search

GeorgeSeif / Semantic-Segmentation-Suite

Semantic Segmentation Suite in TensorFlow. Implement, train, and test new Semantic Segmentation models easily!
2.51k stars 878 forks source link

current_loss is always 0.0000 #174

Open RobinHan24 opened 5 years ago

RobinHan24 commented 5 years ago

[2019-01-10 16:11:28] Epoch = 3 Count = 980 Current_Loss = 0.0000 Time = 1.85 [2019-01-10 16:11:29] Epoch = 3 Count = 1000 Current_Loss = 0.0000 Time = 1.90 [2019-01-10 16:11:31] Epoch = 3 Count = 1020 Current_Loss = 0.0000 Time = 1.89 [2019-01-10 16:11:33] Epoch = 3 Count = 1040 Current_Loss = 0.0000 Time = 1.92 [2019-01-10 16:11:35] Epoch = 3 Count = 1060 Current_Loss = 0.0000 Time = 1.89 [2019-01-10 16:11:37] Epoch = 3 Count = 1080 Current_Loss = 0.0000 Time = 1.79 [2019-01-10 16:11:39] Epoch = 3 Count = 1100 Current_Loss = 0.0000 Time = 1.85 [2019-01-10 16:11:41] Epoch = 3 Count = 1120 Current_Loss = 0.0000 Time = 1.83 [2019-01-10 16:11:42] Epoch = 3 Count = 1140 Current_Loss = 0.0000 Time = 1.82 [2019-01-10 16:11:44] Epoch = 3 Count = 1160 Current_Loss = 0.0000 Time = 1.57 [2019-01-10 16:11:46] Epoch = 3 Count = 1180 Current_Loss = 0.0000 Time = 1.83 [2019-01-10 16:11:48] Epoch = 3 Count = 1200 Current_Loss = 0.0000 Time = 1.74

I modified as the readme, could you help me please, thank you .

ryohachiuma commented 5 years ago

You should definitely provide more information on the issue. such as... What kind of model and frontend did you use? Did you modify the code? Which dataset did you use?

RobinHan24 commented 5 years ago

You should definitely provide more information on the issue. such as... What kind of model and frontend did you use? Did you modify the code? Which dataset did you use?

model:DeepLabV3 frontend:MobileNetV2 I didn't modify the code, I just replace the Camvid data with my own data. Maybe there is error in my own dataset? 42 json_img 42 json_label

developpeur3d commented 5 years ago

Did you changed the size of the images ? because your images does not seems to be 512x512.

burhr2 commented 5 years ago

You can also try to check your color dict in your dataset because i had the same problem and i rechecked it was the wrong rgb numbers in my color dict (class_dict.csv) file

YyChen818 commented 5 years ago

I see the image above which is a binary problem, so do I. So, could you tell me that have you solved this problem and how. thank you