wang7776/Mistral-7B-Instruct-v0.2-sparsity-20-v0.1
TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:8kPublished:Jan 17, 2024License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

The wang7776/Mistral-7B-Instruct-v0.2-sparsity-20-v0.1 is a 7 billion parameter instruction-tuned language model, derived from Mistral-7B-Instruct-v0.2, which has been pruned to 20% sparsity using the Wanda method. This pruning technique maintains competitive performance without retraining or weight updates. It is designed for efficient inference while retaining the instruction-following capabilities of its base model, making it suitable for various natural language processing tasks.

Loading preview...