[training_utils] fix: Resolved bugs and conflicts in the fully async caused by multiple PRs

[ZLiao097]

What does this PR do?

Add concise overview of what this PR aims to achieve or accomplish. Reference related GitHub issues and PRs that help with the review.

Problem1: ConfigAttributeError('Missing key config\n full_key: config\n object_type=dict')
This error was introduced by this PR: dataset_config type was changed from DictConfig to DictConfigWrap in AgentLoopBase initialization (using dataset_config.config for passing), but the fully async agentloop failed to update dataset_config to DictConfigWrap, causing the error.

The following two problems were introduced by this PR:

Problem2: TypeError: got an unexpected keyword argument 'cuda_visible_devices'
The PR added cuda_visible_devices to vLLMHttpServer, but its subclass vLLMHttpServerForPartial in fully async was not updated accordingly, causing conflicts.

Problem3: KeyError: 'ASCEND_RT_VISIBLE_DEVICES'
The PR references the environment variable ASCEND_RT_VISIBLE_DEVICES in get_device_uuid but does not handle its absence or set a default value, leading to potential errors.

Checklist Before Starting

• Search for similar PRs. Paste at least one query link here: ...
• Format the PR title as [{modules}] {type}: {description} (This will be checked by the CI)
  • {modules} include fsdp, megatron, veomni, sglang, vllm, rollout, trainer, ci, training_utils, recipe, hardware, deployment, ray, worker, single_controller, misc, perf, model, algo, env, tool, ckpt, doc, data, cfg, reward
  • If this PR involves multiple modules, separate them with , like [megatron, fsdp, doc]
  • {type} is in feat, fix, refactor, chore, test
  • If this PR breaks any API (CLI arguments, config, function signature, etc.), add [BREAKING] to the beginning of the title.
  • Example: [BREAKING][fsdp, megatron] feat: dynamic batching

Test

For changes that can not be tested by CI (e.g., algorithm implementation, new model support), validate by experiment(s) and show results like training curve plots, evaluation results, etc.

API and Usage Example

Demonstrate how the API changes if any, and provide usage example(s) if possible.

# Add code snippet or script demonstrating how to use this

Design & Code Changes

Demonstrate the high-level design if this PR is complex, and list the specific changes.

Checklist Before Submitting

Important

Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review.

• Read the Contribute Guide.
• Apply pre-commit checks: pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always
• Add / Update the documentation.
• Add unit or end-to-end test(s) to the CI workflow to cover all the code. If not feasible, explain why: ...
• Once your PR is ready for CI, send a message in the ci-request channel in the verl Slack workspace. (If not accessible, please try the Feishu group (飞书群).)
• If your PR is related to the recipe submodule, please also update the reference to the submodule commit via git submodule update --remote or cd recipe && git pull origin main.

[tardis-key]

Could u confirm if there are any other places that also require this fix?

[Shangwei-Li]

Would you please add CUDA_VISIBLE_DEVICES related fixes to this PR?

[ZLiao097]

Could u confirm if there are any other places that also require this fix?

I've checked other potential error points.

[ZLiao097]

Would you please add CUDA_VISIBLE_DEVICES related fixes to this PR?

Done.

[ZLiao097]

@tardis-key I've additionally fixed two other conflict points and merged them into this PR. Please re-trigger the CI for me ;)

[ZLiao097]

Could u confirm if there are any other places that also require this fix?

I've checked other potential error points.

@tardis-key I've also checked the second problem – only one place needs modification.