RCE in Hiyouga Llama-Factory

CVE-2025-53002

LLaMA-Factory is a tuning library for large language models. A remote code execution vulnerability was discovered in LLaMA-Factory versions up to and including 0.9.3 during the LLaMA-Factory training process. This vulnerability arises because the `vhead_file` is loaded without proper safeguards, allowing malicious attackers to execute arbitrary malicious code on the host system simply by passing a malicious `Checkpoint path` parameter through the `WebUI` interface. The attack is stealthy, as the victim remains unaware of the exploitation. The root cause is that the `vhead_file` argument is loaded without the secure parameter `weights_only=True`. Version 0.9.4 contains a fix for the issue.

Vulnerability class: RCE (Remote Code Execution)

EPSS: 0.011 (60.9th percentile) — read the EPSS interpretation.

CVSS v3 metric

CVSS v3 base score 8.3 (High). Vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:H.

Affected products

Weakness classification (CWE)

Public proof-of-concept exploits

References

Frequently asked questions

What is CVE-2025-53002?
CVE-2025-53002 is a high-severity vulnerability in Hiyouga Llama-Factory, classified under Code Injection. CVSS score: 8.3/10. Published 2025-06-26.
How severe is CVE-2025-53002?
High severity. CVSS v3 base score is 8.3 out of 10.
Is CVE-2025-53002 known to be exploited?
1 public proof-of-concept repositories are indexed. Not currently listed in the CISA KEV catalog.