https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct
#1265
by kkwu - opened
We already did that model 3 month ago:
Static quants: https://huggingface.co/mradermacher/Qwen2.5-VL-7B-Instruct-GGUF
Wighted/imatrix quants: https://huggingface.co/mradermacher/Qwen2.5-VL-7B-Instruct-i1-GGUF
I noticed that for the mmproj vision stack we only did F16 so I requeued to also do Q8_0.
@mradermacher For some reason we in the past named F16 mmproj FP16 so it got reuploaded with the new name and now exists twice. Maybe delete the old one manually if you have time but very low priority.
Yes, we did some manually, and then some with a different name, which turned out to be an evolutionary dead end. The quantize script knows nothing about these old files. I deleted it.