fix the use of training a model on the colab

[JinFoolish]

When training the model on colab, the program reported an error when saving dataset_statistic. I found in the Issue that others also encountered the same problem.. I checked the ddsp library and found that framerate was not passed as a parameter into the calculation process. When training with the default configuration in the notebook, the default value of frame_rate is 250, but for the compute_dataset_statistics function, the default value of frame_rate is 50, so I add it as a parameter in the save_dataset_statistics function and pass it to compute_dataset_statistics.

Directly pip install ddsp==3.6.0 in colab, the training will not be able to use colab's GPU, I modified the pip part to use colab's default tensorflow. I used the modified version and trained several models on colab, and the GPU can be used normally.

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.