search

microsoft / AutonomousDrivingCookbook

Scenarios, tutorials and demos for Autonomous Driving
MIT License
2.3k stars 563 forks source link

ValueError: Received an empty batch. Batches should at least contain one item. #102

Open Spatialghost opened 5 years ago

Spatialghost commented 5 years ago

Hi, i started train model with custom dataset and after 24 epochs i've got error: Traceback (most recent call last): File "train_model.py", line 94, in validation_data=eval_generator, validation_steps=num_eval_examples//batch_size, verbose=2) File "C:\Python36\lib\site-packages\keras\legacy\interfaces.py", line 87, in wrapper return func(*args, *kwargs) File "C:\Python36\lib\site-packages\keras\engine\training.py", line 2169, in fit_generator use_multiprocessing=use_multiprocessing) File "C:\Python36\lib\site-packages\keras\legacy\interfaces.py", line 87, in wrapper return func(args, **kwargs) File "C:\Python36\lib\site-packages\keras\engine\training.py", line 2305, in evaluate_generator raise ValueError('Received an empty batch. ' ValueError: Received an empty batch. Batches should at least contain one item.

batch_size=32

How to solve this problem?

mitchellspryn commented 5 years ago

Is your batch size a multiple of 32?

Spatialghost commented 5 years ago

Is your batch size a multiple of 32?

I use code without any changes.

Spatialghost commented 5 years ago

Solved! Just have to restart. Different problem now, early stopping after 90 - 100 epochs.

martsalz commented 4 years ago

What did you have to restart? I have the problem on several machines.

kutilanak commented 2 years ago

Solved! Just have to restart. Different problem now, early stopping after 90 - 100 epochs.

AFAIK it doesn't have to complete the training to 500 epoch. If the val_loss did not improve after several epoch, it will automatically stop (early stopping)

u can use the latest model it saved.