Skip to content

Conversation

@abhiksingla
Copy link

Why are these changes needed?

The change provides the support to run behavior cloning steps (pre-training stage for Offline-RL training) dynamically according to the data size. In the previous implementation, behavior cloning is done for a fixed number of steps. The changes will set behavior cloning iters dynamically according to the size of the dataset.

Related issue number

Checks

  • I've run scripts/format.sh to lint the changes in this PR.
  • I've included any doc changes needed for https://docs.ray.io/en/master/.
  • I've made sure the tests are passing. Note that there might be a few flaky tests, see the recent failures at https://flakey-tests.ray.io/
  • Testing Strategy
    • Unit tests
    • Release tests
    • This PR is not tested :(

@abhiksingla abhiksingla requested a review from a team as a code owner June 29, 2022 00:21
@abhiksingla abhiksingla requested review from RuofanKong and akanso June 29, 2022 00:21
@abhiksingla abhiksingla requested a review from dmlyubim June 29, 2022 01:13
Copy link
Collaborator

@RuofanKong RuofanKong left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

looks good.

@abhiksingla abhiksingla requested a review from Kiko-Aumond July 14, 2022 17:29
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants