Downloads · 30 days
0
fastsafetensors
n4u/fastsafetensors
fastsafetensors is a machine learning model from n4u. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Security proof-of-concept for a Denial-of-Service in foundation-model-stack/fastsafetensors (HEAD 8d35826). Reported via huntr's Model File Vulnerability program.
Access
Public
Updated Jul 6, 2026
Parameters
1
80 B on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
Other1.5 KB · 61%
At a glance
- Access
- Public
- Created
- Jul 6, 2026
- Updated
- Jul 6, 2026
- SHA
- da2ab715
Model card
From the Hugging Face model README
fastsafetensors header DoS — PoC model file
Security proof-of-concept for a Denial-of-Service in
foundation-model-stack/fastsafetensors
(HEAD 8d35826). Reported via huntr's Model File Vulnerability program.
model.safetensors is a ~80-byte safetensors file whose header carries a tensor
entry with an invalid dtype. Loading it crashes the parser with an uncaught
ValueError, before any tensor data is read. This affects any app using
fastsafetensors, including vLLM / SGLang started with --load-format fastsafetensors.
Reproduce
from fastsafetensors.common import SafeTensorsMetadata
from fastsafetensors.frameworks import get_framework_op
SafeTensorsMetadata.from_file("model.safetensors", get_framework_op("pytorch"))
# -> ValueError: 'NOTADTYPE' is not a valid DType (uncaught -> loader crash)
- Created
- Jul 6, 2026
- Updated
- Jul 6, 2026