Adding information from new training.

README.md

@@ -26,6 +26,8 @@ Trained on +-130hrs
 and upgraded to new dataset +-200hrs 30k samples in 2 days with mostly 20s on A100
 Added 3 days on 3500 speakers from, 60k samples audios joined with from 20 to 25s of
 https://commonvoice.mozilla.org/pt
+Added 2 days on 90k samples audios, colab A100.
+Added 5 days on 90k samples audios, with 2 rtx3090 48gb parallel.
 
 Use lower case, and for numbers use num2words. Sample bellow.
 
@@ -90,13 +92,31 @@ time data : 183:27:23
 min sec : 1.02
 max sec : 30.0
 vocab : 2545
+mostly 5s
 
-around4 days ( 800k steps )
+around 4 days ( 800k steps )
 samples : 128908 
 time data : 196:24:47
 min sec : 1.0
 max sec : 25.0
 vocab : 2545
+mostly 10s
+
+around 7 days (1000k steps)
+samples : 90947
+time data : 447:51:31
+min sec : 3.0
+max sec : 30.0
+vocab : 2545
+
+Total audios: 90,947
+Minimum duration: 1.02 seconds
+Maximum duration: 30.0 seconds
+Mean (average) duration: 17.73 seconds
+Median (P50): 19.60 seconds
+P90 (90th percentile): 23.90 seconds
+P95 (95th percentile): 25.96 seconds
+
 
 
 License