New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
'Missing required flag: input_file' when using TOCO with SavedModel #18557
Comments
The issue is likely that the version of TensorFlow you are using does not have SavedModel support. Try and build with the latest version of TensorFlow. Instructions are available here: https://www.tensorflow.org/install/install_sources. |
Nagging Assignee @gargn: It has been 14 days with no activity and this issue has an assignee. Please update the label and/or status accordingly. |
Nagging Assignee @gargn: It has been 30 days with no activity and this issue has an assignee. Please update the label and/or status accordingly. |
I've updated to 1.8 and now rather than throwing an exception you get: TOCO from pip install is currently not working on command line. Closing out issue. |
There was some issue in version 1.8 and it did not install TOCO. After updating to version 1.10 it installed TOCO and also following error got resolved. |
after updating to tensorflow from 17 to 1.10 (!pip install --upgrade "tensorflow==1.10.*") its resolved my issue |
Describe the problem
I have a SavedModel trained and I am trying to use TOCO to convert it to tflite format. When I use the above command, given as an example in the docs here I get prompted for the input_format. Specifying this as TENSORFLOW_GRAPHDEF gives a new error for output_format. Specifying this as TFLITE now prompts for required parameter input_file. As this is a SavedModel and not a GraphDef, I can't provide the input_file and in fact, shouldn't have to as savedmodel_directory has already been provided.
The model directory 1523530083 contains saved_model.pb and a variables folder. There are no assets.
The text was updated successfully, but these errors were encountered: