-
I trained the model with 4 GPUs (GTX 2080) with batchsize 4, and the PSPNet achieves the mIOU of 76.42%. But I found the visualization results are not correct. Here is an example.
 **layer_factory.hpp:81 …
-
hi, i have a question about the ignore index 255 when i trainning the dataset of citysacpes. while training, the background will be set to label 255, and in the loss function it will not be calculated…
-
@NielsRogge I'm trying to fine-tune Oneformer for instance segmentation and HF throws an exception. Here's a simple reproduction:
[inputs.zip](https://github.com/NielsRogge/Transformers-Tutorials/…
-
Hi there,
Thanks for sharing your codes. I wonder if the experience setting in your current train.py can achieve mIoU=0.6768 on Cityscapes.
I trained the model, but I can get good result.
Can you …
gnsci updated
2 years ago
-
hey! everything performs perfectly when I using singe GPU to implement the testing, However, when I try to make good use of my GPU resource, something dump.
I use 3 GPU and set the step to 483，(144…
-
- branch
use master or dev-1.x could cause the problem.
## command
```
python /opt/mmsegmentation/tools/torchserve/mmseg2torchserve.py /opt/mmsegmentation/configs/pspnet/pspnet_r50-d8_512x1024…
-
我使用的是cityscapes来训练的，配置好了您的预训练权重，运行main.py时不管batchsize设置多少都会报错
RuntimeError: CUDA out of memory. Tried to allocate 86.00 MiB (GPU 0; 4.00 GiB total capacity; 2.09 GiB already allocated; 61.09 MiB free…
-
I used the following command for visualization, but got the following error.
"""
python demo/image_demo.py data/fp_grayscale/fp/images/validation/389808-2model.jpg local_configs/segformer/B4/segform…