flashrt/grouped-moe-gemm
• Updated • 2 • 2
AI & ML interests
training & inference infra
Recent Activity
flashrt/fp8-prefill-attention-blackwell
• Updated • 2 • 1
flashrt/small-matrix-cholesky
• Updated • 1 • 3
flashrt/smallm-ffn-megakernels-blackwell
• Updated • 1 • 1
flashrt/fa2-seqused-runtime
• Updated • 2 • 1
flashrt/flashrt-fp8-ffn
• Updated • 32 • 5
flashrt/flashrt-fp8-swiglu-ffn
• Updated • 37 • 5
flashrt/weight-only-ffn
• Updated • 6 • 1
flashrt/fp4-gemm
• Updated • 19 • 1
flashrt/int4-blackwell
• Updated • 26 • 5
flashrt/flashrt-flex-attention-train
• Updated • 2
flashrt/flashrt-adarms-train
• Updated • 16 • 2
flashrt/flashrt-vocab-ce-train
• Updated • 12 • 1
flashrt/flashrt-rope-train
• Updated • 16 • 1
flashrt/flashrt-siglip-fwd-fusion
• Updated • 13 • 2
flashrt/flashrt-qkv-epilogue-train
• Updated • 12 • 1
flashrt/transformer-layout-primitives
• Updated • 9 • 2
flashrt/int8-transformer-primitives
• Updated • 9 • 2
flashrt/speculative-draft-primitives
• Updated • 9 • 2
flashrt/flashrt-gemm-epilogues
• Updated • 57 • 3
flashrt/sageattention2-blackwell
• Updated • 22 • 1
flashrt/gated-delta-attention
• Updated • 13 • 3
flashrt/fp8-kv-attention
• Updated • 12 • 1
flashrt/adaptive-layernorm-producers
• Updated • 11 • 1
flashrt/causal-conv1d-state
• Updated • 15 • 2
flashrt/fp8-gemm
• Updated • 12 • 2
flashrt/linear-attention-seq-state
• Updated • 48 • 2
flashrt/grouped-moe-gemv
• Updated • 11 • 2
flashrt/bf16-linear-gemv
• Updated • 18 • 1
flashrt/transformer-fused-ops
• Updated • 48 • 1