README.md

metadata

license: apache-2.0
tags:
  - marlin
  - gptq
  - pruned

A Mistral-7B pruned50 with Marlin Kernel and AutoGPTQ

Please see my tutorial to execute this model:

https://vilsonrodrigues.medium.com/sparse-quantize-and-serving-llms-with-neuralmagic-autogptq-and-vllm-03961b72ec3a