https://smartchunks.com/shn5-lossless-bf16-61gb-moe-16gb-gpu/Smart Chunksen2026-11-14T09:04:36-05:00Lossless.shn5 Shrinks BF16 Models 1.5x to Fit a 61GB MoE on a 16GB GPU https://smartchunks.com/flashinfer-v0-7-4-rubin-moe-batch-invariant-mode/Smart Chunksen2026-11-13T13:44:24-05:00FlashInfer v0.7.4 Adds Rubin MoE Kernels and a Batch-Invariant Mode