[V0 Deprecation] Remove vllm.worker and update according imports
#25901
Conversation
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
Signed-off-by: Aaron Pham <contact@aarnphm.xyz>
aarnphm
requested review from
22quinn,
WoosukKwon,
alexm-redhat,
comaniac,
hmellor,
njhill,
robertgshaw2-redhat,
youkaichao,
ywang96 and
zhuohan123
as code owners
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request effectively removes the deprecated vllm.worker directory and migrates its functionality to vllm.v1, successfully deprecating the V0 worker implementation. The changes are clean and well-executed. Key improvements include:
- Consolidating worker base classes into
vllm/v1/worker/worker_base.py. - Removing the insecure and deprecated usage of
cloudpicklefor dynamic worker class loading, replacing it with a clear error for unsupported configurations. - Simplifying the worker selection logic in platform-specific files (
cuda.py,rocm.py) to always use the V1 worker. - Updating all relevant imports across the codebase to point to the new
vllm.v1.workerpath.
The refactoring makes the codebase cleaner, more secure, and easier to maintain by removing the old V0 logic. I have reviewed the changes and found no issues. The implementation is solid.
Signed-off-by: Aaron Pham <contact@aarnphm.xyz>
zhuohan123
approved these changes
Sep 29, 2025
njhill
reviewed
Sep 29, 2025
|
Will wait to trigger CI to make sure there aren't any failures |
WoosukKwon
reviewed
Sep 29, 2025
Comment on lines
+115
to
+119
| def execute_model( | ||
| self, | ||
| execute_model_req: Optional[ExecuteModelRequest] = None | ||
| ) -> Optional[list[SamplerOutput]]: | ||
| raise NotImplementedError |
There was a problem hiding this comment.
FYI, this method is also outdated. V1 doesn't use ExecuteModelRequest at all.
pdasigi
pushed a commit
to pdasigi/vllm
that referenced
this pull request
Oct 2, 2025
yewentao256
pushed a commit
that referenced
this pull request
Oct 3, 2025
…25901) Signed-off-by: yewentao256 <zhyanwentao@126.com>
tomeras91
pushed a commit
to tomeras91/vllm
that referenced
this pull request
Oct 6, 2025
…llm-project#25901) Signed-off-by: Tomer Asida <57313761+tomeras91@users.noreply.github.com>
|
@aarnphm better to import the v1 worker here to keep compatibility? people might still import |
|
Are you suggesting a small shim? afaict in v0.11.0 we announce that we completely remove v0? |
HollowMan6
added a commit
to HollowMan6/verl
that referenced
this pull request
Oct 6, 2025
Related to: - vllm-project/vllm#25901 - vllm-project/vllm#25345 Now we first try to import `WorkerWrapperBase` from `vllm.worker.worker_base`, if we have an error, we append `v1` there. For `compute_logits` patch, we can just remove the import of `SamplingMetadata`, create a wrapper that accepts any arguments with *args, **kwargs, and pass them through to the original method, so that it can be more flexible and future-proof. Signed-off-by: Hollow Man <hollowman@opensuse.org>
7 tasks
HollowMan6
added a commit
to HollowMan6/verl
that referenced
this pull request
Oct 6, 2025
Related to: - vllm-project/vllm#25901 - vllm-project/vllm#25345 Now we first try to import `WorkerWrapperBase` from `vllm.worker.worker_base`, if we have an error, we append `v1` there. For `compute_logits` patch, we can just remove the import of `SamplingMetadata`, create a wrapper that accepts any arguments with *args, **kwargs, and pass them through to the original method, so that it can be more flexible and future-proof. Signed-off-by: Hollow Man <hollowman@opensuse.org>
wuxibin89
pushed a commit
to volcengine/verl
that referenced
this pull request
Oct 9, 2025
### 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. Related to: - vllm-project/vllm#25901 - vllm-project/vllm#25345 Now we first try to import `WorkerWrapperBase` from `vllm.worker.worker_base`, if we have an error, we append `v1` there. For `compute_logits` patch, we can just remove the import of `SamplingMetadata`, create a wrapper that accepts any arguments with *args, **kwargs, and pass them through to the original method, so that it can be more flexible and future-proof. ### Checklist Before Starting - [X] Search for similar PRs. Paste at least one query link here: ... - [X] Format the PR title as `[{modules}] {type}: {description}` (This will be checked by the CI) - `{modules}` include `fsdp`, `megatron`, `sglang`, `vllm`, `rollout`, `trainer`, `ci`, `training_utils`, `recipe`, `hardware`, `deployment`, `ray`, `worker`, `single_controller`, `misc`, `perf`, `model`, `algo`, `env`, `tool`, `ckpt`, `doc`, `data` - 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. ```python # 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. - [X] Read the [Contribute Guide](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md). - [X] Apply [pre-commit checks](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md#code-linting-and-formatting): `pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always` - [X] Add / Update [the documentation](https://github.com/volcengine/verl/tree/main/docs). - [X] Add unit or end-to-end test(s) to [the CI workflow](https://github.com/volcengine/verl/tree/main/.github/workflows) to cover all the code. If not feasible, explain why: ... - [X] Once your PR is ready for CI, send a message in [the `ci-request` channel](https://verl-project.slack.com/archives/C091TCESWB1) in [the `verl` Slack workspace](https://join.slack.com/t/verl-project/shared_invite/zt-3855yhg8g-CTkqXu~hKojPCmo7k_yXTQ). (If not accessible, please try [the Feishu group (飞书群)](https://applink.larkoffice.com/client/chat/chatter/add_by_link?link_token=772jd4f1-cd91-441e-a820-498c6614126a).) Signed-off-by: Hollow Man <hollowman@opensuse.org>
xuebwang-amd
pushed a commit
to xuebwang-amd/vllm
that referenced
this pull request
Oct 10, 2025
…llm-project#25901) Signed-off-by: xuebwang-amd <xuebwang@amd.com>
masoudhashemi
pushed a commit
to masoudhashemi/verl
that referenced
this pull request
Oct 19, 2025
### 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. Related to: - vllm-project/vllm#25901 - vllm-project/vllm#25345 Now we first try to import `WorkerWrapperBase` from `vllm.worker.worker_base`, if we have an error, we append `v1` there. For `compute_logits` patch, we can just remove the import of `SamplingMetadata`, create a wrapper that accepts any arguments with *args, **kwargs, and pass them through to the original method, so that it can be more flexible and future-proof. ### Checklist Before Starting - [X] Search for similar PRs. Paste at least one query link here: ... - [X] Format the PR title as `[{modules}] {type}: {description}` (This will be checked by the CI) - `{modules}` include `fsdp`, `megatron`, `sglang`, `vllm`, `rollout`, `trainer`, `ci`, `training_utils`, `recipe`, `hardware`, `deployment`, `ray`, `worker`, `single_controller`, `misc`, `perf`, `model`, `algo`, `env`, `tool`, `ckpt`, `doc`, `data` - 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. ```python # 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. - [X] Read the [Contribute Guide](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md). - [X] Apply [pre-commit checks](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md#code-linting-and-formatting): `pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always` - [X] Add / Update [the documentation](https://github.com/volcengine/verl/tree/main/docs). - [X] Add unit or end-to-end test(s) to [the CI workflow](https://github.com/volcengine/verl/tree/main/.github/workflows) to cover all the code. If not feasible, explain why: ... - [X] Once your PR is ready for CI, send a message in [the `ci-request` channel](https://verl-project.slack.com/archives/C091TCESWB1) in [the `verl` Slack workspace](https://join.slack.com/t/verl-project/shared_invite/zt-3855yhg8g-CTkqXu~hKojPCmo7k_yXTQ). (If not accessible, please try [the Feishu group (飞书群)](https://applink.larkoffice.com/client/chat/chatter/add_by_link?link_token=772jd4f1-cd91-441e-a820-498c6614126a).) Signed-off-by: Hollow Man <hollowman@opensuse.org>
lywa1998
pushed a commit
to lywa1998/vllm
that referenced
this pull request
Oct 20, 2025
alhridoy
pushed a commit
to alhridoy/vllm
that referenced
this pull request
Oct 24, 2025
xuebwang-amd
pushed a commit
to xuebwang-amd/vllm
that referenced
this pull request
Oct 24, 2025
…llm-project#25901) Signed-off-by: xuebwang-amd <xuebwang@amd.com>
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.
This PR removes the
vllm/workerdirectory and migrate relevant class tovllm/v1Some change behaviours include:
init_worker(cc @russellb)I haven't touched the ray executor part yet.
Signed-off-by: Aaron Pham contact@aarnphm.xyz