-
The break occurs when I train the rtdetr-l model with 300 epoches going to 90, but when I use resume, the epoches start at 91 but the map,R and P value become 0 and stay at 0 。the training code as fol…
-
Hi Sir,
As a first part of my project, I have to transfer the weights of the three first 3D convolutional layers of a pretrained model (conv3d_deepnetA_sport1m_iter_1900000 for example) to a new mode…
-
Hello Sir ,
Could you please let me know what's best_quant mean ?
torch.save(ckpt, best)
torch.save(model, best_quant, _use_new_zipfile_serialization=False)
-
Things to consider/steps to be taken:
## Switch to Iceland
- [x] create a maintenance branch of OSS > tag https://github.com/52North/OpenSensorSearch/releases/tag/pre-iceland
- [x] add new 52N parent …
nuest updated
8 years ago
-
Hello sir, when I use java-small preprocess dataset to execute the code2vec model, the following problem occurs.I changed the suffix of the dataset c2s to c2v
![1661616103(1)](https://user-images.git…
-
Sir,do you have any idea how to implement this?I have a pseudocode but Im not able to use it on this model.Can you help?
ghost updated
3 years ago
-
Dear Dev Team,
I am using seq_gan model for generating text sequence. I am trying to use list of valid "subdomains" as input and generate new "text" as output. I am making changes in run/run_seqga…
-
Hi Sir,
Thanks for publishing your work, it is very interesting to review.
One question I have is about the time when testing network configuration.
You are using D2-Net models but this model …
-
**API Platform version(s) affected**: 2.6.2 to 3.3.6
**Description**
PR #4019 introduces an anonymous service argument in `openapi.xml` for the service `api_platform.openapi.normalizer`. Whilst …
-
Hello, sir,I have tried your code, but currently I have found that the model I trained fused the image results in blurry structural details. I would like to know if you have also experienced this prob…