Share via

Model path for a model folder that has json file and weights.

SHUBHAM JAIN 016 - 70271019016 1 Reputation point
2020-12-20T18:40:00.887+00:00

I am trying to create a score.py file for model deployment. In the folder where the model is present, there are two files present. One is the JSON model file another is a .h5 file which has weights of the model. how do I configure the file in this case in the init function of the score.py file?

Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
2,729 questions
0 comments

1 answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. romungi-MSFT 43,696 Reputation points Microsoft Employee
    2020-12-21T10:34:00.973+00:00

    @SHUBHAM JAIN 016 - 70271019016 For the above scenario you need to save the model as a single file HDF5 file and then register the model using model.register() to register the model with the workspace and load the model in the scoring script using load_model.