THUDM / THUDM/slime

TypeError: must be real number, not NoneType when running Fully Asynchronous Rollout Example

Open
#488 2 comments 3 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
8.5k
Forks
1.3k
Avg merge
5h 36m
Merged PRs (30d)
22

Description

Description

While launching the official Fully Asynchronous Rollout Example (examples/fully_async/run-qwen3-4b-fully_async.sh), the script crashes inside the Ray actor RolloutManager.generate().
The list raw_rewards contains None values, so the line

rewards = torch.tensor(raw_rewards, dtype=torch.float)

raises

TypeError: must be real number, not NoneType.

Full traceback and training logs are attached below.

Image

Questions

What is the recommended upstream fix or configuration change?

Thanks!

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reproducing the failure with examples/fully_async/run-qwen3-4b-fully_async.sh and trace the Ray actor entry point RolloutManager.generate(), using the attached traceback and logs. Determine why raw_rewards contains None before the torch.tensor conversion; done means the official example completes with a documented upstream fix or configuration change.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
distributed-systems, machine-learning
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.