-
Step: 25561. Epoch: 78/5000. Iteration: 212/325. Cls loss: 0.00248. Reg loss: 0.00373. Total loss: 0.00620: 65%|▋| 21
请教下我用自己的数据集,更改num_classes参数后(变小),训练收敛,测试结果乱七八糟。
但是用预训练模型跑img.png结果正常,请问作者能给个建…
-
Thanks for your error report and we appreciate it a lot.
**Checklist**
1. I have searched related issues but cannot get the expected help.
2. I have read the [FAQ documentation](https://mmdetec…
-
Hi, I have changed num_classes = 1 in all of the needed configuration files and my tensors also has the correct shape but still, I am receiving the error. Can you please point me in the right directio…
-
If you do not know the root cause of the problem / bug, and wish someone to help you, please
post according to this template:
## Instructions To Reproduce the Issue:
1. what changes you made (`…
-
Hello,
I am experiencing some odd behavior when trying to train on a custom dataset in COCO format. So far I have tried training faster-rcnn and retinanet models but the bbox loss is always zero. T…
-
After updating to v2.5.0, training faster_rcnn model on customized dataset (COCO-like format) seem to have broken. The validation accuracy does stays close to 0 starting at the second epoch and stays …
-
I tried training the mask rcnn with both detectron2 and mmdetection and the training speed gap is huge
mmdetection offers a lot of preimplemented models that I want to run my tests with but I'm st…
-
## ❓ Questions and Help
So let's say that I have trained a model on keypoints or instance segmentation with datasets that are on coco-format, and used that trained model to get some outputs (Infere…
-
i try to train htc
`python /mmdetection/tools/train.py 'htc_hrnetv2p_w18_20e_coco.py'`
and this is the config file
```
dataset_type = 'CocoDataset'
data_root = 'data/coco/'
img_norm_cfg = …
-
I meet a ERROR when I trained my custom dataset whose format is COCO.
I browsed some **issues** but them couldn't help me solve this problem.
# My config file
```python
_base_ = './faster_rcnn/f…