gemini-code-assist[bot] commented on code in PR #18734:
URL: https://github.com/apache/tvm/pull/18734#discussion_r2779360621
##########
tests/python/codegen/test_target_codegen_cuda_fp8.py:
##########
@@ -926,12 +926,12 @@ def _pipeline(mod: tvm.ir.IRModule) -> tvm.ir.IRModule:
seq = tvm.transform.Sequential(
[
tvm.relax.transform.LegalizeOps(),
- tvm.dlight.ApplyDefaultSchedule(
- tvm.dlight.gpu.Matmul(),
- tvm.dlight.gpu.GEMV(),
- tvm.dlight.gpu.Reduction(),
- tvm.dlight.gpu.GeneralReduction(),
- tvm.dlight.gpu.Fallback(),
+ tvm.s_tir.dlight.ApplyDefaultSchedule(
+ tvm.s_tir.dlight.gpu.Matmul(),
+ tvm.s_tir.dlight.gpu.GEMV(),
+ tvm.s_tir.dlight.gpu.Reduction(),
+ tvm.s_tir.dlight.gpu.GeneralReduction(),
+ tvm.s_tir.dlight.gpu.Fallback(),
),
Review Comment:
For consistency and better readability, you can use the `dl` alias that was
imported on line 27 (`from tvm.s_tir import dlight as dl`).
```suggestion
dl.ApplyDefaultSchedule(
dl.gpu.Matmul(),
dl.gpu.GEMV(),
dl.gpu.Reduction(),
dl.gpu.GeneralReduction(),
dl.gpu.Fallback(),
),
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
