Radial Attention: O(nlogn) Attention for Long Video Generation with 2-4× Speeduphanlab.mit.edu·6 pts·lmxyy·1
SVDQuant+NVFP4: 4× Smaller, 3× Faster FLUX with 16-bit Quality on Blackwell GPUshanlab.mit.edu·52 pts·lmxyy·10
SVDQuant: 4-Bit Quantization Powers 12B Flux on a 16GB 4090 GPU with 3x Speeduphanlab.mit.edu·179 pts·lmxyy·65