codeislife99 commented on a change in pull request #7442:
URL: https://github.com/apache/tvm/pull/7442#discussion_r576394193
##########
File path: python/tvm/relay/op/_transform.py
##########
@@ -445,6 +465,44 @@ def argwhere_shape_func(attrs, inputs, out_ndims):
_reg.register_shape_func("scatter_add", False, elemwise_shape_func)
+@script
+def _sparse_fill_empty_rows_shape_func(sparse_indices, dense_shape):
+
+ new_sparse_indices_shape = output_tensor((2,), "int64")
+ new_sparse_values_shape = output_tensor((1,), "int64")
+ empty_row_indicator_shape = output_tensor((1,), "int64")
+ num_dense_rows = int64(dense_shape[0])
+
+ if int64(sparse_indices.shape[0]) == int64(0):
Review comment:
I have added test cases for fully dynamic input shapes and it works.
Were you talking about something else ?
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
