Jinghua Zhao
Jinghua Zhao
hi, i met same issue, i use vscode jupyter to show the fig, only show the axis, but i use `ani.save("animation.mp4", fps=20, writer="ffmpeg")` in util.py: interactive_diarization function, the file i...
sorry, the fps might not be 20, i think 16 might be great?
i think i solve this problem
> @anonymous-atom. The weights currently on Huggingface are for the older codebase. The new checkpoint will be uploaded soon with the updated version—stay tuned! Is there a plan for when...
> @ChocoWu 运行 PyTorch 训练脚本 train.py 时遇到了一个断言错误(AssertionError)。错误发生在 common/utils.py 文件的 l2_loss 函数中,具体是因为两个张量 u 和 v 的形状不匹配。 AssertionError: (torch.Size([1, 77, 768]), torch.Size([1, 77, 1024])) 你能成功运行demo_app.py吗,输入的音频或图像信息可以正常分析吗
> > > > 运行 PyTorch 训练脚本 train.py 时遇到了一个断言错误(AssertionError)。错误发生在 common/utils.py 文件的 l2_loss 函数中,具体是因为两个张量 u 和 v 的形状不匹配。AssertionError: (火炬.大小([1, 77, 768]),火把。大小 ([1, 77, 1024])) > > > > > > >...
> > > 运行 PyTorch 训练脚本 train.py 时遇到了一个断言错误(AssertionError)。错误发生在 common/utils.py 文件的 l2_loss 函数中,具体是因为两个张量 u 和 v 的形状不匹配。AssertionError: (火炬.大小([1, 77, 768]),火把。大小 ([1, 77, 1024])) > > > > > > 你能成功运行demo_app.py吗,输入的音频或图像信息可以正常分析吗 >...
> We only do the training and tuning on the audio captioning task. If you want to perform emotion, human vocal sound, music genre, or music instrument recognition. I guess...
> Could you check the training loss over time to see if the performance is actually improving? > > If the training loss isn't improving, it's possible that your model...
i found this link: [issues](https://github.com/hhj1897/face_detection/issues/4) to download the model can solve this!