Branch	Bits	lm_head bits	Dataset	Size	Description
8_0	8.0	8.0	Default	9.8 GB	Maximum quality that ExLlamaV2 can produce, near unquantized performance.
6_5	6.5	8.0	Default	8.6 GB	Very similar to 8.0, good tradeoff of size vs performance, recommended.
5_0	5.0	6.0	Default	7.4 GB	Slightly lower perplexity vs 6.5.
4_0	4.0	6.0	Default	6.5 GB	Just under GPTQ equivalent bits per weight.
3_5	3.5	6.0	Default	6.1 GB	Lower quality, only use if you have to.

All VRAM requirements estimated from 16k context. For 32k context add ~2 GB.

Download instructions

With git:

git clone --single-branch --branch 6_5 https://huggingface.co/bartowski/UNA-dolphin-2.6-mistral-7b-dpo-laser-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 UNA-dolphin-2.6-mistral-7b-dpo-laser-exl2:

shell

mkdir UNA-dolphin-2.6-mistral-7b-dpo-laser-exl2
huggingface-cli download bartowski/UNA-dolphin-2.6-mistral-7b-dpo-laser-exl2 --local-dir UNA-dolphin-2.6-mistral-7b-dpo-laser-exl2 --local-dir-use-symlinks False

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

shell

mkdir UNA-dolphin-2.6-mistral-7b-dpo-laser-exl2-6_5
huggingface-cli download bartowski/UNA-dolphin-2.6-mistral-7b-dpo-laser-exl2 --revision 6_5 --local-dir UNA-dolphin-2.6-mistral-7b-dpo-laser-exl2-6_5 --local-dir-use-symlinks False