bartowski/

neural-chat-7b-v3-2-exl2

Using turboderp's ExLlamaV2 v0.0.10 for quantization.

Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.

Conversion was done using wikitext-103-raw-v1-test.parquet as calibration dataset.

Default arguments used except when the bits per weight is above 6.0, at that point the lm_head layer is quantized at 8 bits per weight instead of the default 6.

Original model: https://huggingface.co/Intel/neural-chat-7b-v3-2/

With git:

With huggingface hub (credit to TheBloke for instructions):

To download the main (only useful if you only care about measurement.json) branch to a folder called neural-chat-7b-v3-2-exl2:

To download from a different branch, add the --revision parameter: