Update about.md
Browse files
about.md
CHANGED
@@ -22,12 +22,15 @@ Here you'll be able to find all the information regarding our models π΅ Matxa
|
|
22 |
|
23 |
The significance of open-source text-to-speech (TTS) technologies for minority languages cannot be overstated. These technologies democratize access to TTS solutions by providing a framework for communities to develop and adapt models according to their linguistic needs. This is why we have developed different open-source TTS solutions in Catalan, using an ensemble of technologies.
|
24 |
|
25 |
-
|
26 |
|
27 |
* Balear
|
|
|
28 |
* North-Occidental
|
29 |
* Valencian
|
30 |
|
|
|
|
|
31 |
|
32 |
## Intended Uses and Limitations
|
33 |
|
@@ -228,13 +231,13 @@ The original Matcha-TTS model excels in English, but to bring its capabilities t
|
|
228 |
* [Our version of the Festcat dataset.](https://huggingface.co/datasets/projecte-aina/festcat_trimmed_denoised)
|
229 |
|
230 |
Then we further fine-tuned the single accent Catalan Matxa-based model with the soon to be published LaFrescat dataset that has 8.5 hours of recordings for four dialectal variants:
|
231 |
-
|
232 |
-
* Central
|
233 |
-
|
234 |
* Valencian
|
235 |
|
236 |
* Occidental
|
237 |
-
|
|
|
|
|
238 |
* Balear
|
239 |
|
240 |
With a male and a female speaker for each dialect.
|
|
|
22 |
|
23 |
The significance of open-source text-to-speech (TTS) technologies for minority languages cannot be overstated. These technologies democratize access to TTS solutions by providing a framework for communities to develop and adapt models according to their linguistic needs. This is why we have developed different open-source TTS solutions in Catalan, using an ensemble of technologies.
|
24 |
|
25 |
+
Here we present π΅ Matxa, the first multispeaker, multidialectal neural TTS model. It comes together with the vocoder model π₯ alVoCat, to generate high quality and expressive speech efficiently in four dialects:
|
26 |
|
27 |
* Balear
|
28 |
+
* Central
|
29 |
* North-Occidental
|
30 |
* Valencian
|
31 |
|
32 |
+
Both models are trained with open data; π΅ Matxa models are free (as in freedom) to use for non-comercial purposes, but for commercial purposes it needs licensing from the voice artist. For details please consult the [License](#additional-information) section and the [model page](https://huggingface.co/BSC-LT/matcha-tts-cat-multiaccent/).
|
33 |
+
|
34 |
|
35 |
## Intended Uses and Limitations
|
36 |
|
|
|
231 |
* [Our version of the Festcat dataset.](https://huggingface.co/datasets/projecte-aina/festcat_trimmed_denoised)
|
232 |
|
233 |
Then we further fine-tuned the single accent Catalan Matxa-based model with the soon to be published LaFrescat dataset that has 8.5 hours of recordings for four dialectal variants:
|
234 |
+
|
|
|
|
|
235 |
* Valencian
|
236 |
|
237 |
* Occidental
|
238 |
+
|
239 |
+
* Central
|
240 |
+
|
241 |
* Balear
|
242 |
|
243 |
With a male and a female speaker for each dialect.
|