eval 代码验证
- 作者仓库星标 1,155
- 许可证 BSD-2-Clause
- 作者仓库 LearningHumanoidWalking
/eval — Evaluate a Trained Checkpoint
Parse the user's request from $ARGUMENTS and run evaluation.
Command Template
uv run python run_experiment.py eval --path <PATH> [OPTIONS...]Path Resolution
The user may provide:
- A .pt file: Use directly (
--path /tmp/.../actor_999.pt) - A run directory: Contains actor*.pt files (
--path /tmp/.../26-03-07-00-26-36_cartpole/) - A logdir: Contains timestamped run subdirectories (
--logdir /tmp/training_runs)
If no path is given, check /tmp/training_runs for the most recent run.
Use Glob to verify the path exists and resolve it before running.
Options
| Flag | Default | Description |
|---|---|---|
--ep-len |
10 | Episode length in seconds |
--seed |
None | Random seed for reproducible eval |
--out-dir |
None | Directory to save videos |
Instructions
- Resolve the model path from the user's input. If ambiguous, list available checkpoints and ask.
- Show the user which checkpoint will be evaluated (full path).
- Run the eval command. This opens an interactive MuJoCo viewer window — it is NOT a background job.
- Report the results when done.
- 流狐分类
- 通用
- 作者声明 Agent
- 未找到明确声明；不据此推断已兼容或已测试
- 静态检查
- 94 / 100 · 启发式扫描，不代表运行安全
- 作者 / 版本 / 许可
- @rohanpsingh · BSD-2-Clause
- 流狐 Token 估算
- 低消耗
- 流狐接入估算
- 即装即用
- 是否需要外部 API Key
- 未发现要求
- 检测到的系统要求
- 未声明
- 底层运行要求
- Python
- 检测到的文件与系统行为
-
- 只读
- 检测到的网络行为
- 仅限本地
- 安装命令数
- 无（仅作为资料）
档案由构建时根据 SKILL.md 与安装命令自动衍生，可能与作者实际意图存在差异。
需要注意： 未限定 allowed-tools，默认拥有全部工具权限。
作者没有在当前 SKILL.md 中定义固定输出样例。Command Template
The user may provide: A .pt file: Use directly (--path /tmp/.../actor999.pt) A run directory: Contains actor.pt files (--path /tmp/.../26-03-07-00-26-36cartpole/)
Flag · Default · Description --ep-len · 10 · Episode length in seconds --seed · None · Random seed for reproducible eval
Resolve the model path from the user's input. If ambiguous, list available checkpoints and ask. Show the user which checkpoint will be evaluated (full path). Run the eval command. This opens an interactive MuJoCo viewer window — it is NOT a background job.
# /eval — Evaluate a Trained Checkpoint
Parse the user's request from `$ARGUMENTS` and run evaluation.
## Command Template
```
uv run python run_experiment.py eval --path <PATH> [OPTIONS...]
```
## Path Resolution
The user may provide:
- **A .pt file**: Use directly (`--path /tmp/.../actor_999.pt`)
- **A run directory**: Contains actor*.pt files (`--path /tmp/.../26-03-07-00-26-36_cartpole/`)
- **A logdir**: Contains timestamped run subdirectories (`--logdir /tmp/training_runs`)
If no path is given, check `/tmp/training_runs` for the most recent run.
Use `Glob` to verify the path exists and resolve it before running.
## Options
| Flag | Default | Description |
|------|---------|-------------|
| `--ep-len` | 10 | Episode length in seconds |
| `--seed` | None | Random seed for reproducible eval |
| `--out-dir` | None | Directory to save videos |
## Instructions
1. Resolve the model path from the user's input. If ambiguous, list available checkpoints and ask.
2. Show the user which checkpoint will be evaluated (full path).
3. Run the eval command. This opens an interactive MuJoCo viewer window — it is NOT a background job.
4. Report the results when done.作者原文负责流程事实；流狐只索引当前章节、要点、文件与命令。
章节 -> Command Template → Path Resolution → Options → Instructions
要点 -> A .pt file · A run directory · A logdir · Parse the user's request from $ARGUMENTS and run evaluation. · If no path is given, check /tmp/trainingruns for the most recent run. · Use Glob to verify the path exists and resolve it before running. · 1. Resolve the model path from the user's input.
文件/命令 -> $ARGUMENTS · --path /tmp/.../actor999.pt · --path /tmp/.../26-03-07-00-26-36cartpole/ · --logdir /tmp/trainingruns · /tmp/trainingruns · Glob · --ep-len · --seed
内容 SHA-256 -> a33ca52ffbb1
原文结构
适用与边界
原文中的明确线索
$ARGUMENTS、--path /tmp/.../actor999.pt、--path /tmp/.../26-03-07-00-26-36cartpole/、--logdir /tmp/trainingruns、/tmp/trainingruns、Glob、--ep-len、--seed