mehrdadh opened a new pull request #7903: URL: https://github.com/apache/tvm/pull/7903
This PR: - changes graph executor debugger to have two modes {timing, accuracy}. These two modes are different in a sense that in timing we measure the performance of each layer, however in accuracy we look for the correctness of each layer. - debug_mode=timing does not dump parameter anymore since the parameter placeholders are override after executing the entire model. - debug_mode=accuracy runs the model up to each layer and saves the output tensor. -- 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