[GitHub] [incubator-tvm] Haoran-Young commented on a change in pull request #6022: Add support for checking big ONNX models

[GitBox <gi...@apache.org>]

Haoran-Young commented on a change in pull request #6022:
URL: https://github.com/apache/incubator-tvm/pull/6022#discussion_r452643452



##########
File path: python/tvm/relay/frontend/onnx.py
##########
@@ -2263,6 +2265,14 @@ def from_onnx(model,
     dtype : str or dict of str to str
         The input types to the graph
 
+    big_model : bool
+        Whether the size of the input model is larger than 2GB
+
+    model_path : str, optional
+        The path of input model, and the external data
+        files need under the same directory
+        Must be set if 'big_model' is True

Review comment:
       @liangfu Thanks for your advice.
   As I commented on @adityaatluri 's advice, I think it is better to introduce two arguments, 'model_path' and 'external_data_path,' to replace 'model,' 'big_model,' 'model_path.' in my commit.
   The newest ONNX API can load a model via model path and external data files' path (external files' path is needed if they are not under the same directory with the model file) without any other arguments.




----------------------------------------------------------------
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:
users@infra.apache.org