CODEWITHSUNDEEP
tensorflowtf.keras
Zil Chaudhary
Zil Chaudhary

Reputation: 1

fit() got an unexpected keyword argument 'generator'

epochs = 150
callbacks_list=[ModelCheckpoint(save_best_only=False,filepath=checkpoint_path),TensorBoard(log_dir='logs')]
history = model.fit(generator=gen_tr,steps_per_epoch=200,
                                  epochs=epochs,
                                  validation_data=gen_val,
                                  validation_steps=1,
                                  callbacks = callbacks_list )

Error message- TypeError : fit() got an unexpected keyword argument 'generator'

In documentation it is written that Model.fit supports generator now, but still it shows error.

enter image description here

Upvotes: 0

Views: 6051

Answers (2)

user7603816
user7603816

Reputation: 21

Model.fit() takes in the generator. But remove keyword generator.

model.fit(generate_batch(X_train, y_train, batch_size = batch_size),...)

Upvotes: 2

user11530462
user11530462

Reputation:

From comments

epochs = 150
callbacks_list=[ModelCheckpoint(save_best_only=False,filepath=checkpoint_path),TensorBoard(log_dir='logs')]
history = model.fit(gen_tr,steps_per_epoch=200,
                                  epochs=epochs,
                                  validation_data=gen_val,
                                  validation_steps=1,
                                  callbacks = callbacks_list )

(paraphrased from Frightera)

For more details you can refer model.fit.

Upvotes: 0

Related Questions