forked from PaddlePaddle/FastDeploy
-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Description
Add unit tests for speculate_rebuild_append_padding.
⚠️ Important: XPU-only op — no GPU CUDA kernel
This op only has an XPU implementation. There is no GPU CUDA source in custom_ops/gpu_ops/.
XPU source: custom_ops/xpu_ops/src/ops/mtp/speculate_rebuild_append_padding.cc
XPU test reference: custom_ops/xpu_ops/test/test_speculate_rebuild_append_padding.py
Options:
- Write a GPU-side test if the op can be invoked through Python bindings on GPU
- Write an XPU test following the existing XPU test pattern
- Write a CPU-level reference test if possible
Study the XPU test at custom_ops/xpu_ops/test/test_speculate_rebuild_append_padding.py for the expected behavior.
Test file: tests/operators/test_speculate_rebuild_append_padding.py
Branch: task/058-spec-rebuild-append-test
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels