search

seth814 / Semantic-Shapes

A semantic segmentation pipeline for custom image annotation
MIT License
128 stars 55 forks source link

Cant Train on a larger image size. #5

Closed Dan201210 closed 4 years ago

Dan201210 commented 4 years ago

im try to train on a larger image with shape (1088,1920,3). when i run the train.py i get an error.

File "C:\Users\user\anaconda3\envs\shapes\lib\site-packages\tensorflow_core\python\keras\engine\training.py", line 908, in predict use_multiprocessing=use_multiprocessing) File "C:\Users\user\anaconda3\envs\shapes\lib\site-packages\tensorflow_core\python\keras\engine\training_arrays.py", line 716, in predict x, check_steps=True, steps_name='steps', steps=steps) File "C:\Users\user\anaconda3\envs\shapes\lib\site-packages\tensorflow_core\python\keras\engine\training.py", line 2471, in _standardize_user_data exception_prefix='input') File "C:\Users\user\anaconda3\envs\shapes\lib\site-packages\tensorflow_core\python\keras\engine\training_utils.py", line 572, in standardize_input_data str(data_shape)) ValueError: Error when checking input: expected input_1 to have shape (1088, 1920, 3) but got array with shape (256, 256, 3)

im not sure if this is a tensorflow issue or if your code is casting to (256,256,3) somewhere that i cant find. I have no trouble training on (256,256,3)

Dan201210 commented 4 years ago

i figured it out. i left 256,256 images in the logs folder and tensorflow_callbacks was loading them and throwing the error.