cvlab-epfl/tf-lift

Dockerfile

Opened this issue · 5 comments

breakingflower commented

Hi

I've made a docker file that contains all of your code and dependencies.

docker pull fremmen/tf-lift
coolbeam commented

Hi

I've made a docker file that contains all of your code and dependencies.

docker pull fremmen/tf-lift

hello，thanks for your docker, i've pull your docker file and run the code. I have successfully run the first phase: subtask=kp. However, when i'm going to run the second phase: subtask=ori, it turns out:

[joint] Checking if old pre-trained weights exists in
[joint] No pretrained weights from the old framework
[joint] Checking if previous Tensorflow run exists in
[joint] No previous Tensorflow result
[joint] Checking if previous Tensorflow run exists in logs/test/joint
[joint] No previous Tensorflow result
[ori] Checking if old pre-trained weights exists in
[ori] No pretrained weights from the old framework
[ori] Checking if previous Tensorflow run exists in
[ori] No previous Tensorflow result
[ori] Checking if previous Tensorflow run exists in logs/test/ori
[ori] No previous Tensorflow result
Traceback (most recent call last):
File "main.py", line 98, in
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/platform/app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "main.py", line 70, in main
task.run()
File "/tf-lift/tester.py", line 110, in run
raise RuntimeError("Could not load network weights!")
RuntimeError: Could not load network weights!

I think maybe something run in the pretrained model ori

coolbeam commented

Hi
I've made a docker file that contains all of your code and dependencies.

docker pull fremmen/tf-lift

hello，thanks for your docker, i've pull your docker file and run the code. I have successfully run the first phase: subtask=kp. However, when i'm going to run the second phase: subtask=ori, it turns out:

[joint] Checking if old pre-trained weights exists in
[joint] No pretrained weights from the old framework
[joint] Checking if previous Tensorflow run exists in
[joint] No previous Tensorflow result
[joint] Checking if previous Tensorflow run exists in logs/test/joint
[joint] No previous Tensorflow result
[ori] Checking if old pre-trained weights exists in
[ori] No pretrained weights from the old framework
[ori] Checking if previous Tensorflow run exists in
[ori] No previous Tensorflow result
[ori] Checking if previous Tensorflow run exists in logs/test/ori
[ori] No previous Tensorflow result
Traceback (most recent call last):
File "main.py", line 98, in
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
File "/usr/local/lib/python3.5/dist-packages/tensorflow/python/platform/app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "main.py", line 70, in main
task.run()
File "/tf-lift/tester.py", line 110, in run
raise RuntimeError("Could not load network weights!")
RuntimeError: Could not load network weights!

I think maybe something run in the pretrained model ori

hello，thanks for your docker, i've pull your docker file and run the code.
Thanks a lot!
The code is all right, I found that if you want to run the pretrained model, you should use: '--logdir=release-aug', not the default '--logdir=logs/test'

kmyi commented

Closing, since not a bug

kmyi commented

Well, actually, leaving it open since the docker is useful! (thought this was a bug report. Sorry for spam!)

breakingflower commented

@coolbeam running the code inside the docker is the same as in the original implementation.