amd-shark
/

sdxl-quant-int8

Inference Endpoints

Model card Files Files and versions Community

sdxl-quant-int8 / all_linear_sym_7_5_noconvout

4 contributors

History: 2 commits

GiusFra's picture

MI250 QKV fused and all linear layers sym, guidance scale 7.5, no conv_out

d30c97c verified 4 months ago

params.safetensors

5.14 GB
LFS

MI250 QKV fused and all linear layers sym, guidance scale 7.5, no conv_out 4 months ago
quant_params.json

62.1 MB
LFS

MI250 QKV fused and all linear layers sym, guidance scale 7.5, no conv_out 4 months ago