Winnie202

Results 25 comments of Winnie202

> Hi~ Have you followed the steps in https://mmcv.readthedocs.io/en/latest/deployment/tensorrt_plugin.html?highlight=trt? It seems that you've not TensorRT correctly I installed tensorRT and mmcv-full first, and then followed the command: git clone https://github.com/open-mmlab/mmcv.git...

> 您好,看来您使用的是 tensorRT 8.10。如何尝试 TensorRT-7.2.1.6 并按照文档中的步骤操作。 ![图片](https://user-images.githubusercontent.com/57566630/179195022-62160d1e-e955-481e-892f-9b9424c6c1f8.png) Yes, I installed tensorRT8.2.1.8, but I changed the version number in the command during the actual operation. The reason is that the version...

How should the value of lr_config and num_input_frames be set? Is the value of num_input_frames the bigger the better? ![2022-07-12 11-04-48屏幕截图](https://user-images.githubusercontent.com/86768013/178403213-ff809d63-2e6e-44ff-832a-e59f462c94de.png)

I got this kind of result after changing it according to similar code ![2022-07-06 16-07-16屏幕截图](https://user-images.githubusercontent.com/86768013/177501907-1c39b149-cb06-4ba4-813c-e8abfb8d5d77.png) ![2022-07-06 16-08-47屏幕截图](https://user-images.githubusercontent.com/86768013/177502399-8f78f645-33ca-43d9-8f94-f78161bbe96e.png) I would like to know why it is not saved as separate sheets

>my obtained lq_path is a list,like this:['data/test/test_lq/001/0000.jpg', 'data/test/test_lq/001/0001.jpg', 'data/test/test_lq/001/0002.jpg', 'data/test/test_lq/001/0003.jpg', 'data/test/test_lq/001/0004.jpg']

> 是tensor2img函数和real_esrgan.py中的forward_test函数都需要修改对吗?另外我还想请问一下,训练时如下警告会不会对精度有影响 ![2022-07-06 17-32-26屏幕截图](https://user-images.githubusercontent.com/86768013/177520126-c4709e56-df2f-4ba3-b167-401a770685aa.png)

> 感谢您的及时回复,给了我很大帮助

> 你好,这是正常的。我们的一些测试结果可能不太令人满意,尤其是在人脸上(需要在未来的工作中改进)。 thank you for your reply。I use the weights you provided, follow the “python test.py --resume './pretrained_weights/masa.pth' --testset TestSet_multi --name masa_TestSet_multi ”command to test, the input size is the...

> Yes, I modified the category in the configuration file to be consistent with my data, this is my json file and classes ![2022-08-10 10-00-59屏幕截图](https://user-images.githubusercontent.com/86768013/183794120-b6d2cc50-4ea1-4d23-9993-8b06c8fee4cb.png) ![2022-08-10 10-01-16屏幕截图](https://user-images.githubusercontent.com/86768013/183794176-3a6ba863-d76c-4b13-913e-07a3ad734967.png) ![2022-08-10 10-01-37屏幕截图](https://user-images.githubusercontent.com/86768013/183794211-826c56c9-7d2c-475e-934e-140715566e10.png) ![2022-08-10...

> 看来你的训练过程被打破了。您可以从头开始训练并提供包含验证指标的整个 vdl 吗?我在训练配置中看不到任何错误。一些参数与您的数据相关。 Because cuda memory is not enough, it is interrupted. Is it normal that the value of train/lr does not change?