lindadamama

Results 12 issues of lindadamama

missing futures tf.keras.callbacks

![1](https://user-images.githubusercontent.com/48863486/215397250-01b4352e-0e67-4d15-afd7-2f40f7fb1edc.PNG) 左边为训练集,右边为测试结果,请问这种情况该怎么处理?

Do I need to read all data into memory and then add new data and save it? Can you provide an example? Can I insert data directly as I operate...

![1](https://user-images.githubusercontent.com/48863486/215396944-5b8b1ee7-c191-4dbc-88aa-0f0a84210774.PNG) 左边为训练集,右边为测试结果,请问这种情况该怎么处理?

有没有不包含mmseg的版本?看起来比较费劲,没用的信息太多了

调用torch.jit.script(self.net).save("model.pt"),程序一直报错TypeError: cannot create weak reference to 'numpy.ufunc' object 调用toch.jit.trace(self.net,torch.rand(1,3,256,320)).save("model.pt")可以正常导出模型,但是测试结果和原来的模型(self.net)不一致 ![捕获](https://user-images.githubusercontent.com/48863486/220063748-de9235a4-f8b6-4e06-8d3a-0f188bd3ac03.PNG) 左边是直接预测的,右边是用toch.jit.trace(self.net,torch.rand(1,3,256,320)).save("model.pt")导出的模型预测的

### 📚 The doc issue ```pthon pip install mmcv-full==1.3.14 ``` I encountered this problem when installing mmcv. I don't know how to solve it ### Suggest a potential alternative/fix _No...

![捕获](https://user-images.githubusercontent.com/48863486/227129642-f3ff904f-7912-423b-bc01-2b971c33ea63.PNG)