ConvolutedDog

Other compilation error messages include: ```bash ../test/EQueue/conv2d.mlir:6:10: error: 'equeue.create_mem' op requires attribute 'data_bit' %1 = "equeue.create_mem"() {data = "f32", name = "mem", shape = dense : tensor, type = "RegisterFile"}...

@ziyuhuang123 You can try to compile cutlass_fpA_intB_gemm independently. If it is a problem with the machine environment, it is likely to fail.

I have checked the following minimal reproduction example, and it runs without errors in my environment. ```python import torch import torch.nn as nn from tvm.relax.frontend.torch import from_exported_program input_size = 28...

```bash UserWarning: The given buffer is not writable, and PyTorch does not support non-writable tensors. This means you can write to the underlying (supposedly non-writable) buffer using the tensor. You...

The error occurs because the `expand_shape` parameter you passed to the main function is `tvm.nd.array([1, 1, 4, 4])`, but TVM's ***Exapnd*** node requires that the `expand_shape` passed in is the...

Yes this is a good suggestion, it needs to be better documented. Maybe there should be an operator documentation similar to ONNX or PyTorch.

Don't use None for pragma_unroll_explicit. Use 1/0 or True/False instead. https://github.com/apache/tvm/blob/8ab96af40e9f966f579436ca801e08d41b569516/src/tir/transforms/unroll_loop.cc#L111-L112