Update README.md

test
Michael Nguyen 2018-09-14 14:27:52 -05:00 committed by GitHub
parent 2ee34ceaa8
commit c69fad46c6
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 0 deletions

View File

@ -242,3 +242,6 @@ Below is an example of a bad distribution for number of samples. This distributi
![phonemedist](example_visuals/phonemedist.png)
This only output if you use the `--cmu_dict_path` parameter. The X axis are the unique phonemes and the Y axis shows how many times that phoneme shows up in your dataset. We are still experimenting with how the distribution should look, but the theory is having a balanced distribution of phonemes will increase quality in pronounciation.
### Tips
If your data looks bad, you can try resampling methods to change the shape of your data. For example, to fix the bad number of samples chart. You can try to delete samples and copy samples to make the chart look more normally distributed. For standard deviation, you can remove data that causes your standard deviation of have a large spread.