Closed
Conversation
torch.barrier has a fixed limited timeout depending on its backend so won't help keep all MCT managed processes alive. So we need a new barrier mechanism. This implementation is a refactor of existing SyncActor but made it more general to serve as a barrier between clients.
Added a single arg parse for `file_path` for omegaconf yaml path Also added support for omegaconf yamls
Example run: compose-rl-distributed-ppo-test-yRLEyN This is tested with 2x8 GPUs
<img width="1072" height="668" alt="image" src="https://github.com/user-attachments/assets/a6166c79-b136-4290-9344-fb68f1aedfa8" /> still missing a couple steps for some reason, going to ask the mlflow team about it but wanted to get this up asap so we can have logging.
Uses a llama 8b math dataset instead of qwen open_r1. Modifies the generation length and sequence length to see rewards and hillclimbing in 5 steps. [mlflow green curve](https://dbc-559ffd80-2bfc.cloud.databricks.com/ml/experiments/3739963835954932/runs?o=7395834863327820&searchFilter=&orderByKey=attributes.start_time&orderByAsc=false&startTime=ALL&lifecycleFilter=Active&modelVersionFilter=All+Runs&datasetsFilter=W10%3D&compareRunsMode=CHART)
<img width="926" height="400" alt="image" src="https://github.com/user-attachments/assets/77ec07e3-982a-4252-a3bb-785d1273eca5" /> all 39 steps!
bowenyang008
requested review from
abaheti95,
gupta-abhay and
jdchang1
as code owners
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.