CVE-2025-46567

LLaMA-Factory Allows Arbitrary Code Execution via Unsafe Deserialization in Ilamafy_baichuan2.py

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.

Affected

hiyouga · LLaMA-Factory

CVSS v3.1 NVD

7.8 HIGHCVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H

CVSS v3.1 GITHUB

6.1 MEDIUMCVSS:3.1/AV:L/AC:L/PR:L/UI:R/S:U/C:H/I:L/A:L

EPSS

0.29% chance of exploitation in the next 30 days, 20th percentile

Published

2025-05-01

Updated

2025-05-02

Proof-of-concept exploits (2)

• hiyouga/LLaMA-Factory/security/advisories/GHSA-f2f7-gj54-6vpv
• advisories/GHSA-f2f7-gj54-6vpv

Security advisories

• hiyouga/LLaMA-Factory/commit/2989d39239d2f46e584c1e1180ba46b9768afb2a

References

• CVE Record
• NVD entry