search

dengdan / seglink

An Implementation of the seglink alogrithm in paper Detecting Oriented Text in Natural Images by Linking Segments
GNU General Public License v3.0
494 stars 177 forks source link

import error #26

Open shengzhang90 opened 6 years ago

shengzhang90 commented 6 years ago

Dear Deng： I'm sorry I am a newer for TensorFlow framework and get an error of following：

++ set -e ++ export CUDA_VISIBLE_DEVICES=0 ++ CUDA_VISIBLE_DEVICES=0 ++ CHECKPOINT_PATH=/home/jin/zs/seglink/seglink-512/ ++ DATASET_DIR=/home/jin/data/ch4/ch4_test_imgs ++ python test_seglink.py --checkpoint_path=/home/jin/zs/seglink/seglink-512/ --gpu_memory_fraction=-1 --seg_conf_threshold=0.8 --link_conf_threshold=0.5 --dataset_dir=/home/jin/data/ch4/ch4_test_imgs I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcublas.so locally I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcudnn.so locally I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcufft.so locally I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcuda.so.1 locally I tensorflow/stream_executor/dso_loader.cc:111] successfully opened CUDA library libcurand.so locally Traceback (most recent call last): File "test_seglink.py", line 9, in from tensorflow.contrib.training.python.training import evaluation ImportError: cannot import name evaluation

I will appreciate it very much！ Thanks！

dengdan commented 6 years ago

What your tensorflow version? 1.1.0 should be used, higher versions should be OK too, but I am not sure.