4 people 5426311d91 [Kernel][ROCm][Perf] FlyDSL decode-attention kernel for 4-bit TurboQuant KV cache (#47896) ...

Signed-off-by: root <root@smci355-ccs-aus-m02-09.cs-aus.dcgpu>
Co-authored-by: root <root@smci355-ccs-aus-m02-09.cs-aus.dcgpu>
Co-authored-by: Andreas Karatzas <akaratza@amd.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

2026-08-11 08:45:39 -07:00

..

__init__.py

[Core] Ensure memory is pinned prior to async h2d copy (#45424)

2026-06-20 20:02:24 -07:00

cpu.py

[Bugfix][CPU] Make the Apple Silicon BF16 probe fall back instead of raising (#51627)

2026-08-11 21:35:07 +08:00

cuda.py

[Bugfix][Platform] Stop re-initializing NVML on every device-capability check (fixes #50381) (#50393)

2026-08-07 08:30:00 +00:00

interface.py

[New model] Kimi K3 (#50000)

2026-07-30 03:49:58 -07:00

rocm.py

[Kernel][ROCm][Perf] FlyDSL decode-attention kernel for 4-bit TurboQuant KV cache (#47896)

2026-08-11 08:45:39 -07:00

tpu.py

[Refactor][TPU] Remove torch_xla path and use tpu-inference (#30808)

2026-01-07 16:07:16 +08:00

xpu.py

[XPU] Fix UVA weight offloading (non-pinned-tensor views and static Triton launcher) (#51770)

2026-08-11 21:48:44 +08:00

zen_cpu.py

[ZenCPU] AMD Zen CPU Backend with supported dtypes via zentorch weekly (#39967)

2026-04-18 06:22:37 +00:00