-
### System information
- **What is the top-level directory of the model you are using**: research/object_detection
- **Have I written custom code**: no
- **OS Platform and Distribution (e.g., Linux…
-
Hi! I'm trying to train deformable detr with swin-tiny backbone, but I get the error :
- **RuntimeError: CUDA out of memory. Tried to allocate 84.00 MiB (GPU 0; 14.76 GiB total capacity; 12.12 GiB…
-
Hello~
the error is
ValueError: bbox_params must be specified for bbox transformations
the version of mmdetection is 2.23.0
and l follow the #7143 ,use `pip install -U albumentations --no-bin…
-
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…
-
My image size is 416*416, I changed the config.file like that but still I got this error. How can I solve this problem?
My custom config file:
```
# The new config inherits a base config to h…
-
**Describe the bug**
RuntimeError: CUDA error: device-side assert triggered; training HTC with the custom dataset.
It looks like related to the shape of **gt_sematic_seg** {torch.Size([2, 1, 100, 16…
-
when i try to train scnet_x101_64x4d_fpn_20e_coco , it happens error.
my config file is that.
```
# 这个新的配置文件继承自一个原始配置文件，只需要突出必要的修改部分即可
_base_ = '../configs/scnet/scnet_x101_64x4d_fpn_20e_coco…
-
------------------------
### System information
- **What is the top-level directory of the model you are using**:
`research/object_detection`
- **Have I written custom code (as opposed to using …
ghost updated
2 years ago
-
I am trying to train a YOLOX model with default configuration settings.I have only changed the number of classes according to my dataset and dataset paths inside the cfg.You can see my configuration b…
-
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…