Uploaded V3S weights trained with a balanced dataset
Browse files- v3s_balanced/configs/vad_train_config.yaml +55 -0
- v3s_balanced/vad_model_epoch_10_step_1840.pth +3 -0
- v3s_balanced/vad_model_epoch_11_step_2024.pth +3 -0
- v3s_balanced/vad_model_epoch_12_step_2208.pth +3 -0
- v3s_balanced/vad_model_epoch_13_step_2392.pth +3 -0
- v3s_balanced/vad_model_epoch_14_step_2576.pth +3 -0
- v3s_balanced/vad_model_epoch_15_step_2760.pth +3 -0
- v3s_balanced/vad_model_epoch_16_step_2944.pth +3 -0
- v3s_balanced/vad_model_epoch_17_step_3128.pth +3 -0
- v3s_balanced/vad_model_epoch_18_step_3312.pth +3 -0
- v3s_balanced/vad_model_epoch_19_step_3496.pth +3 -0
- v3s_balanced/vad_model_epoch_1_step_184.pth +3 -0
- v3s_balanced/vad_model_epoch_20_step_3680.pth +3 -0
- v3s_balanced/vad_model_epoch_2_step_368.pth +3 -0
- v3s_balanced/vad_model_epoch_3_step_552.pth +3 -0
- v3s_balanced/vad_model_epoch_4_step_736.pth +3 -0
- v3s_balanced/vad_model_epoch_5_step_920.pth +3 -0
- v3s_balanced/vad_model_epoch_6_step_1104.pth +3 -0
- v3s_balanced/vad_model_epoch_7_step_1288.pth +3 -0
- v3s_balanced/vad_model_epoch_8_step_1472.pth +3 -0
- v3s_balanced/vad_model_epoch_9_step_1656.pth +3 -0
v3s_balanced/configs/vad_train_config.yaml
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
batch_size: 128
|
| 2 |
+
epochs: 20
|
| 3 |
+
learning_rate: 0.00001
|
| 4 |
+
weight_decay: 0.01
|
| 5 |
+
accamulation_steps: 2
|
| 6 |
+
|
| 7 |
+
is_half: false
|
| 8 |
+
num_workers: 8
|
| 9 |
+
pin_memory: true
|
| 10 |
+
|
| 11 |
+
mel_encoder_path: /mnt/ML/models/GPT-SoVITS/vad_specific/ref_enc.pth
|
| 12 |
+
mel_encoder_sr: 32000
|
| 13 |
+
filter_length: 2048
|
| 14 |
+
gin_channels: 1024
|
| 15 |
+
hop_length: 640
|
| 16 |
+
projection_weights_path: /mnt/ML/models/GPT-SoVITS/vad_specific/linear_proj.pth
|
| 17 |
+
prelu_weights_path : /mnt/ML/models/GPT-SoVITS/vad_specific/prelu.pth
|
| 18 |
+
sv_model_path: /mnt/ML/models/GPT-SoVITS/sv/pretrained_eres2netv2w24s4ep4.ckpt
|
| 19 |
+
sv_model_sr: 16000
|
| 20 |
+
use_sv: true
|
| 21 |
+
win_length: 2048
|
| 22 |
+
|
| 23 |
+
shuffle: true
|
| 24 |
+
data_dir: /mnt/ML/Datasets/laions_got_talent_enhanced_flash_annotations_and_long_captions/extracted_audio/
|
| 25 |
+
annotations_file: vad_descriptions_resampled_normalized_v3_balanced.csv
|
| 26 |
+
test_size: 5000 # set 0.1 for test run
|
| 27 |
+
limit_train_samples: -1
|
| 28 |
+
min_samples_threshold: -1
|
| 29 |
+
random_state: 42
|
| 30 |
+
|
| 31 |
+
save_dir: /mnt/ML/models/GPT-SoVITS/vad_specific/vad_scoring_model/scorer/
|
| 32 |
+
save_interval_epoch: 1
|
| 33 |
+
logging_interval: 100
|
| 34 |
+
save_interval_steps: -1
|
| 35 |
+
|
| 36 |
+
model_version: v3s
|
| 37 |
+
model_id: mkrausio/EmoWhisper-AnS-Small-v0.1
|
| 38 |
+
#vXs
|
| 39 |
+
out_channels: 16
|
| 40 |
+
#v1s
|
| 41 |
+
n_layers: 4
|
| 42 |
+
n_heads: 8
|
| 43 |
+
|
| 44 |
+
kl_cycle_length: 1
|
| 45 |
+
kl_max_weight: 0.1
|
| 46 |
+
kl_min_weight: 0.1
|
| 47 |
+
mean_loss_weight: 1.0
|
| 48 |
+
|
| 49 |
+
loss_type: kl
|
| 50 |
+
|
| 51 |
+
normalize_embeddings: true
|
| 52 |
+
|
| 53 |
+
#v4
|
| 54 |
+
emo2vec_config: /mnt/ML/models/emotion2vec/config.yaml
|
| 55 |
+
emo2vec_model_weights: /mnt/ML/models/emotion2vec/model.pt
|
v3s_balanced/vad_model_epoch_10_step_1840.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e1fba8735bfe4b6d29a712586050f1224fa7df1986fe3f93792b787d789a1d8c
