{"data":{"kind":"file","path":"README.md","version_id":"k2gh61d763s9uck2q6lcpydg","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":1806,"modified_at":"2026-02-06T11:15:15.108000","content_hash":"bfbff31e6487663c5a6093400bffd4928fec4c523b4e692a7659939f0fd5e5bd"},"entries":[],"content":"# wordle\n\n<a href=\"https://github.com/PrimeIntellect-ai/verifiers/tree/main/environments/wordle\">\n<img src=\"https://img.shields.io/badge/GitHub-181717?style=for-the-badge&logo=github&logoColor=white\" alt=\"Source Code\">\n</a>\n\n### Overview\n- **Environment ID**: `wordle`\n- **Short description**: Wordle game environment\n- **Tags**: games, train, eval, multi-turn, wordle\n\n### Datasets\n- **Primary dataset(s)**: TextArena `Wordle-v0` (environment provides episodes)\n- **Source links**: TextArena\n- **Split sizes**: Number of episodes controlled via args\n\n### Task\n- **Type**: multi-turn (game interaction)\n- **Rubric overview**: Exact guess match, partial credit from feedback, length bonus, and format check\n\n### Quickstart\nRun an evaluation with default settings:\n\n```bash\nprime eval run wordle\n```\n\nConfigure model and sampling:\n\n```bash\nprime eval run wordle \\\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\": 20}'\n```\n\n### Environment Arguments\n| Arg | Type | Default | Description |\n| --- | ---- | ------- | ----------- |\n| `num_train_examples` | int | `2000` | Number of training episodes |\n| `num_eval_examples` | int | `20` | Number of evaluation episodes |\n\n### Metrics\n| Metric | Meaning |\n| ------ | ------- |\n| `correct_answer` | 1.0 if final guess equals target, else 0.0 |\n| `partial_answer` | Partial credit from greens/yellows in feedback |\n| `length_bonus` | Higher score for solving in fewer turns |\n| `format_reward` | Adherence to expected XML format |\n\n### Changelog\n\n#### v0.1.6 (Dec 10, 2025)\n\n- Setup environment in `setup_state` instead of `env_response`\n- Fix checking game completion condition\n- Fix feedback parsing to correctly handle feedback on game completion and invalid guesses\n- Add logger\n- Rename reward functions\n","encoding":"utf-8","truncated":false,"total_bytes":1806},"status":null}