Update Jan 27: This model was done before some config updates from internlm, please try the new one here and report any differences: https://huggingface.co/bartowski/internlm2-chat-20b-llama-exl2/
Special thanks to Charles Goddard for the conversion script to create llama models from internlm
Exllama v2 Quantizations of internlm2-chat-20b-llama
Using turboderp's ExLlamaV2 v0.0.11 for quantization.
The "main" branch only contains the measurement.json, download one of the other branches for the model (see below)
Each branch contains an individual bits per weight, with the main one containing only the meaurement.json for further conversions.
Conversion was done using the default 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/internlm/internlm2-chat-20b
Download instructions
With git:
git clone --single-branch --branch 4_0 https://huggingface.co/bartowski/internlm2-chat-20b-llama-exl2-old
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 internlm2-chat-20b-llama-exl2:
mkdir internlm2-chat-20b-llama-exl2
huggingface-cli download bartowski/internlm2-chat-20b-llama-exl2-old --local-dir internlm2-chat-20b-llama-exl2 --local-dir-use-symlinks False
To download from a different branch, add the --revision parameter:
mkdir internlm2-chat-20b-llama-exl2
huggingface-cli download bartowski/internlm2-chat-20b-llama-exl2-old --revision 4_0 --local-dir internlm2-chat-20b-llama-exl2 --local-dir-use-symlinks False