-
你好，我用的是DeepKE-cnSchema开箱即用版
python 3.9
torch 1.13.1 cpu版本
, loss becomes nan after several iterations. i have tried to de…
-
Three columns:
1. hash value
2. No-Intro name
3. SMDB name
(a lot of false-positive due to `top X lists` and chronological folders)
```sh
NOINTRO="Nintendo - Super Nintendo Entertainment Syste…
-
已将bert-base-chinese下载到/DeepKE/example/ner/standard/bert-base-chinese，执行run_bert.py报错如下：ValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': '/root//PythonProject/DeepKE/e…
-
您好，我通过以下代码导入模型后想要微调CLIP模型，我先冻结了CLIP的参数，添加了全连接层在我的任务上训练, 之后再解冻CLIP部分参数，但是CLIP模型输出为Nan，请问我应该如何微调CLIP模型呢？
```
model, preprocess = load_from_name("ViT-B-16", device=device, download_root='./')
```
-
### 🚀 The feature, motivation and pitch
Unable to export to ONNX a model that uses the operation 'aten::lgamma'.
### Alternatives
I have tried adding the operation to symbolic_opsetxx.py as p…
-
## Keyword: detection
### Video Anomaly Detection by Estimating Likelihood of Representations
- **Authors:** Yuqi Ouyang, Victor Sanchez
- **Subjects:** Computer Vision and Pattern Recognition (cs.C…
-
hello，
I encountered an error while reproducing the code. I used the first 100 entries of ag_news as test samples, and then ran it on my laptop. There were no issues during initialization, but an e…
-
Share your Chinese synthesis results or mandrain model training questions.
-
如题，当我有一份自定义的图片以及文档数据集时，我应该如何使用该数据集对PreFLMR的模型进行微调以达到我想要的通过图片来检索相关文档的目的？是否需要做图片以及文档的向量对齐的训练？期望能够解答我的疑惑，不甚感激。