Hugging Face's logo

alphacep
/
vosk-vc-ru

Audio-to-Audio
Transformers
Russian
Model card Files
xet
 Community
nshmyrevgmail commited on
Commit
ce8c740
·
1 Parent(s): 64d2f17

Update checkpoint

Browse files
Files changed (3) hide show
  1. D_1642000.pth → D_191000.pth +2 -2
  2. G_1642000.pth → G_191000.pth +2 -2
  3. README.md +5 -3
D_1642000.pth → D_191000.pth RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:afb0b7fc90dcee66e95f68da224ce0a5880a45345ae5e7c3b6293d2c61639f4e
3
- size 564536506
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aae8184e99156f26e2fbd8e748e8aa26906a0dbba83ef627a252e29b8e75371d
3
+ size 564535842
G_1642000.pth → G_191000.pth RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d707c03236c35b9252018799cca2fbcd1ef4d4c7943e7acfb0524229af14a891
3
- size 480945612
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b6acc4bccf4e942f02fa5e82ad806d3c29e925a9d4a58cedd84f160d0cbc1d1
3
+ size 482123406
README.md CHANGED
@@ -7,7 +7,7 @@ pipeline_tag: audio-to-audio
7
 
8
  # About
9
 
10
- This is a basic zero-shot voice conversion model trained with VITS + softhubert
11
 
12
  See:
13
 
@@ -15,11 +15,13 @@ See:
15
 
16
  <https://github.com/quickvc/QuickVC-VoiceConversion>
17
 
 
 
18
  # Speaker Similarity
19
 
20
  Computed with eval.py with Resemblyzer
21
 
22
  ```
23
  Original QuickVC (trained on VCTK) Average: 0.667 Min: 0.477
24
- New model Average: 0.836 Min: 0.692
25
- ```
 
7
 
8
  # About
9
 
10
+ This is a basic zero-shot voice conversion model trained with VITS + contentvec
11
 
12
  See:
13
 
 
15
 
16
  <https://github.com/quickvc/QuickVC-VoiceConversion>
17
 
18
+ <https://github.com/auspicious3000/contentvec>
19
+
20
  # Speaker Similarity
21
 
22
  Computed with eval.py with Resemblyzer
23
 
24
  ```
25
  Original QuickVC (trained on VCTK) Average: 0.667 Min: 0.477
26
+ New model Average: 0.880 Min: 0.712
27
+ ```