Update README.md
Browse files
README.md
CHANGED
@@ -7,30 +7,42 @@ Test RVC2 models on the DDLC character Monika, via various hyperparams and datas
|
|
7 |
# monika-test-0 (~07/2023)
|
8 |
* Trained on augmented dataset of ~10 10 second clips
|
9 |
* Trained for ~100 epochs
|
10 |
-
*
|
|
|
11 |
|
12 |
# monika-test-2 (~07/2023)
|
13 |
* Trained on augmented dataset of ~10 10 second clips (augmented via tortoise tts)
|
14 |
* Trained for 100 epochs
|
|
|
15 |
|
16 |
# monika-test-4 (~07/2023)
|
17 |
* Trained on smaller but better dataset of ~2 10 second clips (augmented via 11labs)
|
18 |
* Trained for 150 epochs
|
|
|
19 |
|
20 |
# monika-test-7 (08/22/2023)
|
21 |
* Trained on augmented dataset of ~10+ 10 second clips (augmented via tortoise tts)
|
22 |
* Trained for 60 epochs (720 steps)
|
23 |
* Better quality than others
|
24 |
-
* "Version 1"
|
|
|
25 |
|
26 |
# monika-test-8 (08/22/2023)
|
27 |
* Trained on smaller but better dataset of ~5 10 second clips (some augmented via 11labs)
|
28 |
* Trained for 60 epochs (660 steps)
|
29 |
* Even clearer quality but with slightly more artifacting than monika-test-7 (still better than pre 7th ones)
|
30 |
-
* "Version 1a"
|
|
|
31 |
|
32 |
# ct-m4 (~10/2023)
|
33 |
* Trained on preprocessed version of dataset of ~5 10 second clips
|
34 |
* Trained for ~200 epochs
|
35 |
-
*
|
36 |
-
*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
# monika-test-0 (~07/2023)
|
8 |
* Trained on augmented dataset of ~10 10 second clips
|
9 |
* Trained for ~100 epochs
|
10 |
+
* RVC1
|
11 |
+
* "Version 1" ("0" in the old numbering)
|
12 |
|
13 |
# monika-test-2 (~07/2023)
|
14 |
* Trained on augmented dataset of ~10 10 second clips (augmented via tortoise tts)
|
15 |
* Trained for 100 epochs
|
16 |
+
* RVC1
|
17 |
|
18 |
# monika-test-4 (~07/2023)
|
19 |
* Trained on smaller but better dataset of ~2 10 second clips (augmented via 11labs)
|
20 |
* Trained for 150 epochs
|
21 |
+
* RVC1
|
22 |
|
23 |
# monika-test-7 (08/22/2023)
|
24 |
* Trained on augmented dataset of ~10+ 10 second clips (augmented via tortoise tts)
|
25 |
* Trained for 60 epochs (720 steps)
|
26 |
* Better quality than others
|
27 |
+
* "Version 2" ("1" in old numbering)
|
28 |
+
* RVC2
|
29 |
|
30 |
# monika-test-8 (08/22/2023)
|
31 |
* Trained on smaller but better dataset of ~5 10 second clips (some augmented via 11labs)
|
32 |
* Trained for 60 epochs (660 steps)
|
33 |
* Even clearer quality but with slightly more artifacting than monika-test-7 (still better than pre 7th ones)
|
34 |
+
* "Version 2a" ("1a" in old numbering)
|
35 |
+
* RVC2
|
36 |
|
37 |
# ct-m4 (~10/2023)
|
38 |
* Trained on preprocessed version of dataset of ~5 10 second clips
|
39 |
* Trained for ~200 epochs
|
40 |
+
* Test model
|
41 |
+
* RVC1
|
42 |
+
|
43 |
+
# ct-m4 (~10/2023)
|
44 |
+
* Trained on preprocessed version of dataset of ~5 10 second clips
|
45 |
+
* Trained for ~200 epochs
|
46 |
+
* "Version 4" ("3" in old numbering)
|
47 |
+
* RVC2, WIP
|
48 |
+
|