|
EXL2 quants of [Smaug-72B-v0.1](https://huggingface.co/abacusai/Smaug-72B-v0.1). |
|
|
|
Couldn't find a description of the prompt format, but it looks like OpenChat? |
|
|
|
[2.50 bits per weight](https://huggingface.co/turboderp/Smaug-72B-exl2/tree/2.5bpw) (may be a little broken?) |
|
[3.00 bits per weight](https://huggingface.co/turboderp/Smaug-72B-exl2/tree/3.0bpw) |
|
[4.00 bits per weight](https://huggingface.co/turboderp/Smaug-72B-exl2/tree/4.0bpw) |
|
[5.00 bits per weight](https://huggingface.co/turboderp/Smaug-72B-exl2/tree/5.0bpw) |
|
|
|
[measurement.json](https://huggingface.co/turboderp/Smaug-72B-exl2/blob/main/measurement.json) |