CVE-2025-5197

Source
https://nvd.nist.gov/vuln/detail/CVE-2025-5197
Import Source
https://storage.googleapis.com/cve-osv-conversion/osv-output/CVE-2025-5197.json
JSON Data
https://api.osv.dev/v1/vulns/CVE-2025-5197
Aliases
Related
Published
2025-08-06T12:15:26Z
Modified
2025-08-06T20:37:48Z
Summary
[none]
Details

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.

References

Affected packages

Git / github.com/huggingface/transformers

Affected ranges

Type
GIT
Repo
https://github.com/huggingface/transformers
Events
Introduced
0 Unknown introduced commit / All previous commits are affected
Fixed