Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
rical730 21d1adfd77 | 3 years ago | |
---|---|---|
.. | ||
benchmark | 4 years ago | |
cmake/Torch | 3 years ago | |
core | 4 years ago | |
demo | 3 years ago | |
paddle | 4 years ago | |
scripts | 3 years ago | |
test | 3 years ago | |
torch/include/evo_kit | 4 years ago | |
CMakeLists.txt | 3 years ago | |
DeepES.gif | 4 years ago | |
README.md | 3 years ago |
EvoKit 是一个集合了多种进化算法、兼容多种类预测框架的进化算法库,主打快速上线验证 。
//实例化一个预测,根据配置文件加载模型,采样方式(Gaussian\CMA sampling..)、更新方式(SGD\Adam)等
auto agent = ESAgent(config);
for (int i = 0; i < 10; ++i) {
auto sampling_agnet = agent->clone(); // clone出一个sampling agent
SamplingInfo info;
sampling_agent->add_noise(info); // 参数扰动,同时保存随机种子到info中
int reward = evaluate(env, sampling_agent); //评估参数
noisy_info.push_back(info); // 记录随机噪声对应种子
noisy_rewards.push_back(reward); // 记录评估结果
}
//根据评估结果、随机种子更新参数,然后重复以上过程,直到收敛。
agent->update(noisy_info, noisy_rewards);
下载PaddleLite的X86预编译库,或者编译PaddleLite源码,得到inference_lite_lib文件夹,放在当前目录中。(可参考:PaddleLite使用X86预测部署)
下载libtorch或者编译torch源码,得到libtorch文件夹,放在当前目录中。
PARL 是一个高性能、灵活的强化学习框架
Python C++ JavaScript Shell Markdown other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》