New attempt

Update onnx files again after new fix from Xenova

Update model with fix from xenova

Replace with conversion from transformers.js/scripts/convert.py

Rename to int8

Add quant config files

Add int8 quants

Upload the model

initial commit