llama-2-chat-onnx

llama-2-chat-onnx is an ONNX int4 quantized version of Llama-2-Chat, providing a fast, small inference implementation, optimized for AI PCs and Windows x86-64 architectures.

llama-2-chat is the official chat finetune of the classic Llama 2 model, one of the most most iconic (and still one the best) 7B instruct trained models.

Model Description

• Developed by: meta-llama
• Quantized by: llmware
• Model type: llama2
• Parameters: 7 billion
• Model Parent: meta-llama/Llama-2-7b-chat-hf
• Language(s) (NLP): English
• License: Llama-2 Community License
• Uses: Fact-based question-answering
• RAG Benchmark Accuracy Score: NA
• Quantization: int4

Model Card Contact

llmware on github

llmware on hf

llmware website