Unpruned FP8 Text Encoder Request
#11
by makisekurisu-jp - opened
Request to release an unpruned FP8‑quantized version of the text encoder.
Hey bro, I saw you’ve done a lot of quantization work in your model repository.
Could you also help out by making an unpruned W4A8‑quantized version of the Minimax H3 model?
https://huggingface.co/Kijai/MiniMax-H3-experimental/discussions/27