CVE-2025-5197
Last modified
CVE-2025-5197 is a vulnerability of currently unknown severity. A Regular Expression Denial of Service (ReDoS) vulnerability exists in the Hugging Face Transformers library, specifically in the `convert_tf_weight_name_to_pt_weight_name()` function. This function, responsible for converting TensorFlow weight names to PyTorch format, uses a regex pattern `/[^/]*___([^/]*)/` that can be exploited to cause excessive CPU consumption through crafted input strings due to catastrophic backtracking. EPSS estimates a 0.36% chance of exploitation in the next 30 days.
Description
A Regular Expression Denial of Service (ReDoS) vulnerability exists in the Hugging Face Transformers library, specifically in the `convert_tf_weight_name_to_pt_weight_name()` function. This function, responsible for converting TensorFlow weight names to PyTorch format, uses a regex pattern `/[^/]*___([^/]*)/` that can be exploited to cause excessive CPU consumption through crafted input strings due to catastrophic backtracking. The vulnerability affects versions up to 4.51.3 and is fixed in version 4.53.0. This issue can lead to service disruption, resource exhaustion, and potential API service vulnerabilities, impacting model conversion processes between TensorFlow and PyTorch formats.
Metrics
Weakness Enumeration
Affected Software
| Vendor | Product | Versions |
|---|---|---|
| Huggingface | Transformers | < 4.53.0 |
References
- https://huntr.com/bounties/3f8b3fd0-166b-46e7-b60f-60dd9d2678bfExploit, Issue Tracking, Patch, Third Party Advisory
- https://huntr.com/bounties/3f8b3fd0-166b-46e7-b60f-60dd9d2678bfExploit, Issue Tracking, Patch, Third Party Advisory
Timeline
- Published
- Last Modified
- Status
- Analyzed
Frequently Asked Questions
What is CVE-2025-5197?
How severe is CVE-2025-5197?
How do I fix CVE-2025-5197?
Are you affected by CVE-2025-5197?
Run a free Strix scan to check your systems for this vulnerability.
Scan your code nowSource: NVD / NIST