NVIDIA TRT-LLM
- < v1.2
A vulnerability exists in NVIDIA TensorRT-LLM on all platforms, specifically within RPC testing. This vulnerability allows an attacker to cause unsafe deserialization of data. Exploiting this issue could lead to unauthorized code execution, denial of service, data tampering, and information disclosure.
Exploitation of this vulnerability could result in unauthorized code execution, a denial of service, unauthorized data modification, and unauthorized information disclosure.
Our algorithm analyzes dozens of metrics to generate these 8 key vulnerability categories, which are then combined to calculate the overall risk score.