CVE-2024-34359

llama-cpp-python is the Python bindings for llama.cpp. `llama-cpp-python` depends on class `Llama` in `llama.py` to load `.gguf` llama.cpp or Latency Machine Learning Models. The `__init__` constructor built in the `Llama` takes several parameters to configure the loading and running of the model. Other than `NUMA, LoRa settings`, `loading tokenizers,` and `hardware settings`, `__init__` also loads the `chat template` from targeted `.gguf` 's Metadata and furtherly parses it to `llama_chat_format.Jinja2ChatFormatter.to_chat_handler()` to construct the `self.chat_handler` for this model. Nevertheless, `Jinja2ChatFormatter` parse the `chat template` within the Metadate with sandbox-less `jinja2.Environment`, which is furthermore rendered in `__call__` to construct the `prompt` of interaction. This allows `jinja2` Server Side Template Injection which leads to remote code execution by a carefully constructed payload.
Configurations

No configuration.

History

14 May 2024, 15:38

Type Values Removed Values Added
New CVE

Information

Published : 2024-05-14 15:38

Updated : 2024-05-14 16:12


NVD link : CVE-2024-34359

Mitre link : CVE-2024-34359

CVE.ORG link : CVE-2024-34359


JSON object : View

Products Affected

No product.

CWE
CWE-76

Improper Neutralization of Equivalent Special Elements