feat(wan): fast inference optimizations (fused cross-attention, token patch embed, shard-major A2A, lane padding) - #491

feat(wan): fast inference optimizations (fused cross-attention, token patch embed, shard-major A2A, lane padding)#491

Perseus14 wants to merge 1 commit into

feat/wan-custom-kernelsAI-Hypercomputer/maxdiffusion:feat/wan-custom-kernelsfrom

feat/wan-fast-inference-optimizationsAI-Hypercomputer/maxdiffusion:feat/wan-fast-inference-optimizationsCopy head branch name to clipboard