search

pagand / ORL_optimizer

offline RL optimizer
0 stars 0 forks source link

Preliminaries #3

Closed pagand closed 1 month ago

pagand commented 1 month ago

Getting familiar with the required background to get ready for the projects

@tzhao-ooc @jnqian99 Manage your time to finish this stage by our next internal meeting.

jnqian99 commented 1 month ago

I will pick project 2 if that is OK with you guys.

pagand commented 1 month ago

Thanks @jnqian99 You can work on project 2 (GYM model). Please look (in order) at this paper table 2 for Gym-mujoco tasks, table 5 for v-D4RL table 3 for antmaze, and table 4 for adroidt tasks. ideally we want overall 3-5 models from different task, preferably simple (not transformer).

@tzhao-ooc You can also work on project 1 (vessel model) as you wanted. @jnqian99 will join you after he finish project 1 and both can work collaboratively. Start looking at paper 1 and 2. Then try converting the ipynb to python file and checking the code to migrate the code from the previous repo to this one. Feel free to ask questions and check our previous repo.

Both of you, create a branch under your name in this repo and try to push code to your branch and issue a pull request after you an stage is approved to merge to main.