Text-to-Speech
Piper
Italian
tts
checkpoint
finetuning
bacca87 commited on
Commit
ab22683
·
1 Parent(s): fb5a5bf

Reword README: drop v1 defect narrative

Browse files
Files changed (1) hide show
  1. README.md +8 -5
README.md CHANGED
@@ -28,11 +28,14 @@ Finetuning base checkpoint for Piper TTS (VITS), fine-tuned on Italian.
28
 
29
  ## Why this checkpoint
30
 
31
- This is the **fixed** version of the published Italian medium checkpoint: the
32
- previous release had a defect (sentence-initial words after `.?!` spoken with
33
- collapsed duration). This checkpoint is the fine-tuned base used to build the
34
- fixed `it_IT-serena-medium` voice, and can be used as a base for further
35
- finetuning of other Italian voices.
 
 
 
36
 
37
  ## Sample
38
 
 
28
 
29
  ## Why this checkpoint
30
 
31
+ Piper publishes **no official pretrained checkpoint for Italian** (the `rhasspy/piper-checkpoints`
32
+ catalog has no `it` entry; the only official Italian voices are `it_IT-paola-medium`, which was
33
+ finetuned from a U.S. English base, and `it_IT-riccardo-x_low`, trained from scratch at x_low
34
+ quality neither exposes a trainable checkpoint). This is the only public Italian **medium**
35
+ checkpoint that can be used as a finetuning base. It is the base used to build
36
+ `it_IT-serena-medium`, and it is fine-tuned on an extended dataset
37
+ (`serena-synthetic-it-28h`) with targeted sentence-initial phrases for improved
38
+ prosody at the start of sentences.
39
 
40
  ## Sample
41