Hugging Face's logo

fierce-cats
/
beatrice-trainer

Audio-to-Audio
audio
speech
voice-conversion
Model card Files Community
7
Project Beatrice commited on
Commit
a9864f2
•
1 Parent(s): d95893e

Add pretrained models, reduce noise files, and implement minor improvements

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitignore +3 -1
  2. README.md +13 -2
  3. assets/default_config.json +3 -3
  4. assets/images/noimage.png +0 -0
  5. assets/noise/-A9bKRkhJo0.flac +0 -3
  6. assets/noise/-Ss70a9E8zk.flac +0 -3
  7. assets/noise/-Xa-AmWQ71E.flac +0 -3
  8. assets/noise/-__L3T1Yv_4.flac +0 -3
  9. assets/noise/-hdolK7F5lM.flac +0 -3
  10. assets/noise/-jT_4hrpYvo.flac +0 -3
  11. assets/noise/-n4BEVTbMP8.flac +0 -3
  12. assets/noise/-oUuF7nuv9M.flac +0 -3
  13. assets/noise/-pM0aXt3C3s.flac +0 -3
  14. assets/noise/-rxWFrTlJEY.flac +0 -3
  15. assets/noise/-u5Kjg1CBgk.flac +0 -3
  16. assets/noise/-wBISA-_DNM.flac +0 -3
  17. assets/noise/00kiylShRxU.flac +0 -3
  18. assets/noise/08INOGepJ9A.flac +0 -3
  19. assets/noise/0CMjmud6fp4.flac +0 -3
  20. assets/noise/0DNXCuOfI6A.flac +0 -3
  21. assets/noise/0Fym_Eo51p0.flac +0 -3
  22. assets/noise/0Hwm8Dj8Z5w.flac +0 -3
  23. assets/noise/0TzyMHA3mrM.flac +0 -3
  24. assets/noise/0Vzluh9sH3c.flac +0 -3
  25. assets/noise/0WTlb9Y3VdI.flac +0 -3
  26. assets/noise/0WijQqjABIs.flac +0 -3
  27. assets/noise/0fs3n8vKgO0.flac +0 -3
  28. assets/noise/0ge-wKTSBnE.flac +0 -3
  29. assets/noise/0hAEfrHwZc8.flac +0 -3
  30. assets/noise/0pU2FRZFXPw.flac +0 -3
  31. assets/noise/0wBhkeneS_0.flac +0 -3
  32. assets/noise/0wRkcrgpFCM.flac +0 -3
  33. assets/noise/0yneHvRg0dE.flac +0 -3
  34. assets/noise/1737qgz80bY.flac +0 -3
  35. assets/noise/1AXjgucDtO0.flac +0 -3
  36. assets/noise/1BZWUrkr2pw.flac +0 -3
  37. assets/noise/1CMIR6B-fbM.flac +0 -3
  38. assets/noise/1LeKrUh4e5A.flac +0 -3
  39. assets/noise/1NOjAAinHL0.flac +0 -3
  40. assets/noise/1OS22tA0s9w.flac +0 -3
  41. assets/noise/1OTi6NpPEtM.flac +0 -3
  42. assets/noise/1R53GvXZb7s.flac +0 -3
  43. assets/noise/1aSpZnfxKJk.flac +0 -3
  44. assets/noise/1fqKeYfXUT8.flac +0 -3
  45. assets/noise/1gb_KKKHf5U.flac +0 -3
  46. assets/noise/1ipPuv0QE5k.flac +0 -3
  47. assets/noise/1tViAVnYScA.flac +0 -3
  48. assets/noise/1zUvA2WId78.flac +0 -3
  49. assets/noise/21iXfj6AkqQ.flac +0 -3
  50. assets/noise/22Esf3dk-M4.flac +0 -3
.gitignore CHANGED
@@ -1 +1,3 @@
1
- poetry.lock
 
 
 
1
+ poetry.lock
2
+ work/*
3
+ __pycache__
README.md CHANGED
@@ -33,12 +33,23 @@ cd beatrice-trainer
33
  Poetry などを使用して、依存ライブラリをインストールしてください。
34
  ```sh
35
  poetry install
 
36
  ```
37
 
