| { | |
| "generated_at": "2025-12-04T17:12:40.045804Z", | |
| "stage": "extract", | |
| "dataset": { | |
| "name": "HuggingFaceM4/FineVision", | |
| "config": "olmOCR-mix-0225-documents", | |
| "split": "train" | |
| }, | |
| "model": { | |
| "served_model_name": "deepseek-ocr", | |
| "prompt": "<image>\n<|grounding|>Convert this document to Markdown.", | |
| "max_tokens": 4096, | |
| "temperature": 0.4 | |
| }, | |
| "inference": { | |
| "max_batch_size": 64, | |
| "max_concurrency": 4, | |
| "request_timeout": 120, | |
| "max_retries": 3, | |
| "retry_backoff_seconds": 2.0, | |
| "max_retry_wait_seconds": 60.0 | |
| }, | |
| "documents": [], | |
| "documents_path": "document_batches", | |
| "documents_batches": [ | |
| "docs/document_batches/batch_00000.json" | |
| ], | |
| "documents_count": 10, | |
| "failures": [] | |
| } |