Skip to content

block_fp8: per-shard scale-row indexing for TP-interleaved fused qkv

dc76957
Select commit
Loading
Failed to load commit list.
Open

block_fp8: 2D-block FP8 quantized matmul + MoE kernels (DeepSeek-V3 / MiMo) #3600

block_fp8: per-shard scale-row indexing for TP-interleaved fused qkv
dc76957
Select commit
Loading
Failed to load commit list.

Workflow runs completed with no jobs