-
I try train coco like dataset with 8 classes, here is my model config:
```yaml
model = dict(
type='SOLOv2',
pretrained='torchvision://resnet50',
backbone=dict(
type='ResNet…
-
To get a ResNet50 training baseline of loss value, I running the [resnet.py](https://github.com/wangkuiyi/gotorch/blob/develop/example/resnet/resnet.py) example and got the following logs:
``` text…
-
After training finished, I ran this command:
python tools/test_net.py --dataset my_dataset --cfg configs/baselines/e2e_faster_rcnn_R-50-C4_1x.yaml --load_ckpt Outputs/e2e_faster_rcnn_R-50-C4_1x/Feb…
-
您好，非常感谢你的工作，我是做视频分割的，我现在的backbone是resnet50，我想把经过预训练的resnet50换成预训练好的swin，可以这样做吗？具体要怎么做？
-
## Instructions To Reproduce the Issue:
I have trained a pretrained ImageNet resnet on a custom dataset with 12 classes.
For training I used following yaml file:
[training_yaml_file](https://git…
-
1. [Deep Learning in Diagnosis of Maxillary Sinusitis on Conventional Radiograph](https://openreview.net/references/pdf?id=SJUKtFssf) (@seunghyunni)
2. [Improvement diagnostic accuracy of sinusitis…
-
Hi, i train coco with config scratch_e2e_faster_rcnn_R_50_FPN_Xconv1fc_3x_gn.yaml, it trains nomally and loss decrease to 0.3-0.4, when i test the model,the results of AP and AR are all 0.I use the sy…
-
@AlexeyAB
I have 15 class dataset and there is no problem in images and text files as I trained detector perfectly from the same dataset.
For training resnet50 classifier, my files are as followin…
-
- https://arxiv.org/abs/2104.12533
- 2021
昨年は、Transformerモジュールを視覚問題に適用することが急速に発展しました。
一部の研究者は、Transformerベースのモデルがデータをフィットさせる能力に優れていることを実証しているが、特に学習データが限られている場合には、これらのモデルが過剰にフィットしてしまうことを示す証拠がまだ増えて…
e4exp updated
3 years ago
-
## ❓ Questions and Help
Hi, I was trying to train the mrcnn on a custom dataset but seems the loss is very high. Is this correct and also I dont see any signs of convergence.
![image](https://user-i…