38
  正しくインストールできていれば、 `python3 beatrice_trainer -h` で以下のようなヘルプが表示されます。
39
 
40
  ```
41
- (TBW)
 
 
 
 
 
 
 
 
 
 
42
  ```
43
 
44
  ### 3. Prepare Your Training Data
@@ -85,7 +96,7 @@ your_training_data_dir_with_only_one_speaker
85
  python3 beatrice_trainer -d <your_training_data_dir> -o <output_dir>
86
  ```
87
 
88
- 学習の状況は、 TensorBoard で確認することができます。
89
 
90
  ```sh
91
  tensorboard --logdir <output_dir>
 
33
  Poetry などを使用して、依存ライブラリをインストールしてください。
34
  ```sh
35
  poetry install
36
+ poetry shell
37
  ```
38
 
39
  正しくインストールできていれば、 `python3 beatrice_trainer -h` で以下のようなヘルプが表示されます。
40
 
41
  ```
42
+ usage: beatrice_trainer [-h] [-d DATA_DIR] [-o OUT_DIR] [-r] [-c CONFIG]
43
+
44
+ options:
45
+ -h, --help show this help message and exit
46
+ -d DATA_DIR, --data_dir DATA_DIR
47
+ directory containing the training data
48
+ -o OUT_DIR, --out_dir OUT_DIR
49
+ output directory
50
+ -r, --resume resume training
51
+ -c CONFIG, --config CONFIG
52
+ path to the config file
53
  ```
54
 
55
  ### 3. Prepare Your Training Data
 
96
  python3 beatrice_trainer -d <your_training_data_dir> -o <output_dir>
97
  ```
98
 
99
+ 学習の状況は、 TensorBoard で確認できます。
100
 
101
  ```sh
102
  tensorboard --logdir <output_dir>
assets/default_config.json CHANGED
@@ -12,8 +12,8 @@
12
  "grad_weight_fm": 1.0,
13
  "grad_balancer_ema_decay": 0.995,
14
  "use_amp": true,
15
- "num_workers": 8,
16
- "n_steps": 3000000,
17
  "warmup_steps": 10000,
18
  "in_sample_rate": 16000,
19
  "out_sample_rate": 24000,
@@ -24,7 +24,7 @@
24
  "in_ir_wav_dir": "assets/ir",
25
  "in_noise_wav_dir": "assets/noise",
26
  "in_test_wav_dir": "assets/test",
27
- "pretrained_file": null,
28
  "hidden_channels": 256,
29
  "san": false
30
  }
 
12
  "grad_weight_fm": 1.0,
13
  "grad_balancer_ema_decay": 0.995,
14
  "use_amp": true,
15
+ "num_workers": 16,
16
+ "n_steps": 100000,
17
  "warmup_steps": 10000,
18
  "in_sample_rate": 16000,
19
  "out_sample_rate": 24000,
 
24
  "in_ir_wav_dir": "assets/ir",
25
  "in_noise_wav_dir": "assets/noise",
26
  "in_test_wav_dir": "assets/test",
27
+ "pretrained_file": "assets/pretrained/040c_checkpoint_libritts_r_200_02300000.pt",
28
  "hidden_channels": 256,
29
  "san": false
30
  }
assets/images/noimage.png ADDED
assets/noise/-A9bKRkhJo0.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:8831477c14bc07c9f700f8094439e01f5ca1ddc915c1c7ef018d9faa01ce168d
3
- size 210350
 
 
 
 
assets/noise/-Ss70a9E8zk.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c2515daaff3b4aebe187a756cc620964b3eabe62673636ac8af14f1619c99e26
3
- size 239568
 
 
 
 
assets/noise/-Xa-AmWQ71E.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3f12b78052806adc2076b702bacef203a35f18900b8fa2d474ff94d0c43b9528
3
- size 234305
 
 
 
 
assets/noise/-__L3T1Yv_4.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:32f5f64e617fa9cc4f907a2866d8ebf91139beff46a63719f1822cff7cd2e264
3
- size 389162
 
