{"data":{"kind":"file","path":"README.md","version_id":"wr9vrr8u9i871kg5z4jakn75","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":1472,"modified_at":"2025-09-06T15:19:01.266000","content_hash":"aa1912d3aca241818f16fd4159124960b324b6e6669262408f9ad837accdb6ea"},"entries":[],"content":"# reverse-text\n\n### Overview\n- **Environment ID**: `reverse-text`\n- **Short description**: Reverse a given paragraph; evaluated by LCS similarity to the exact reversal.\n- **Tags**: text, transformation, single-turn, xml\n\n### Datasets\n- **Primary dataset(s)**: `agentlans/wikipedia-paragraphs` mapped to question/answer pairs\n- **Source links**: Hugging Face Datasets\n- **Split sizes**: Train/eval split controlled by `num_train_examples` and `num_eval_examples`\n\n### Task\n- **Type**: single-turn\n- **Parser**: `XMLParser([\"think\",\"answer\"])`\n- **Rubric overview**: LCS similarity between parsed answer and ground-truth reversed text; optional format check\n\n### Quickstart\nRun an evaluation with default settings:\n\n```bash\nuv run vf-eval reverse-text\n```\n\nConfigure model and sampling:\n\n```bash\nuv run vf-eval reverse-text \\\n  -m gpt-4.1-mini \\\n  -n 20 -r 3 -t 1024 -T 0.7 \\\n  -a '{\"num_train_examples\": 2000, \"num_eval_examples\": 200}'\n```\n\nNotes:\n- Use `-a` / `--env-args` to pass environment-specific configuration as a JSON object.\n\n### Environment Arguments\n| Arg | Type | Default | Description |\n| --- | ---- | ------- | ----------- |\n| `num_train_examples` | int | `2000` | Number of training examples |\n| `num_eval_examples` | int | `200` | Number of evaluation examples |\n\n### Metrics\n| Metric | Meaning |\n| ------ | ------- |\n| `reward` | LCS similarity between reversed text and parsed answer |\n| `format_reward` | Adherence to `<think>`/`<answer>` XML format |\n","encoding":"utf-8","truncated":false,"total_bytes":1472},"status":null}