3 people 9709b7513b [feat] Port NVFP4 QAT/QAD to modular train framework (#1619) ...

Co-authored-by: Peiyuan Zhang <email>

Co-authored-by: Peiyuan <a>

2026-07-18 15:01:09 -07:00

..

attn_qat_train.py

[feat] Port NVFP4 QAT/QAD to modular train framework (#1619)

2026-07-18 15:01:09 -07:00

block_sparse_attn_triton.py

[bugfix] fastvideo-kernel: fix VSA Triton padding NaNs and support q/kv length mismatch (#1094)

2026-02-13 12:39:49 -08:00

fused_attention.py

[kernel] QAD 5090: Add Attn-QAT training Triton kernels (11/12) (#1460)

2026-06-17 13:35:12 -07:00

fused_compress_topk.py

[perf] optimize compress & topk kernel (#1421)

2026-06-23 18:51:22 +00:00

index.py

[Perf] Skip bool-mask round-trip in block-sparse VSA attention (#1243)

2026-04-26 15:14:37 -07:00

nvfp4_utils.py

[kernel] QAD 5090: Add Attn-QAT training Triton kernels (11/12) (#1460)

2026-06-17 13:35:12 -07:00

quant_utils.py

[kernel] QAD 5090: Add Attn-QAT training Triton kernels (11/12) (#1460)

2026-06-17 13:35:12 -07:00

sla_triton.py

[New Model] Turbodiffusion (#971)

2026-01-02 17:55:56 -06:00

st_attn_triton.py

[fix]: fix STA trition kernel for AMD RDNA archs (#969)

2025-12-29 14:25:32 -06:00