 
 
 
assets/noise/-hdolK7F5lM.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:e2c543cf9966c9636886f183bf9d5cd5cbc2788702cc17cdcf4da1c58a770da7
3
- size 361831
 
 
 
 
assets/noise/-jT_4hrpYvo.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:de0be10906c278b88852f4a9925241b591706b004a082a28f1be87c00db01996
3
- size 393231
 
 
 
 
assets/noise/-n4BEVTbMP8.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3558f6d20510ff0aa1ddbbe51c03b05bebfc0bc3bce09b862eb78c548e2e093e
3
- size 294133
 
 
 
 
assets/noise/-oUuF7nuv9M.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1f375a59a53e276e7fb34da370dbc0a6cbe373bd78e13632023ab4d507aad016
3
- size 253486
 
 
 
 
assets/noise/-pM0aXt3C3s.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:2b001ca895b8f140dbd86473c5c486425bf8f1813c40a668ac39c2ca92470f32
3
- size 359934
 
 
 
 
assets/noise/-rxWFrTlJEY.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1501c95a153fbdf46b877c79f90df5ca69b8d8f9c86e751a95305529d58da355
3
- size 200294
 
 
 
 
assets/noise/-u5Kjg1CBgk.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:daf1dfa7a08ec955ea634ad7dc71ecc766a7f0a6a801b52ed9c797814c3ee1f5
3
- size 210098
 
 
 
 
assets/noise/-wBISA-_DNM.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:414cfa239d234310a348823827324e2b4a3bfa660285ed7596ebe98a5c0212cf
3
- size 260563
 
 
 
 
assets/noise/00kiylShRxU.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:249dce5140933edb6bc93cd10788820b0b97c58547c3a0b66b4ce5e55a220034
3
- size 169200
 
 
 
 
assets/noise/08INOGepJ9A.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:fac6cbe53d3f8854106a490561e37c9d07a91fb2eef93709675ece36f1b1ed4a
3
- size 339608
 
 
 
 
assets/noise/0CMjmud6fp4.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:b51f1b25cfa14497cb99b710601ee0c2b9a791d67a4f740914baaedb9c4d3279
3
- size 336632
 
 
 
 
assets/noise/0DNXCuOfI6A.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:41748d5fc95b49b7f1bceda8712be443edccb93a42323e4e366e268e4680ad94
3
- size 230957
 
 
 
 
assets/noise/0Fym_Eo51p0.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d0a1918c17f2accfeed5e0c9cd146f1c3e074a73e5af4deefe6e12e37a5c25fa
3
- size 225951
 
 
 
 
assets/noise/0Hwm8Dj8Z5w.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:652d0665744bc1d7bfbb51d48559f29641177985af3556be98b7552e4d4737aa
3
- size 274150
 
 
 
 
assets/noise/0TzyMHA3mrM.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3a8f75933331f4e7894fef7c9fb8934f5e21d6c7e99017346e2c1ad5c7a0f559
3
- size 147038
 
 
 
 
assets/noise/0Vzluh9sH3c.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7aa3dff34e64d85e87bb40d7629e19ef2a3798736820770b824c84d5eccded42
3
- size 337686
 
 
 
 
assets/noise/0WTlb9Y3VdI.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d3275b06f809de70b8401130cbc6f8dd55d9fc3a85904f04de78e37bb9b47038
3
- size 192873
 
 
 
 
assets/noise/0WijQqjABIs.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:29f93246aadfdcca4d256a5056988adb6ef31d6c6658755a96aba029896a0613
3
- size 331514
 
 
 
 
assets/noise/0fs3n8vKgO0.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:10580fae9f93bcb89a161a1c4758f18211e8e8f578a2740d1b6138f89a764345
3
- size 270221
 
 
 
 
assets/noise/0ge-wKTSBnE.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:5405c9bd67123941e4528aeb8506cb9ba1f89593abc67e8222dae83039b3430c
3
- size 310586
 
 
 
 
assets/noise/0hAEfrHwZc8.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c53ec08530731e9abfd54c3c1a899e1f3b431c9b989752163af7cdcd8816a451
3
- size 245171
 
