mathstral-7b-ov

mathstral-7b-ov is an OpenVino int4 quantized version of Mathstral, providing a very inference implementation, optimized for AI PCs using Intel GPU, CPU and NPU.

mathstral-7b-ov is an instruct-trained model from Mistral with excellent capabilities for answering mathematical oriented problems and questions.

Model Description

• Developed by: mistral
• Quantized by: llmware
• Model type: mistral-7b
• Parameters: 7 billion
• Model Parent: mistralai/Mathstral-7B-v0.1
• Language(s) (NLP): English
• License: Apache 2.0
• Uses: Chat-oriented Math problems
• RAG Benchmark Accuracy Score: NA
• Quantization: int4

Model Card Contact

llmware on github

llmware on hf

llmware website