# Just Bench It: RL Algorithm Benchmarking Tool
这个项目提供了一个简单的工具，用于对强化学习（RL）算法在Atari游戏上进行基准测试。
WEBSITE: https://justbechit.github.io/rl_ladder/
## 安装
## PYPI
1. 安装：
```
pip install just-bench-it
```
### Build from source
1. 克隆这个仓库：
```
git clone https://github.com/your_username/just_bench_it.git
cd just_bench_it
```
2. 安装依赖：
```
pip install -e .
```
## 使用方法
1. 创建你的RL agent类，并使用`@benchmark`装饰器。
2. 在你的agent类中实现以下方法：
- `set_env_info(self, env_info)`: 设置环境信息
- `act(self, state)`: 根据当前状态选择动作
- `update(self, state, action, reward, next_state, done)`: 更新agent的内部状态或模型
3. 运行你的脚本来执行基准测试。
## 示例
这里有一个DQN agent的示例实现：
```python
from just_bench_it import benchmark
@benchmark(pretrained=False, train_episodes=1000, eval_episodes=100)
class DQNAgent:
def __init__(self):
# 初始化你的DQN agent
pass
def set_env_info(self, env_info):
# 设置环境信息： bench_it 会提供当前动作空间和观察空间
# input_shape = env_info['observation_space'].shape
# output_dim = env_info['action_space'].n
# 不同的环境其输入可能不同，确保您的算法能够应对不同环境
pass
def act(self, state):
# 根据状态选择动作
pass
def update(self, state, action, reward, next_state, done):
# 更新agent
pass
if __name__ == "__main__":
agent = DQNAgent()
results = agent.bench()
print(results)
```
## 自定义
你可以通过修改`@benchmark`装饰器的参数来自定义基准测试：
- `pretrained`: 是否使用预训练模型（默认为False）
- `train_episodes`: 训练的回合数（默认为1000）
- `eval_episodes`: 评估的回合数（默认为100）
## 结果
基准测试的结果会自动发布为GitHub issue，包含每个环境的平均得分和其他相关信息。
## 贡献
欢迎提交问题报告和拉取请求。对于重大更改，请先开issue讨论您想要更改的内容。
## 许可证
[MIT](https://choosealicense.com/licenses/mit/)
Raw data
{
"_id": null,
"home_page": "https://github.com/justbechit/just_bench_it",
"name": "just-bench-it",
"maintainer": null,
"docs_url": null,
"requires_python": null,
"maintainer_email": null,
"keywords": null,
"author": "stone91",
"author_email": "370025263@qq.com",
"download_url": "https://files.pythonhosted.org/packages/aa/e1/8e591c6875814af77f46ad7323356ae200ad1865bb3bb8662275184c9d3e/just_bench_it-0.1.9.tar.gz",
"platform": null,
"description": "\n# Just Bench It: RL Algorithm Benchmarking Tool\n\n\u8fd9\u4e2a\u9879\u76ee\u63d0\u4f9b\u4e86\u4e00\u4e2a\u7b80\u5355\u7684\u5de5\u5177\uff0c\u7528\u4e8e\u5bf9\u5f3a\u5316\u5b66\u4e60\uff08RL\uff09\u7b97\u6cd5\u5728Atari\u6e38\u620f\u4e0a\u8fdb\u884c\u57fa\u51c6\u6d4b\u8bd5\u3002\nWEBSITE: https://justbechit.github.io/rl_ladder/\n\n## \u5b89\u88c5\n## PYPI\n\n1. \u5b89\u88c5\uff1a\n ```\n pip install just-bench-it\n ```\n\n### Build from source\n\n1. \u514b\u9686\u8fd9\u4e2a\u4ed3\u5e93\uff1a\n ```\n git clone https://github.com/your_username/just_bench_it.git\n cd just_bench_it\n ```\n\n2. \u5b89\u88c5\u4f9d\u8d56\uff1a\n ```\n pip install -e .\n ```\n\n## \u4f7f\u7528\u65b9\u6cd5\n\n1. \u521b\u5efa\u4f60\u7684RL agent\u7c7b\uff0c\u5e76\u4f7f\u7528`@benchmark`\u88c5\u9970\u5668\u3002\n\n2. \u5728\u4f60\u7684agent\u7c7b\u4e2d\u5b9e\u73b0\u4ee5\u4e0b\u65b9\u6cd5\uff1a\n - `set_env_info(self, env_info)`: \u8bbe\u7f6e\u73af\u5883\u4fe1\u606f\n - `act(self, state)`: \u6839\u636e\u5f53\u524d\u72b6\u6001\u9009\u62e9\u52a8\u4f5c\n - `update(self, state, action, reward, next_state, done)`: \u66f4\u65b0agent\u7684\u5185\u90e8\u72b6\u6001\u6216\u6a21\u578b\n\n3. \u8fd0\u884c\u4f60\u7684\u811a\u672c\u6765\u6267\u884c\u57fa\u51c6\u6d4b\u8bd5\u3002\n\n## \u793a\u4f8b\n\n\u8fd9\u91cc\u6709\u4e00\u4e2aDQN agent\u7684\u793a\u4f8b\u5b9e\u73b0\uff1a\n\n```python\nfrom just_bench_it import benchmark\n\n@benchmark(pretrained=False, train_episodes=1000, eval_episodes=100)\nclass DQNAgent:\n def __init__(self):\n # \u521d\u59cb\u5316\u4f60\u7684DQN agent\n pass\n\n def