taomiao opened a new issue, #16389: URL: https://github.com/apache/tvm/issues/16389
Thanks for participating in the TVM community! We use https://discuss.tvm.ai for any general usage questions and discussions. The issue tracker is used for actionable items such as feature proposals discussion, roadmaps, and bug tracking. You are always welcomed to post on the forum first :smile_cat: Issues that are inactive for a period of time may get closed. We adopt this policy so that we won't lose track of actionable issues that may fall at the bottom of the pile. Feel free to reopen a new one if you feel there is an additional problem that needs attention when an old one gets closed. ### Expected behavior work well ### Actual behavior ``` test_nonzero_numpy.py:45: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ../../../../python/tvm/relay/frontend/pytorch.py:5418: in from_pytorch outputs = converter.convert_operators(operator_nodes, outputs, ret_name) ../../../../python/tvm/relay/frontend/pytorch.py:4528: in convert_operators unpacked = _unpack_tuple(inputs[0]) ../../../../python/tvm/relay/frontend/pytorch.py:5137: in _unpack_tuple elif isinstance(tup.type_annotation, TupleType): _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ self = CallNode(Op(argwhere), [Var(input1, ty=TensorType([2, 10], bool))], (nullptr), []) name = 'type_annotation' def __getattr__(self, name): # specially check handle since # this is required for PackedFunc calls if name == "handle": raise AttributeError("handle is not set") try: return _ffi_node_api.NodeGetAttr(self, name) except AttributeError: > raise AttributeError(f"{type(self)} has no attribute {name}") from None E AttributeError: <class 'tvm.relay.expr.Call'> has no attribute type_annotation ``` ### Environment os: windows 10 python: 3.9 pytorch: 2.0 tvm: main branch ### Steps to reproduce ``` from torch import nn import torch import tvm class NonZeroModule(nn.Module): """Module that performs nonzero""" def __init__(self): super().__init__() def forward(self, x, mask): mask_index = torch.nonzero(mask, as_tuple=True) x[mask_index] = torch.ones_like(x[mask_index]) return x def test_pytorch_nonzero(): model = NonZeroModule() x = torch.zeros((2, 10), dtype=torch.float32) mask = torch.randint(0, 2, (2, 10)).bool() with torch.no_grad(): traced_torch_model = torch.jit.trace(model, (x, mask)) import_input = [("input0", (2, 10)), ("input1", (2, 10))] relay_model_ir, relay_model_params = tvm.relay.frontend.from_pytorch( traced_torch_model, import_input ) ``` ### Triage Please refer to the list of label tags [here](https://github.com/apache/tvm/wiki/Issue-Triage-Labels) to find the relevant tags and add them below in a bullet format (example below). * needs-triage -- 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]
