{"data":{"kind":"file","path":"README.md","version_id":"hqm1piyi9m7idbht9evdfq84","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":1845,"modified_at":"2025-09-07T21:44:21.051000","content_hash":"a1702e812ba4a5542ba51f08d0c88c3643cd50e2e7d63b3b40d709e9e10bb3bf"},"entries":[],"content":"# trading-bot-bench\n\n### Overview\n- **Environment ID**: `trading-bot-bench`\n- **Short description**: Single-turn python algorithmic trading script creation. \n- **Tags**: single-turn, trading, stocks, finance\n\n### Datasets\n- **Primary dataset**: paperswithbacktest/Stocks-Daily-Price (Hugging Face datasets)\n- **Source links**: https://huggingface.co/datasets/paperswithbacktest/Stocks-Daily-Price\n- **Split sizes**: 80/20 train/test split\n\n### Task\n- **Type**: single-turn\n- **Parser**: XMLParser\n- **Rubric overview**: The rubric combines return, risk-adjusted performance, and drawdown control into a single score between 0 and 1. It normalizes total return and Sharpe ratio to bounded values, applies a penalty for large drawdowns, and then weights them (45% return, 45% Sharpe, 10% drawdown).\n\n### Quickstart\nRun an evaluation with default settings:\n\n```bash\nuv run vf-eval trading-bot-bench\n```\n\nConfigure model and sampling:\n\n```bash\nuv run vf-eval trading-bot-bench   -m gpt-4.1-mini   -n 20 -r 3 -t 1024 -T 0.7   -a '{\"key\": \"value\"}'  # env-specific args as JSON\n```\n\nNotes:\n- Use `-a` / `--env-args` to pass environment-specific configuration as a JSON object.\n\n### Environment Arguments\nDocument any supported environment arguments and their meaning. Example:\n\n| Arg | Type | Default | Description |\n| --- | ---- | ------- | ----------- |\n| `verbose` | bool | `\"False\"` | enables verbose logging |\n\n### Metrics\nSummarize key metrics your rubric emits and how they’re interpreted.\n\n| Metric | Meaning |\n| ------ | ------- |\n| `reward` | Main scalar reward (weighted sum of normalized return, Sharpe ratio, and drawdown penalty) |\n| `return_reward` | Normalized measure of total return |\n| `sharpe_reward` | Normalized measure of Sharpe ratio |\n| `drawdown_penalty` | Penalty factor for maximum drawdown; discourages large losses |\n\n\n","encoding":"utf-8","truncated":false,"total_bytes":1845},"status":null}