biondizzle 5c94dbbc37 Fix CuTeDSL from_dlpack device mismatch in CUDA graph capture (v2) ...

Previous fix (set_device) caused 'Capture must end on the same stream'.
New fix: wrap tensor in _DLPatchTensor during graph capture, which forces
dl_device in __dlpack__ to bypass the device check without changing the stream.

This enables CUDA graph capture on all 8 GPUs, not just cuda:0.

2026-06-03 20:54:18 +00:00

..

__init__.py

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00

custom_ops.py

Stage E: head-packed MQA/GQA, batch dim, custom_op, integration API

2026-05-27 15:15:03 +00:00

gemm_runner.py

Fix CuTeDSL from_dlpack device mismatch in CUDA graph capture (v2)

2026-06-03 20:54:18 +00:00

layouts.py

Restructure: cutedsl/ -> dsv4/ with proper layering

2026-05-21 17:30:44 +00:00

quantize.py

CUDA graph: Fix gsa broadcast — contiguous for prefill, reshape for decode

2026-06-03 18:08:18 +00:00

rope_cuda.py

fix: rope_cuda path — kernels/cuda not ops/cuda

2026-06-02 09:06:36 +00:00

router.py

router: catch CuTeDSL warmup failures fast, don't let MLIR errors slow down init

2026-06-01 00:00:07 +00:00