metadata

license: apache-2.0
tags:
  - nougat
  - quant
  - avx512_vnni

nougat-small-onnx-quant_avx512_vnni

This was quantized from pszemraj/nougat-small-onnx using the --avx512_vnni flag. You need to have a processor with avx512_vnni instructions for this to work properly.

per_channel is set to True for better accuracy
Usage is the same as: https://huggingface.co/pszemraj/nougat-small-onnx
verify that you have the requirements by checking for results with lscpu | grep avx512_vnni