HongHongHongL opened a new pull request, #15893: URL: https://github.com/apache/tvm/pull/15893
In the documentation for pad_einsum, the example uses > sch.pad_einsum(block, [0, 1, 1]) However, it will cause > Error message: The padding for the block tir.Block#0 are invalid. It should be a list of 3 positive integers. On a block with trivial binding, this primitive pads the iteration domain of the block by the given padding factors, for example, 127 -> 128, 132 -> 144 when padding factor is 16. So the example should be > sch.pad_einsum(block, [padding factor, padding factor, padding factor]) In the test file tests/python/relax/test_optimize_layout_transform.py, this **fused_mode** is unused. I think it should be **fused_mod**. -- 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]
