lixiaoquan commented on a change in pull request #5429:
URL: https://github.com/apache/incubator-tvm/pull/5429#discussion_r415197766



##########
File path: include/tvm/relay/op_attr_types.h
##########
@@ -81,10 +81,16 @@ using TOpIsStateful = bool;
  */
 using TNonComputational = bool;
 
+enum ShapeDependantKind {
+  kShapeDependantShape = 0,
+  kShapeDependantData = 1,
+  kShapeDependantBoth = 2,

Review comment:
       Thanks for you comment.
   
   I tried two ways to use data tensor's shape directly.
   First, I passed data tensor's shape to _reshape_shape_func but I got this 
error
   ```
   ValueError: All indices are supposed to be constants
   ```
   Second, I tried to construct a tensor by tvm.te.placeholder based on data 
tensor's shape, but it doesn't work because it can't be passed to tvm.lower 
called later.
   




----------------------------------------------------------------
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:
us...@infra.apache.org


Reply via email to