3.4 Changelog#
3.4.0#
Compatibility with Kalpy 0.9.0
Introduced new command for Comparing and evaluating alignments (mfa compare_alignments)
Introduced new MFA model format for distribution on HuggingFace
Introduced new commands that use the new MFA model format (i.e. Align with a model hosted on Hugging Face (mfa align_hf)) which will replace the default commands (i.e. Align with an acoustic model (mfa align)) in MFA 4.0. The default commands will still be available using a legacy command and will be removed in MFA 5.0.
Deprecated Train a word tokenizer (mfa train_tokenizer) and Tokenize utterances (mfa tokenize) for removal in MFA 4.0
Added calculation for Intensity deviation and Signal-to-Noise Ratio (SNR) to be included as diagnostic metrics for alignment
Updated calculation for phone duration deviation to use log-duration to better model the duration distribution
Added column for Max running short intervals as a measure of alignment quality