 
 
 
assets/noise/0pU2FRZFXPw.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:3f5628d5d6516725700b5858162d4cd00b5c9ef6ec2bdbe7f57cbc72f614f66c
3
- size 294910
 
 
 
 
assets/noise/0wBhkeneS_0.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:07100c9a2b3718d1f4f0bf308d539b20915f9f4b6387620305951a7e7e61ed83
3
- size 174550
 
 
 
 
assets/noise/0wRkcrgpFCM.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:db7842fc105365cd06ed68767088760b0634dbce0807273a902be8ce241a4103
3
- size 172493
 
 
 
 
assets/noise/0yneHvRg0dE.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:e67b0931fbcf6e55faa1e6a152d04214b2003864f1f3cf8de13a4ea347890dcf
3
- size 181030
 
 
 
 
assets/noise/1737qgz80bY.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ef9de7e45b0079e998b4ab8c69af242bbc45b06d928190216a8aabd00b8c61d9
3
- size 297763
 
 
 
 
assets/noise/1AXjgucDtO0.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ac0dd3de9cbf8aa5f901ed6a19ab80a6dfb67268f59b314d2e5f3c1c117906e9
3
- size 280709
 
 
 
 
assets/noise/1BZWUrkr2pw.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c85f772590bf220afa48351106e98c8c269489d02e498ac7623894bd75a406ef
3
- size 274302
 
 
 
 
assets/noise/1CMIR6B-fbM.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:2ba56bb37a20258864358f338f3ab499d375658e4008af492185c1b0bf68ae0c
3
- size 199680
 
 
 
 
assets/noise/1LeKrUh4e5A.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:04fe748541e1dc79896abe6f614b9af9ad1e831562a0cdae318de6e5ca8381f1
3
- size 392647
 
 
 
 
assets/noise/1NOjAAinHL0.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:351b6d7422e63dcbdc821d3845569050340cb64e81ecf158abd65a0d00f4a83a
3
- size 337522
 
 
 
 
assets/noise/1OS22tA0s9w.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d9c44a102e84e86f11d3fb1f75becc5fbe0d5ee8d4cf3372e03bca6217ea8f5b
3
- size 404735
 
 
 
 
assets/noise/1OTi6NpPEtM.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:1f596c034e83f62442a93ac1639e689fc6e72ee098a471830ba79b4d44569dc7
3
- size 222837
 
 
 
 
assets/noise/1R53GvXZb7s.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:a74265765c1b1d97cb282326831ebe056caf0cfa9b9a2d96449b01df5e7889f9
3
- size 306838
 
 
 
 
assets/noise/1aSpZnfxKJk.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:b6bb290f14b1efeacbbac1ad8e7f70325ce41d10c30fd9b829b1c2a1dc0951a6
3
- size 245944
 
 
 
 
assets/noise/1fqKeYfXUT8.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f86102aeac8ef1947eccf4b866377b5ca75e5557d25be0d6edcaf3dc74eda2ed
3
- size 325983
 
 
 
 
assets/noise/1gb_KKKHf5U.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:013b70c58e15f3ec4ccdb207c22d3d9438755ca59c2db375aac70d53a7f8e3af
3
- size 279831
 
 
 
 
assets/noise/1ipPuv0QE5k.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9d9ba0660098e324832e5c1a5288ee0f40ed39ab44e35679d8c1711e33f75ec7
3
- size 395875
 
 
 
 
assets/noise/1tViAVnYScA.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:2fc54832224908113a409b0d2b10d90877620d2b066a6b6b10d106aa6cf5fd33
3
- size 379462
 
 
 
 
assets/noise/1zUvA2WId78.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:fd77d126c90f8f15d0d26ae156bae8590524986408fe1e4997c17203528c13aa
3
- size 351434
 
 
 
 
assets/noise/21iXfj6AkqQ.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:b1daceb9015b45f815cea48081615773de85b8803b23e780dcc8669d6eaaa17e
3
- size 227499
 
 
 
 
assets/noise/22Esf3dk-M4.flac DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:4372b2cc4dce65b715bccab9c18bbae9144ac636bef8eb4ee553c564b1a21c09
3
- size 248360