elvin-nnov edited a comment on pull request #7903:
URL: https://github.com/apache/tvm/pull/7903#issuecomment-825181927


   > @elvin-nnov actually `debug_get_output` by itself doesn't execute the 
network and only returns the output of a tensor that was already captured by 
`self. _get_output_by_layer()` and recorded in 
`self.debug_datum._output_tensor_list`.
   > So assuming we fix the _run_debug, then we can use `debug_get_output` 
function.
   
   here 
https://github.com/apache/tvm/blob/main/src/runtime/graph_executor/debug/graph_executor_debug.cc#L292
 `debug_get_output` calls `DebugGetNodeOutput` and in its turn 
`DebugGetNodeOutput` executes required number of ops here 
https://github.com/apache/tvm/blob/main/src/runtime/graph_executor/debug/graph_executor_debug.cc#L225
   


-- 
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]


Reply via email to