YXY-0922 opened a new pull request, #17171:
URL: https://github.com/apache/tvm/pull/17171
For a `conv2d` workload that requires padding:
### Details
Test Script: `test_padding_
Location: `tests/python/meta_schedule` directory
```Python
mod = te.create_prim_func(
te_workload.conv2d_nhwc(
1,
224,
224,
3,
64,
7,
2,
3,
in_dtype="float16",
out_dtype="float32",
)
)
```
This configuration sets up a conv2d workload where the product of reduce
axes is 147. Therefore, padding is needed when applying the rule
multi_level_tiling_tensorcore.
--
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]
