eval_validation_stack_loss": 0.657370388507843, "eval_validation_stack_runtime": 150.3713, "eval_validation_stack_samples_per_second": 1.456, "eval_validation_stack_steps_per_second": 0.732, "step": 2992 }, { "epoch": 0.44, "grad_norm": 2.8125, "learning_rate": 1e-05, "loss": 1.2216, "step": 3179 }, { "epoch": 0.44, "eval_validation_loss": 0.7953833341598511, "eval_validation_runtime": 194.5265, "eval_validation_samples_per_second": 1.455, "eval_validation_steps_per_second": 0.73, "step": 3179 }, { "epoch": 0.44, "eval_validation_privacy_sources_loss": 0.7953833341598511, "eval_validation_privacy_sources_runtime": 194.6007, "eval_validation_privacy_sources_samples_per_second": 1.454, "eval_validation_privacy_sources_steps_per_second": 0.73, "step": 3179 }, { "epoch": 0.44, "eval_validation_agenda_digitale_loss": 1.4409934282302856, "eval_validation_agenda_digitale_runtime": 174.5993, "eval_validation_agenda_digitale_samples_per_second": 1.455, "eval_validation_agenda_digitale_steps_per_second": 0.727, "step": 3179 }, { "epoch": 0.44, "eval_validation_legal_articles_loss": 1.3360470533370972, "eval_validation_legal_articles_runtime": 281.1454, "eval_validation_legal_articles_samples_per_second": 1.455, "eval_validation_legal_articles_steps_per_second": 0.729, "step": 3179 }, { "epoch": 0.44, "eval_validation_dolma_loss": 1.8454322814941406, "eval_validation_dolma_runtime": 111.309, "eval_validation_dolma_samples_per_second": 1.455, "eval_validation_dolma_steps_per_second": 0.728, "step": 3179 }, { "epoch": 0.44, "eval_validation_leggepertutti_loss": 1.2207506895065308, "eval_validation_leggepertutti_runtime": 32.3094, "eval_validation_leggepertutti_samples_per_second": 1.455, "eval_validation_leggepertutti_steps_per_second": 0.743, "step": 3179 }, { "epoch": 0.44, "eval_validation_stack_loss": 0.6579957604408264, "eval_validation_stack_runtime": 150.3599, "eval_validation_stack_samples_per_second": 1.457, "eval_validation_stack_steps_per_second": 0.732, "step": 3179 }, { "epoch": 0.47, "grad_norm": 3.3125, "learning_rate": 1e-05, "loss": 1.2183, "step": 3366 }, { "epoch": 0.47, "eval_validation_loss": 0.7924418449401855, "eval_validation_runtime": 194.1975, "eval_validation_samples_per_second": 1.457, "eval_validation_steps_per_second": 0.731, "step": 3366 }, { "epoch": 0.47, "eval_validation_privacy_sources_loss": 0.7924418449401855, "eval_validation_privacy_sources_runtime": 194.1952, "eval_validation_privacy_sources_samples_per_second": 1.457, "eval_validation_privacy_sources_steps_per_second": 0.731, "step": 3366 }, { "epoch": 0.47, "eval_validation_agenda_digitale_loss": 1.4390149116516113, "eval_validation_agenda_digitale_runtime": 174.2539, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 3366 }, { "epoch": 0.47, "eval_validation_legal_articles_loss": 1.3332501649856567, "eval_validation_legal_articles_runtime": 280.6326, "eval_validation_legal_articles_samples_per_second": 1.457, "eval_validation_legal_articles_steps_per_second": 0.73, "step": 3366 }, { "epoch": 0.47, "eval_validation_dolma_loss": 1.848328709602356, "eval_validation_dolma_runtime": 111.0824, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 3366 }, { "epoch": 0.47, "eval_validation_leggepertutti_loss": 1.2147657871246338, "eval_validation_leggepertutti_runtime": 32.2807, "eval_validation_leggepertutti_samples_per_second": 1.456, "eval_validation_leggepertutti_steps_per_second": 0.743, "step": 3366 }, { "epoch": 0.47, "eval_validation_stack_loss": 0.6573349833488464, "eval_validation_stack_runtime": 150.3498, "eval_validation_stack_samples_per_second": 1.457, "eval_validation_stack_steps_per_second": 0.732, "step": 3366 }, { "epoch": 0.5, "grad_norm": 3.140625, "learning_rate": 1e-05, "loss": 1.217, "step": 3553 }, { "epoch": 0.5, "eval_validation_loss": 0.7858894467353821, "eval_validation_runtime": 194.4212, "eval_validation_samples_per_second": 1.456, "eval_validation_steps_per_second": 0.73, "step": 3553 }, { "epoch": 0.5, "eval_validation_privacy_sources_loss": 0.7858894467353821, "eval_validation_privacy_sources_runtime": 194.417, "eval_validation_privacy_sources_samples_per_second": 1.456, "eval_validation_privacy_sources_steps_per_second": 0.73, "step": 3553 }, { "epoch": 0.5, "eval_validation_agenda_digitale_loss": 1.4353201389312744, "eval_validation_agenda_digitale_runtime": 174.2645, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 3553 }, { "epoch": 0.5, "eval_validation_legal_articles_loss": 1.329630732536316, "eval_validation_legal_articles_runtime": 280.8801, "eval_validation_legal_articles_samples_per_second": 1.456, "eval_validation_legal_articles_steps_per_second": 0.73, "step": 3553 }, { "epoch": 0.5, "eval_validation_dolma_loss": 1.847051978111267, "eval_validation_dolma_runtime": 111.2168, "eval_validation_dolma_samples_per_second": 1.457, "eval_validation_dolma_steps_per_second": 0.728, "step": 3553 }, { "epoch": 0.5, "eval_validation_leggepertutti_loss": 1.2128466367721558, "eval_validation_leggepertutti_runtime": 32.2731, "eval_validation_leggepertutti_samples_per_second": 1.456, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 3553 }, { "epoch": 0.5, "eval_validation_stack_loss": 0.660595715045929, "eval_validation_stack_runtime": 150.4122, "eval_validation_stack_samples_per_second": 1.456, "eval_validation_stack_steps_per_second": 0.731, "step": 3553 }, { "epoch": 0.52, "grad_norm": 4.59375, "learning_rate": 1e-05, "loss": 1.2025, "step": 3740 }, { "epoch": 0.52, "eval_validation_loss": 0.7821725606918335, "eval_validation_runtime": 194.102, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 3740 }, { "epoch": 0.52, "eval_validation_privacy_sources_loss": 0.7821725606918335, "eval_validation_privacy_sources_runtime": 194.0576, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 3740 }, { "epoch": 0.52, "eval_validation_agenda_digitale_loss": 1.4339957237243652, "eval_validation_agenda_digitale_runtime": 174.1039, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 3740 }, { "epoch": 0.52, "eval_validation_legal_articles_loss": 1.3280164003372192, "eval_validation_legal_articles_runtime": 280.5513, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 3740 }, { "epoch": 0.52, "eval_validation_dolma_loss": 1.84745454788208, "eval_validation_dolma_runtime": 111.0532, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.729, "step": 3740 }, { "epoch": 0.52, "eval_validation_leggepertutti_loss": 1.2140834331512451, "eval_validation_leggepertutti_runtime": 32.2501, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 3740 }, { "epoch": 0.52, "eval_validation_stack_loss": 0.6579498052597046, "eval_validation_stack_runtime": 150.177, "eval_validation_stack_samples_per_second": 1.458, "eval_validation_stack_steps_per_second": 0.732, "step": 3740 }, { "epoch": 0.55, "grad_norm": 2.90625, "learning_rate": 1e-05, "loss": 1.1892, "step": 3927 }, { "epoch": 0.55, "eval_validation_loss": 0.778718888759613, "eval_validation_runtime": 194.0252, "eval_validation_samples_per_second": 1.459, "eval_validation_steps_per_second": 0.732, "step": 3927 }, { "epoch": 0.55, "eval_validation_privacy_sources_loss": 0.778718888759613, "eval_validation_privacy_sources_runtime": 194.0094, "eval_validation_privacy_sources_samples_per_second": 1.459, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 3927 }, { "epoch": 0.55, "eval_validation_agenda_digitale_loss": 1.4309070110321045, "eval_validation_agenda_digitale_runtime": 174.153, "eval_validation_agenda_digitale_samples_per_second": 1.458, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 3927 }, { "epoch": 0.55, "eval_validation_legal_articles_loss": 1.3250170946121216, "eval_validation_legal_articles_runtime": 280.4271, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 3927 }, { "epoch": 0.55, "eval_validation_dolma_loss": 1.8441972732543945, "eval_validation_dolma_runtime": 111.0278, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.73, "step": 