fix the use of training a model on the colab #513
+8
−551
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
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.