search

aredden / flux-fp8-api

Flux diffusion model implementation using quantized fp8 matmul & remaining layers use faster half precision accumulate, which is ~2x faster on consumer devices.
Apache License 2.0
109 stars 12 forks source link

update README #1

Closed dsingal0 closed 3 weeks ago

aredden commented 3 weeks ago

Ah nice catch