Repository navigation
Docs: vLLM backend needs a vLLM nightly and supports Qwen3.5 only - #9
Merged
Merged
Conversation
- README: install the vLLM nightly for the vLLM backend (vLLM 0.32.0 on PyPI predates its decision-model support) and note the GPU requirement (Ampere or newer for non-causal checkpoints such as vllm-sr/d3-edge). - Qwen3.5/3.8 -> Qwen3.5 in the backend docstring, its refusal message and the --backend help: only Qwen3_5Model checkpoints are accepted. - The [vllm] extra's comment says the nightly is needed.
akanyaani
added a commit
to OthmanAdi/decisionformer
that referenced
this pull request
Oct 27, 2026
Resolves conflicts with FrontiersMindAI#9 in the vLLM backend description and the --backend help: Qwen3.5 only for code readouts, which need a vLLM nightly; native Qwen3 classifier exports use vLLM's pooling runner.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes the install instructions and support claims for the vLLM backend.
pip install "decisionformer[vllm]", which installs vLLM 0.32.0. That release (PyPI, 2026-10-18) has no Decision 3.0 model:vllm/model_executor/models/d3.pyis missing at thev0.32.0tag and was added on vLLM main on 2026-10-15. Soserve vllm-sr/d3-edge --backend vllmfailed withUnsupportedBackend. The README now installs the vLLM nightly, and notes that non-causal checkpoints such as d3-edge need an Ampere or newer GPU on vLLM. The install table'svllmrow points there.--backendhelp. The backend accepts onlyqwen3_5/Qwen3_5Model(vllm_backend.py:52), and 3.8 has never been tested.[vllm]extra's comment says the nightly is needed.Text only: +12/−9 across 4 files, with no code paths changed.