MasterJH5574 opened a new pull request, #14212: URL: https://github.com/apache/tvm/pull/14212
Previously we found that the FX GraphModule captured by torch.dynamo will always return a tuple at the end, even if the Module being traced returns a single object. Unwrapping the unit tuple in that case can help and ease model deployment. Therefore, this PR introduces an option "unwrap_return_unit_tuple" to `from_fx`, to indicate if we caller wants to unwrap the returned unit tuple. `dynamo_subgraph_capture` now will always use True on this option. -- 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]
