Hugofernandez/Mistral-7B-v0.1-colab-sharded
TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Nov 28, 2023License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

Hugofernandez/Mistral-7B-v0.1-colab-sharded is a re-sharded version of the Mistral-7B-v0.1 Large Language Model, developed by the Mistral AI Team. This 7 billion parameter generative text model is optimized with Grouped-Query Attention and Sliding-Window Attention, and is specifically re-sharded into 6 parts to facilitate easier loading on machines with limited RAM, such as free Google Colab instances. It outperforms Llama 2 13B on all tested benchmarks, making it suitable for general generative text tasks where resource efficiency is critical.

Loading preview...