High severityNVD Advisory· Published Jan 27, 2025· Updated Feb 12, 2025
vLLM allows a malicious model RCE by torch.load in hf_model_weights_iterator
CVE-2025-24357
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.
AI Insight
LLM-synthesized narrative grounded in this CVE's description and references.
Affected packages
Versions sourced from the GitHub Security Advisory.
| Package | Affected versions | Patched versions |
|---|---|---|
vllmPyPI | < 0.7.0 | 0.7.0 |
Affected products
6- osv-coords5 versionspkg:apk/chainguard/py3.10-vllm-cuda-11.8pkg:apk/chainguard/py3.10-vllm-cuda-12.6pkg:apk/chainguard/py3.10-wheels-vllm-cuda-11.8pkg:apk/chainguard/tritonserver-backend-vllm-24.04pkg:pypi/vllm
< 0.7.2-r0+ 4 more
- (no CPE)range: < 0.7.2-r0
- (no CPE)range: < 0.7.2-r0
- (no CPE)range: < 0.7.2-r0
- (no CPE)range: < 24.04-r9
- (no CPE)range: < 0.7.0
Patches
Vulnerability mechanics
References
8- github.com/advisories/GHSA-rh4j-5rhw-hr54ghsaADVISORY
- nvd.nist.gov/vuln/detail/CVE-2025-24357ghsaADVISORY
- github.com/pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2025-58.yamlghsaWEB
- github.com/vllm-project/vllm/commit/d3d6bb13fb62da3234addf6574922a4ec0513d04ghsax_refsource_MISCWEB
- github.com/vllm-project/vllm/pull/12366ghsax_refsource_MISCWEB
- github.com/vllm-project/vllm/releases/tag/v0.7.0ghsaWEB
- github.com/vllm-project/vllm/security/advisories/GHSA-rh4j-5rhw-hr54ghsax_refsource_CONFIRMWEB
- pytorch.org/docs/stable/generated/torch.load.htmlghsax_refsource_MISCWEB
News mentions
0No linked articles in our index yet.