Nov 11, 2020 · I think the issue has to do with the Keras backend and the fact that _make_predict_function is a part of the v1 library The text was updated successfully, but these errors were encountered:
11.11.2020 · BentoML/Keras AttributeError: 'Functional' object has no attribute '_make_predict_function' #19 ShayanRiyaz opened this issue Nov 11, 2020 · 2 comments Labels
16.01.2020 · AttributeError: 'Sequential' object has no attribute '_get_distribution_strategy' Ask Question Asked 1 year, 11 months ago. Active 10 months ago. Viewed 10k times ... Unpickled tensorflow model fails to make predictions. 2. Azure ML: ''Sequential' object has no attribute '_distribution_strategy'' Related.
Mar 11, 2019 · Oh, I missed that. Well, I am no expert, but, first of all, your loss and the last layer of your model are incompatible. For 'sparse_categorical_crossentropy' you should have a dense layer with the number of categories and a softmax activation.
Nov 17, 2018 · In the older version (which I think you are in) you still have to declare that the model is functional through Scan(functional_model=True) but in the latest version (0.4.4) this is no longer needed and you can run both Sequential and Functional models as-is. Can you make sure you have the latest version:
Nov 13, 2020 · When trying to pack the KerasModelService with a model, the following error occurs: AttributeError: 'Sequential' object has no attribute '_make_predict_function'. To Reproduce.
Sep 25, 2020 · _make_predict_function() is from old version of Keras. Try to install Tensorflow 1.15 (with Keras 2.2.4-tf) instead of 2.x.x as you are trying to call old function on new model_from_json. Share
24.09.2020 · AttributeError: 'Sequential' object has no attribute '_make_predict_function' Ask Question Asked 1 year, 3 months ago. Active 1 year, 2 months ago. Viewed 2k times ... 'Sequential' object has no attribute '_make_predict_function' Process finished with exit code 1 ...
13.11.2020 · KerasModelArtifact - 'Sequential' Object has no attribute '_make_predict_function' #1242 Closed telescopic opened this issue on Nov 13, 2020 · 2 comments Contributor telescopic commented on Nov 13, 2020 • edited Describe the bug When trying to pack the KerasModelService with a model, the following error occurs:
05.10.2018 · Option 3. # using model's methods model.save ("my_model.h5") # deletes the existing model del model # load the saved model back new_model = load_model ('my_model.h5') Option 1 requires the new_model to be compiled before using. Option 2 and 3 are almost similar in syntax. Codes used from:
Apr 03, 2017 · Add documentation for Model._make_predict_function #13116. Closed. mihajenko pushed a commit to mihajenko/ivis that referenced this issue on May 5, 2020. rm call to private Model._make_predict_function. fae6a90. Discussions on why it isn't recommended to call: * keras-team/keras#6124 * keras-team/keras#13116.
11.03.2019 · Oh, I missed that. Well, I am no expert, but, first of all, your loss and the last layer of your model are incompatible. For 'sparse_categorical_crossentropy' you should have a dense layer with the number of categories and a softmax activation. If you want to use the last layer you have in the example above, you should probably use tf.keras.losses.BinaryCrossentropy().
I can not understand it is giving me the error of 'Sequential' object has no attribute '_get_distribution_strategy'. Package: Keras 2.3.1 Keras-Applications 1.0 ...