search

brade31919 / SRGAN-tensorflow

Tensorflow implementation of the SRGAN algorithm for single image super-resolution
MIT License
848 stars 284 forks source link

segmentation fault #43

Open menghuadeng opened 5 years ago

menghuadeng commented 5 years ago

Hi, I copied my dataset to ./data/train folder, then I ran the sh inference_SRGAN command, but the code not work, it reminded "segmentaion fault" and no other error. The detail is as follows, can you tell what the problem may be, thank you. @brade31919

/PycharmProjects/SRGAN-tensorflow$ sh inference_SRGAN.sh /home/menghua/anaconda3/lib/python3.6/site-packages/h5py/init.py:36: FutureWarning: Conversion of the second argument of issubdtype from float to np.floating is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type. from ._conv import register_converters as _register_converters [Configurations]: output_dir: <absl.flags._flag.Flag object at 0x7f0e9163d1d0> summary_dir: <absl.flags._flag.Flag object at 0x7f0e50c65c18> mode: <absl.flags._flag.Flag object at 0x7f0e50c65c88> checkpoint: <absl.flags._flag.Flag object at 0x7f0e50c74048> pre_trained_model: <absl.flags._flag.BooleanFlag object at 0x7f0e50c74128> pre_trained_model_type: <absl.flags._flag.Flag object at 0x7f0e50c869b0> is_training: <absl.flags._flag.BooleanFlag object at 0x7f0e50c86e48> vgg_ckpt: <absl.flags._flag.Flag object at 0x7f0e50c22630> task: <absl.flags._flag.Flag object at 0x7f0e50c226a0> batch_size: <absl.flags._flag.Flag object at 0x7f0e50c22748> input_dir_LR: <absl.flags._flag.Flag object at 0x7f0e50c227f0> input_dir_HR: <absl.flags._flag.Flag object at 0x7f0e50c22860> flip: <absl.flags._flag.BooleanFlag object at 0x7f0e50c22898> random_crop: <absl.flags._flag.BooleanFlag object at 0x7f0e50c22908> crop_size: <absl.flags._flag.Flag object at 0x7f0e50c229b0> name_queue_capacity: <absl.flags._flag.Flag object at 0x7f0e50c22a58> image_queue_capacity: <absl.flags._flag.Flag object at 0x7f0e50c22b00> queue_thread: <absl.flags._flag.Flag object at 0x7f0e50c22ba8> num_resblock: <absl.flags._flag.Flag object at 0x7f0e50c22c50> perceptual_mode: <absl.flags._flag.Flag object at 0x7f0e50c22cf8> EPS: <absl.flags._flag.Flag object at 0x7f0e50c22dd8> ratio: <absl.flags._flag.Flag object at 0x7f0e50c22e80> vgg_scaling: <absl.flags._flag.Flag object at 0x7f0e50c22f28> learning_rate: <absl.flags._flag.Flag object at 0x7f0e50c22f98> decay_step: <absl.flags._flag.Flag object at 0x7f0e50c2b048> decay_rate: <absl.flags._flag.Flag object at 0x7f0e50c2b0b8> stair: <absl.flags._flag.BooleanFlag object at 0x7f0e50c2b0f0> beta: <absl.flags._flag.Flag object at 0x7f0e50c2b1d0> max_epoch: <absl.flags._flag.Flag object at 0x7f0e50c2b278> max_iter: <absl.flags._flag.Flag object at 0x7f0e50c2b2e8> display_freq: <absl.flags._flag.Flag object at 0x7f0e50c2b358> summary_freq: <absl.flags._flag.Flag object at 0x7f0e50c2b400> save_freq: <absl.flags._flag.Flag object at 0x7f0e50c2b4a8> End of configuration Finish building the network 2018-09-04 16:08:21.109872: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX AVX2 FMA 2018-09-04 16:08:21.114524: I tensorflow/core/common_runtime/process_util.cc:69] Creating new thread pool with default inter op setting: 2. Tune using inter_op_parallelism_threads for best performance. Loading weights from the pre-trained model Evaluation starts!! Segmentation fault

lxk1990727 commented 5 years ago

same too

CognitiveClouds-Prasad commented 5 years ago

Same here. I am facing the same problem for few of the images only though.