Vulnerability Description
LLama Factory enables fine-tuning of large language models. Prior to version 1.0.0, a critical vulnerability exists in the `llamafy_baichuan2.py` script of the LLaMA-Factory project. The script performs insecure deserialization using `torch.load()` on user-supplied `.bin` files from an input directory. An attacker can exploit this behavior by crafting a malicious `.bin` file that executes arbitrary commands during deserialization. This issue has been patched in version 1.0.0.
CVSS Score
MEDIUM
Affected Products
| Vendor | Product | Versions |
|---|---|---|
| Hiyouga | Llama-Factory | < 1.0.0 |
Related Weaknesses (CWE)
References
- https://github.com/hiyouga/LLaMA-Factory/commit/2989d39239d2f46e584c1e1180ba46b9Patch
- https://github.com/hiyouga/LLaMA-Factory/security/advisories/GHSA-f2f7-gj54-6vpvExploitVendor Advisory
- https://github.com/hiyouga/LLaMA-Factory/security/advisories/GHSA-f2f7-gj54-6vpvExploitVendor Advisory
FAQ
What is CVE-2025-46567?
CVE-2025-46567 is a vulnerability with a CVSS score of 6.1 (MEDIUM). LLama Factory enables fine-tuning of large language models. Prior to version 1.0.0, a critical vulnerability exists in the `llamafy_baichuan2.py` script of the LLaMA-Factory project. The script perfor...
How severe is CVE-2025-46567?
CVE-2025-46567 has been rated MEDIUM with a CVSS base score of 6.1/10. Review the CVSS metrics above for detailed severity breakdown.
Is there a patch for CVE-2025-46567?
Check the references section above for vendor advisories and patch information. Affected products include: Hiyouga Llama-Factory.