3927 }, { "epoch": 0.55, "eval_validation_leggepertutti_loss": 1.2091401815414429, "eval_validation_leggepertutti_runtime": 32.2467, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 3927 }, { "epoch": 0.55, "eval_validation_stack_loss": 0.6560612916946411, "eval_validation_stack_runtime": 150.1302, "eval_validation_stack_samples_per_second": 1.459, "eval_validation_stack_steps_per_second": 0.733, "step": 3927 }, { "epoch": 0.57, "grad_norm": 2.890625, "learning_rate": 1e-05, "loss": 1.1884, "step": 4114 }, { "epoch": 0.57, "eval_validation_loss": 0.775921106338501, "eval_validation_runtime": 194.1411, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.731, "step": 4114 }, { "epoch": 0.57, "eval_validation_privacy_sources_loss": 0.775921106338501, "eval_validation_privacy_sources_runtime": 194.0466, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 4114 }, { "epoch": 0.57, "eval_validation_agenda_digitale_loss": 1.4286702871322632, "eval_validation_agenda_digitale_runtime": 174.1174, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 4114 }, { "epoch": 0.57, "eval_validation_legal_articles_loss": 1.3244529962539673, "eval_validation_legal_articles_runtime": 280.4889, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 4114 }, { "epoch": 0.57, "eval_validation_dolma_loss": 1.8429710865020752, "eval_validation_dolma_runtime": 111.0663, "eval_validation_dolma_samples_per_second": 1.459, "eval_validation_dolma_steps_per_second": 0.729, "step": 4114 }, { "epoch": 0.57, "eval_validation_leggepertutti_loss": 1.2103228569030762, "eval_validation_leggepertutti_runtime": 32.2408, "eval_validation_leggepertutti_samples_per_second": 1.458, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 4114 }, { "epoch": 0.57, "eval_validation_stack_loss": 0.6574613451957703, "eval_validation_stack_runtime": 150.2725, "eval_validation_stack_samples_per_second": 1.457, "eval_validation_stack_steps_per_second": 0.732, "step": 4114 }, { "epoch": 0.6, "grad_norm": 3.328125, "learning_rate": 1e-05, "loss": 1.216, "step": 4301 }, { "epoch": 0.6, "eval_validation_loss": 0.7726068496704102, "eval_validation_runtime": 194.1083, "eval_validation_samples_per_second": 1.458, "eval_validation_steps_per_second": 0.732, "step": 4301 }, { "epoch": 0.6, "eval_validation_privacy_sources_loss": 0.7726068496704102, "eval_validation_privacy_sources_runtime": 194.0844, "eval_validation_privacy_sources_samples_per_second": 1.458, "eval_validation_privacy_sources_steps_per_second": 0.732, "step": 4301 }, { "epoch": 0.6, "eval_validation_agenda_digitale_loss": 1.4294990301132202, "eval_validation_agenda_digitale_runtime": 174.0978, "eval_validation_agenda_digitale_samples_per_second": 1.459, "eval_validation_agenda_digitale_steps_per_second": 0.729, "step": 4301 }, { "epoch": 0.6, "eval_validation_legal_articles_loss": 1.3229315280914307, "eval_validation_legal_articles_runtime": 280.5081, "eval_validation_legal_articles_samples_per_second": 1.458, "eval_validation_legal_articles_steps_per_second": 0.731, "step": 4301 }, { "epoch": 0.6, "eval_validation_dolma_loss": 1.844489574432373, "eval_validation_dolma_runtime": 111.0784, "eval_validation_dolma_samples_per_second": 1.458, "eval_validation_dolma_steps_per_second": 0.729, "step": 4301 }, { "epoch": 0.6, "eval_validation_leggepertutti_loss": 1.2054811716079712, "eval_validation_leggepertutti_runtime": 32.2642, "eval_validation_leggepertutti_samples_per_second": 1.457, "eval_validation_leggepertutti_steps_per_second": 0.744, "step": 4301 }, { "epoch": 0.6, "eval_validation_stack_loss": 0.657507061958313, "eval_validation_stack_runtime": 150.2647, "eval_validation_stack_samples_per_second": 1.457, "eval_validation_stack_steps_per_second": 0.732, "step": 4301 } ], "logging_steps": 187, "max_steps": 14324, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 187, "total_flos": 2.4051409805295747e+19, "train_batch_size": 1, "trial_name": null, "trial_params": null }