bytedance / bytedance/WMP

训练的时候重点关注那些坐标到达,什么程度是正常的

Open
#10 2 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
381
Forks
35
PR merge metrics
No merged PRs in 30d

Description

我在训练的时候go2的max episode length只有700多,理论上这个值不应该接近1000吗,训练出的go2策略在sim2sim后表现的很不稳定,碰到台阶什么都就乱摆腿,然后摔倒。在gym里面又是正常的。目前a1训出的策略可以实现sim2sim(排除了sim2sim代码的问题)

Contributor guide

No contributing guide indexed for this repository

Research direction

Start by comparing the Go2 and A1 training configurations and their maximum episode-length results in the gym environment. Then reproduce the Go2 policy in sim2sim, especially its behavior when encountering steps, and determine whether the episode length near 700 is expected or indicates a training discrepancy. Done means the cause of the unstable Go2 behavior and the expected episode length are documented.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning, robotics
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.