{ "checkpoint": { "bytes": 465294654, "config_sha256": "a9ba0bf41cfb5477a28494e2500585bbebaa65a9d19057cb48918ddf94a6d654", "data_cursor": 167494, "hugging_face_path": "full-training/sloppa-4b-stage1/checkpoints/step-00002000.pt", "model_state_kind": "trainable_parameters_v1", "model_tensors": 549, "optimizer_states": 549, "resume_contract_sha256": "7476342fbfacd6b9d333670646d23178523adb8e453e82e04a936ccb30085e1c", "sha256": "5a9ef596efaa5a9887be46d13dbc9af7fc2828536289b65073239eae949ee242", "source_tokens": 530125335, "step": 2000, "world_size": 4 }, "identity": { "backbone": "Qwen/Qwen3.5-4B", "backbone_revision": "851bf6e806efd8d0a36b00ddf55e13ccb7b8cd0a", "code_commit": "c52c07f0751f463c54f877f5868be31b55453cc1", "dataset_revision": "9ba8d28f6e5311788314c71326c16bfdcea75f57", "hugging_face_repository": "lordspline/sloppa-4b-training" }, "kind": "sloppa_stage1_progress", "observed": { "all_loss_and_gradient_values_finite": true, "checkpoint_written_at": "2026-08-04T00:42:05Z", "maximum_peak_allocated_bytes": 36318213120, "maximum_peak_reserved_bytes": 51443138560, "maximum_rank_source_token_imbalance": 761, "mean_gradient_reduction_seconds_steps_1001_2000": 4.0530867227807175, "model_tokens_per_second_steps_1001_2000": 15367.127049900993, "source_tokens_per_second_steps_1001_2000": 13780.748763623025, "steps_1951_2000_means": { "file_numeric_huber": 115.36209682938986, "file_ordinal_coral": 1.252781271645986, "gradient_norm_before_clipping": 7.660628485679626, "pair_ranking_bradley_terry": 0.16603217707655674, "total_loss": 5.046765271892137 }, "training_started_at": "2026-08-03T14:10:17Z" }, "schema_version": 1, "status": "in_progress", "training_state": { "best_metric": null, "best_step": null, "lora_gradient_check": { "nonzero_parameter_tensors": 496, "parameter_tensors": 496, "parameters": 32464896 }, "stopped_early": false, "validations_without_improvement": 0 }, "validation": { "fixed_partition_evaluation_completed": false, "online_losses_are_not_fixed_set_metrics": true } }