-
Hi,
I try to reproduce your code (gta5 -> cityscapes) and I found that it has a very slow training speed, which is nearly 4x slower than the DACS. I wonder why... And also, can I ask you the total tr…
-
2021-12-23 11:41:25,981 Segmentron INFO: train_fine data num 2975
2021-12-23 11:41:26,012 Segmentron INFO: train_fine data num 2975
2021-12-23 11:41:26,017 Segmentron INFO: val_fine data num 500
20…
-
我训练初始损失3+最后损失值降低到了0.5,但是利用提供的可视化代码显示效果并不理想.
-
Thanks for sharing your code~
I cannot reproduced the results of GPA (Two-stage alignment), are there any important details that I missed? (e.g., in Foggy City, you used the images of beta = 0.01, ra…
-
Thank you very much for your code. But I had a problem loading the weight: tf.train. latest_checkpoint (checkpoint_dir) return None Type. It seems to lack the "checkpoint" file in the compressed packa…
-
I employ the DataParallelCriterion like below,
```
model = DataParallelModel(deeplab)
model.train()
model.float()
model.cuda()
# .......
criterion = DataParal…
-
作者您好,我看到有人提出相似的问题,但是没有找到答案。
我采用batchsize 8 iter60000 初始0.01学习率的策略,但是最终只能在cityscapes val上到大约67%,68%的miou(STDC1-50)。
我使用的训练脚本就是github README.MD中的训练脚本。
请问我是不是哪里使用错误了么?为什么原封不动的使用您的代码,却无法达到(或者接近)您…
-
Hi,
Thank you for your research and I really like your achievement and sharing.
I evaluated the model "cityscapes_baseline_strongaug_ema_foggy_val_model_best_591_ema2model.pth" with this comman…
-
Dear Mark,
According to the paper, it seems that the backbone of the model for Potsdam-3 dataset is vit-base, but there are no specific values for hyperparameters.
In addition, train_config.yml ha…
-
**Task**: Currently [tfds.show_examples()](https://www.tensorflow.org/datasets/overview#visualization) only works for supervised images datasets. It would be good to extend the heuristic to more datas…