Baybars commited on
Commit
bfe25c0
β€’
1 Parent(s): 4c4db00

Update about.md

Browse files
Files changed (1) hide show
  1. about.md +8 -5
about.md CHANGED
@@ -22,12 +22,15 @@ Here you'll be able to find all the information regarding our models 🍡 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
- Firstly, we created a [TTS model for central Catalan](https://huggingface.co/BSC-LT/matcha-tts-cat-multispeaker) by fine-tuning the Matcha-TTS English model. Matcha-TTS is a state-of-the-art model that employs deep learning, a form of AI, to train models that replicate human speech patterns, allowing it to generate lifelike synthetic voices from written text. After that, we fine-tuned this Catalan central model for four Catalan dialects, central plus three more:
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 🍡 Matxa, the first multispeaker, multidialectal neural TTS model. It comes together with the vocoder model πŸ₯‘ 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; 🍡 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.