set_env_info(self, env_info):\n # \u8bbe\u7f6e\u73af\u5883\u4fe1\u606f\uff1a bench_it \u4f1a\u63d0\u4f9b\u5f53\u524d\u52a8\u4f5c\u7a7a\u95f4\u548c\u89c2\u5bdf\u7a7a\u95f4\n # input_shape = env_info['observation_space'].shape\n # output_dim = env_info['action_space'].n\n # \u4e0d\u540c\u7684\u73af\u5883\u5176\u8f93\u5165\u53ef\u80fd\u4e0d\u540c\uff0c\u786e\u4fdd\u60a8\u7684\u7b97\u6cd5\u80fd\u591f\u5e94\u5bf9\u4e0d\u540c\u73af\u5883\n pass\n\n def act(self, state):\n # \u6839\u636e\u72b6\u6001\u9009\u62e9\u52a8\u4f5c\n pass\n\n def update(self, state, action, reward, next_state, done):\n # \u66f4\u65b0agent\n pass\n\nif __name__ == \"__main__\":\n agent = DQNAgent()\n results = agent.bench()\n print(results)\n```\n\n## \u81ea\u5b9a\u4e49\n\n\u4f60\u53ef\u4ee5\u901a\u8fc7\u4fee\u6539`@benchmark`\u88c5\u9970\u5668\u7684\u53c2\u6570\u6765\u81ea\u5b9a\u4e49\u57fa\u51c6\u6d4b\u8bd5\uff1a\n\n- `pretrained`: \u662f\u5426\u4f7f\u7528\u9884\u8bad\u7ec3\u6a21\u578b\uff08\u9ed8\u8ba4\u4e3aFalse\uff09\n- `train_episodes`: \u8bad\u7ec3\u7684\u56de\u5408\u6570\uff08\u9ed8\u8ba4\u4e3a1000\uff09\n- `eval_episodes`: \u8bc4\u4f30\u7684\u56de\u5408\u6570\uff08\u9ed8\u8ba4\u4e3a100\uff09\n\n## \u7ed3\u679c\n\n\u57fa\u51c6\u6d4b\u8bd5\u7684\u7ed3\u679c\u4f1a\u81ea\u52a8\u53d1\u5e03\u4e3aGitHub issue\uff0c\u5305\u542b\u6bcf\u4e2a\u73af\u5883\u7684\u5e73\u5747\u5f97\u5206\u548c\u5176\u4ed6\u76f8\u5173\u4fe1\u606f\u3002\n\n## \u8d21\u732e\n\n\u6b22\u8fce\u63d0\u4ea4\u95ee\u9898\u62a5\u544a\u548c\u62c9\u53d6\u8bf7\u6c42\u3002\u5bf9\u4e8e\u91cd\u5927\u66f4\u6539\uff0c\u8bf7\u5148\u5f00issue\u8ba8\u8bba\u60a8\u60f3\u8981\u66f4\u6539\u7684\u5185\u5bb9\u3002\n\n## \u8bb8\u53ef\u8bc1\n\n[MIT](https://choosealicense.com/licenses/mit/)\n\n\n",
"bugtrack_url": null,
"license": null,
"summary": "A simple benchmarking tool for RL algorithms on Atari games",
"version": "0.1.9",
"project_urls": {
"Homepage": "https://github.com/justbechit/just_bench_it"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "89ea2cf6ff44d01a45ccdf4704bd7ac487c3aa4918c9e9d20785767348f603f2",
"md5": "faef0b28cf1dd53c0d69bb1be6f29e7a",
"sha256": "1f5e365c4a611fd9dba9503c2cda3772eb401a0c381fdd3df3afa0a6e7383fe0"
},
"downloads": -1,
"filename": "just_bench_it-0.1.9-py3-none-any.whl",
"has_sig": false,
"md5_digest": "faef0b28cf1dd53c0d69bb1be6f29e7a",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": null,
"size": 8991,
"upload_time": "2024-08-02T09:41:14",
"upload_time_iso_8601": "2024-08-02T09:41:14.897951Z",
"url": "https://files.pythonhosted.org/packages/89/ea/2cf6ff44d01a45ccdf4704bd7ac487c3aa4918c9e9d20785767348f603f2/just_bench_it-0.1.9-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "aae18e591c6875814af77f46ad7323356ae200ad1865bb3bb8662275184c9d3e",
"md5": "9163bee98c9864fdb0a80cafedc821b5",
"sha256": "b1928f4839927c39276b92fb6a5445e967ef9656bfc299932260f4742d4d7c63"
},
"downloads": -1,
"filename": "just_bench_it-0.1.9.tar.gz",
"has_sig": false,
"md5_digest": "9163bee98c9864fdb0a80cafedc821b5",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 8670,
"upload_time": "2024-08-02T09:41:17",
"upload_time_iso_8601": "2024-08-02T09:41:17.313424Z",
"url": "https://files.pythonhosted.org/packages/aa/e1/8e591c6875814af77f46ad7323356ae200ad1865bb3bb8662275184c9d3e/just_bench_it-0.1.9.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-08-02 09:41:17",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "justbechit",
"github_project": "just_bench_it",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "just-bench-it"
}
JSON
