Simple Loading of the You Tube Recommendation Model

I have read through the readme which contains a load using flask. I am currently saving the model as such https://github.com/massquantity/LibRecommender/issues/10#issuecomment-707470504

>>> from libreco.utils import save_info, save_model_tf_serving
>>> ytb_ranking = YouTubeRanking(...)
>>> ytb_ranking.fit(...)  # train model
>>> path = "tf_model"  # specify model saving directory
>>> save_model_tf_serving(path, ytb_ranking, "youtube")  # save tf model

and would like to know if there is a simple way to this model without deploying with the flask?

At the moment, Im using the following code:

frozen_graph_filename='C:\\Projects\\RecommendationModel\\data\\rec_data_mainfeat_alldeepFM_0\\1\\saved_model.pb'

with tf.gfile.GFile(frozen_graph_filename, "rb") as f:
    graph_def = tf.GraphDef()
    graph_def.ParseFromString(f.read())

graph = tf.get_default_graph()
tf.import_graph_def(graph_def, name="prefix")

However, it results in the following error:

    graph_def.ParseFromString(f.read())

DecodeError: Error parsing message

[massquantity]

No, it doesn’t. I’ve come up with some ideas about it, and maybe they can be implemented in the next version.

[massquantity]

I have seen a lot of ways about saving and loading TensorFlow models on the internet. Honestly I’m also confused right now because some are about tf1 and some are about tf2. I haven’t tested it yet, but you can try the tf.saved_model.loader API.

I decide to do some research and add some methods to solve this save/load problem. But it will take a few days.