Transformers now runs llama.cpp quants

Hugging Face Transformers can now run models quantized with llama.cpp.

Hugging Face announced that its Transformers library can now run models quantized with llama.cpp. The post is an inference-tooling update for developers using Transformers. The available text does not describe implementation details, supported formats beyond llama.cpp quants, or any security issue.

• Transformers can now run llama.cpp quantized models.
• The change is an inference and developer-tooling update.
• No model launch or security incident is described.