Add Mixtral and better groupwise quantization

microsoft / TransformerCompression

For releasing code related to compression methods for transformers, accompanying our publications

MIT License

354 stars 31 forks source link

Closed nailimixaM closed 2 months ago