search

thanosvlo / MARL-for-Anatomical-Landmark-Detection

Tensorflow Implementation of Multiple Landmark Detection using Multi-Agent Reinforcement Learning Paper
Apache License 2.0
57 stars 13 forks source link

The Training does not start #6

Closed savanigokhale closed 3 years ago

savanigokhale commented 3 years ago

When I run DQN.py, a few errors were thrown. After I fixed those errors the script fetched the data correctly but the training does not start.

Command: python DQN.py --task train --gpu 0 --files './data/filenames/image_files.txt' './data/filenames/landmark_files.txt'

On terminal:

WARNING:tensorflow: The TensorFlow contrib module will not be included in TensorFlow 2.0. For more information, please see:

[0205 09:33:32 @logger.py:90] Argv: DQN.py --task train --gpu 0 --files ./data/filenames/image_files.txt ./data/filenames/landmark_files.txt

Not sure where I made the mistake. I am using the same versions of the packages required.

thanosvlo commented 3 years ago

The terminal output you posted just is a deprecation notice, make sure you are using TF 1.14 or earlier , the code has not been changed to work with TF 2.0, was there anything else in the console output to indicate the error ?

savanigokhale commented 3 years ago

I am using TF 1.14.0. There's nothing in the console output which indicates error.

I printed a few statements to check on what step it is freezing.

Console output: WARNING:tensorflow: The TensorFlow contrib module will not be included in TensorFlow 2.0. For more information, please see:

in class model in class medical player

Landmark files ['./data/landmarks/ADNI_002_S_0816_MR_MPRGradWarpB1_CorrectionN3Scaled_Br_20070217005829488_S18402_I40731.txt', './data/landmarks/ADNI_003_S_1059_MR_MPR-RGradWarpB1_CorrectionN3Scaled_Br_20070501173720913_S22300_I52816.txt', './data/landmarks/ADNI_005_S_0448_MR_MPRGradWarpB1_CorrectionN3Scaled_Br_20061212165540546_S14032_I32875.txt', './data/landmarks/ADNI_006_S_0681_MR_MPRGradWarpB1_CorrectionN3Scaled_Br_20080224150248153_S18451_I92305.txt', './data/landmarks/ADNI_007_S_0293_MR_MPRGradWarpB1_CorrectionN3Scaled_Br_20070118022549519_S12193_I36549.txt', './data/landmarks/ADNI_009_S_1354_MR_MPRGradWarpN3Scaled_Br_20070717183309825_S27634_I60443.txt', './data/landmarks/ADNI_010_S_0786_MR_MPR__N3Scaled_Br_20070807153633938_S19638_I65597.txt', './data/landmarks/ADNI_011_S_0010_MR_MPRGradWarpB1_CorrectionN3Scaled_Br_20061208114538147_S8800_I32270.txt', './data/landmarks/ADNI_133_S_0638_MR_MPRGradWarpB1_CorrectionN3Scaled_Br_20070813172348807_S16608_I67531.txt', './data/landmarks/ADNI_941_S_1363_MR_MPRGradWarp__B1_CorrectionN3Scaled_Br_20070801202916120_S28008_I63897.txt']

length of image files 10

length of landmark files 10 in restart episode in new random game

length of image files in num_files 10 [57, 57] [138, 117] [78, 52]

length of image files in num_files 10 num actions and files 6 10 [0205 15:17:09 @logger.py:90] Argv: DQN.py --task train --gpu 0 --files ./data/filenames/image_files.txt ./data/filenames/landmark_files.txt

in func get config in get_player in class medical player

Landmark files []

length of image files 0

length of landmark files 0 in restart episode in new random game

length of image files in num_files 0