Description
vLLM is a library for LLM inference and serving. vllm/model_executor/weight_utils.py implements hf_model_weights_iterator to load the model checkpoint, which is downloaded from huggingface. It uses the torch.load function and the weights_only parameter defaults to False. When torch.load loads malicious pickle data, it will execute arbitrary code during unpickling. This vulnerability is fixed in v0.7.0.
Metrics
CVSS 3.1 8.8/10
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:HWeakness enumeration
Affected software
| Vendor | Product | Versions |
|---|---|---|
| vllm-project | vllm | < 0.7.0 |
References
- https://github.com/vllm-project/vllm/commit/d3d6bb13fb62da3234addf6574922a4ec0513d04Patch
- https://github.com/vllm-project/vllm/pull/12366Issue Tracking, Patch
- https://github.com/vllm-project/vllm/security/advisories/GHSA-rh4j-5rhw-hr54Vendor Advisory
- https://pytorch.org/docs/stable/generated/torch.load.htmlTechnical Description
Timeline
- Published
- Jan 27, 2025
- Last modified
- Jun 17, 2026
- NVD status
- Analyzed