Exllama v2 Quantizations of openchat_3.5
Using turboderp's ExLlamaV2 v0.0.7 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 WizardLM_evol_instruct_V2_196k.parquet as calibration dataset.
Original model: https://huggingface.co/openchat/openchat_3.5
Download instructions
With git:
git clone --single-branch --branch 4.0 https://huggingface.co/bartowski/openchat_3.5-exl2
With huggingface hub (credit to TheBloke for instructions):
pip3 install huggingface-hub
To download the main (only useful if you only care about measurement.json) branch to a folder called openchat_3.5-exl2:
mkdir openchat_3.5-exl2
huggingface-cli download bartowski/openchat_3.5-exl2 --local-dir openchat_3.5-exl2 --local-dir-use-symlinks False
To download from a different branch, add the --revision parameter:
mkdir openchat_3.5-exl2
huggingface-cli download bartowski/openchat_3.5-exl2 --revision 4.0 --local-dir openchat_3.5-exl2 --local-dir-use-symlinks False