Artem Perevedentsev a16133a0f1 [Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338) ...

Signed-off-by: Artem Perevedentsev <aperevedents@nvidia.com>

2026-03-23 00:37:58 -07:00

..

__init__.py

[Perf] add packed recurrent fast path for decode (#36596)

2026-03-12 04:01:57 -07:00

chunk_delta_h.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

chunk_o.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

chunk_scaled_dot_kkt.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

chunk.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

cumsum.py

Update Optional[x] -> x | None and Union[x, y] to x | y (#26633)

2025-10-12 09:51:31 -07:00

fused_recurrent.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

fused_sigmoid_gating.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

index.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

kda.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00

l2norm.py

[Model] Add support for OLMo Hybrid (#32550)

2026-03-05 14:51:06 -05:00

layernorm_guard.py

[Model] Add support for OLMo Hybrid (#32550)

2026-03-05 14:51:06 -05:00

op.py

[Kernel] Accelerate solve_tril with TMA (#26746)

2025-10-20 05:39:02 +00:00

solve_tril.py

[Kernel] Accelerate solve_tril with TMA (#26746)

2025-10-20 05:39:02 +00:00

utils.py

replace with torch.cuda.device with with torch.accelerator.device_index (#36144)

2026-03-11 23:12:57 -07:00

wy_fast.py

[Perf] [Bugfix] Fix Triton autotuning in inference for Qwen3.5 (#37338)

2026-03-23 00:37:58 -07:00