Open Villiamos opened 3 years ago
the debug process will show scripts run till it meets this line: import tensorflow as tf tf.logging.set_verbosity(tf.logging.ERROR)
I tried many ways and it seems that the simplest way is to downgrade your TensorFlow version to '1.13.1' using pip install. before that I tried to change import tensorflow.compat.v1 as tf, but there are set_session issues ...
maybe try to debug your code in Pycharm? you could see where it ends and figure out the issue. also maybe double-check the tf version(tf.version)? I downgrade my Tensorflow to 1.x and it works fine.
Best,
On Tue, Mar 2, 2021 at 11:26 AM Villiamos notifications@github.com wrote:
@lydiaCS https://github.com/lydiaCS I downgrade TF but still the same issue [image: tensorflo] https://user-images.githubusercontent.com/44847279/109679600-02ea4200-7b7c-11eb-8805-8bc2843d67f0.png I put here my settings file maybe there is some issue Carla folder with exe file D:\CARLA_0.9.5
carla1.txt https://github.com/Sentdex/Carla-RL/files/6070130/carla1.txt [image: tensorflo] https://user-images.githubusercontent.com/44847279/109680085-712f0480-7b7c-11eb-882c-435eb6a9de3d.png
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Sentdex/Carla-RL/issues/14#issuecomment-789033374, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJU6GV6ZVUS6RFKMJRUWRF3TBUGVFANCNFSM4YGUJ5XQ .
@lydiaCS Hello, installing TensorFlow version 1.13.1 helped and the script starts, Carla turned on, so the main problem is fixed, thank you very much. I have one more quick question, after running the script, Carla starts, in cmd I have the messages "(project_env) D: \ CARLA_0.9.5 \ Carla-RL-master> py train.py Starting ... Starting Carla ... " and the question how will I notice that the script itself starts teaching?
I have installed tf 1.13.1 (windows) but there is no output from the script, neither the "Starting...". Any clues? @lydiaCS
I have installed tf 1.13.1 (windows) but there is no output from the script, neither the "Starting...". Any clues? @lydiaCS
Please install tensorflow-gpu==1.13.1 and keras==2.2.4.
I was able to get outputing by commenting out the lines that seeks to mute the error logging to the screen.
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3' stdin = sys.stdin sys.stdin = open(os.devnull, 'w') stderr = sys.stderr sys.stderr = open(os.devnull, 'w')
...
tf.logging.set_verbosity(tf.logging.ERROR)
.......
sys.stdin = stdin sys.stderr = stderr
These lines are in a number of files.
I have installed tf 1.13.1 (windows) but there is no output from the script, neither the "Starting...". Any clues? @lydiaCS
Please install tensorflow-gpu==1.13.1 and keras==2.2.4.
I have installed tensorflow-gpu and keras, but still nothing happened. There are only some warning:2024-03-05 23:42:28.484066: I tensorflow/core/util/port.cc:113] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable TF_ENABLE_ONEDNN_OPTS=0
.
WARNING:tensorflow:From D:\Anaconda3\envs\car\lib\site-packages\keras\src\losses.py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use tf.compat.v1.losses.sparse_softmax_cross_entropy instead.
When I run any of the scripts, nothing happens. There are no errors or warnings