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_11_step_2024.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d9938ca9f2c9d75d9349e910899e1cdcb5199474f5d53e4158254feaf9648c00
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_12_step_2208.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a96009080302d9754e513e5658aa9587da01546bf91c70f0d89dd0fe3c4d8db3
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_13_step_2392.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:777ad473c0df40a91cfa622b5eb2c4a5ceef1576ff8f19672e95cda62d613db9
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_14_step_2576.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:93cdd62af4f95994bb67d4c879077628970829819da61067e71effa5ff9fee80
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_15_step_2760.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4824b2e363aec9048abe2e139536573406c11f8d7e6689e7c5ba0d85f9e57bcd
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_16_step_2944.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ad71bf9770a84fbac1d7acc00087c2bd6748e81d43be1b778c8a3dfd5f820eb
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_17_step_3128.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3009eabf1fa65b43f53231f34afbab5de15ce88b06f520ac597e9b3afefc2c15
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_18_step_3312.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c27a88b7af39d5e88b4f797fc7000c9fd0fc7b209ed24b31f0dc3b9cf66073b4
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_19_step_3496.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7424be77a233457ef98124849403588627b7b658ca2536d07c703ce7b02cdc22
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_1_step_184.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf7190dba77f3b38c5a2dcff53b2f89dfea64be75c53f3b15e30dab021e25226
|
| 3 |
+
size 98373890
|
v3s_balanced/vad_model_epoch_20_step_3680.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52a4a4db02b69880e128551569c3a9d74b352089b6598046e65530a6818c89ce
|
| 3 |
+
size 98373918
|
v3s_balanced/vad_model_epoch_2_step_368.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c9c7e1d4873c14b3b8a34202634c7221e870b3a627dba2411744b650637ed9c
|
| 3 |
+
size 98373890
|
v3s_balanced/vad_model_epoch_3_step_552.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fdf4557b16ae2bbb0a808f2139bb613a09e8d630c7b38c5f717ddc1079fac4b
|
| 3 |
+
size 98373890
|
v3s_balanced/vad_model_epoch_4_step_736.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e9d9bd7429e32dd700d20c8e68bf78a660e500a56191f815711694aefd7748e3
|
| 3 |
+
size 98373890
|
v3s_balanced/vad_model_epoch_5_step_920.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:758dd13b5032bf1eeeb13d7ece4550ec3b711c230b4d9579a96cbd973c301b1c
|
| 3 |
+
size 98373890
|
v3s_balanced/vad_model_epoch_6_step_1104.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c67aabf3ab0764024b08bf23bb953e6382bf0cea07c38248483e1a06adf9ce83
|
| 3 |
+
size 98373904
|
v3s_balanced/vad_model_epoch_7_step_1288.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5009cc5dea4d80b0badb0fb01fe59926e2f3ede429aa10a3d68fd06aa2a2486
|
| 3 |
+
size 98373904
|
v3s_balanced/vad_model_epoch_8_step_1472.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:30c066bfbf0bef24998a8f019608b57f950cfd19e695869dfe99227d81bea9bf
|
| 3 |
+
size 98373904
|
v3s_balanced/vad_model_epoch_9_step_1656.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7854ebb839fb5cccb251ad56899d0f9742136f411054d5a480db8b01b0adca1f
|
| 3 |
+
size 98373904
|