{"data":{"kind":"file","path":"README.md","version_id":"nu7hdbxvt7a2prgsi3dmm21e","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":1390,"modified_at":"2026-02-15T13:29:32.332000","content_hash":"70ca6e37d3b937bdeaaae447a218e3383da015ee8b22b139ccf03d511a6e3032"},"entries":[],"content":"# OpenMed LitCovid\n\nMulti-label biomedical article topic classification using the LitCovid BioCreative VII dataset.\n\n## Task\n\nGiven a biomedical article's title and abstract, predict which research topic labels apply. Articles can have multiple labels simultaneously.\n\n**Format**: Multi-label classification with `<answer></answer>` output tags\n\n## Dataset\n\n- **Source**: [KushT/LitCovid_BioCreative](https://huggingface.co/datasets/KushT/LitCovid_BioCreative) (Apache 2.0)\n- **Size**: 33.7K PubMed articles (25K train / 2.5K val / 6.2K test)\n- **Labels**: Treatment, Diagnosis, Prevention, Mechanism, Transmission, Epidemic Forecasting, Case Report\n\n## Reward Structure\n\n| Component | Weight | Description |\n|-----------|--------|-------------|\n| F1 Score | 0.40 | Set-level F1 between predicted and expected labels |\n| Recall | 0.25 | Partial credit for each correct topic found |\n| Thinking | 0.20 | Encourages literature analysis reasoning |\n| Format | 0.15 | Proper `<answer></answer>` tag usage |\n\n## Example\n\n**Input**: Title: \"Remdesivir for the Treatment of Covid-19\" / Abstract: \"A double-blind, randomized, placebo-controlled trial...\"\n\n**Output**:\n```\n<answer>\nTreatment\n</answer>\n```\n\n## Usage\n\n```toml\n[[env]]\nid = \"maziyar/OpenMed_LitCovid\"\n```\n\n## Citation\n\nChen Q, et al. LitCovid in 2022: an information resource for the COVID-19 literature. Nucleic Acids Research. 2023.\n","encoding":"utf-8","truncated":false,"total_bytes":1390},"status":null}