Instructions to use Supradeepdan/OverlapSpeakerCounter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- speechbrain
How to use Supradeepdan/OverlapSpeakerCounter with speechbrain:
# interface in config.json invalid
- Notebooks
- Google Colab
- Kaggle
files
Browse files- classifier.ckpt +3 -0
- embedding_model.ckpt +3 -0
- hyperparams.yaml +51 -0
- normalizer.ckpt +3 -0
- optimizer.ckpt +3 -0
classifier.ckpt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5685aaf2bf2ebbc83cca96144e452f51f0d1485d83e6413ae3cc7f620a285dad
|
| 3 |
+
size 77986
|
embedding_model.ckpt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b5a73350673d1ecd1eba1ba313c66b07ff5d86a231e06b1ea2b3c641f50d197
|
| 3 |
+
size 350831
|
hyperparams.yaml
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#Feature parameters
|
| 2 |
+
sample_rate: 16000
|
| 3 |
+
n_mels: 40
|
| 4 |
+
|
| 5 |
+
#Model parameters
|
| 6 |
+
emb_dim: 128
|
| 7 |
+
n_classes: 5
|
| 8 |
+
tdnn_channels: 64
|
| 9 |
+
tdnn_channels_out: 128
|
| 10 |
+
|
| 11 |
+
label_encoder: !new:speechbrain.dataio.encoder.CategoricalEncoder
|
| 12 |
+
|
| 13 |
+
|
| 14 |
+
#model
|
| 15 |
+
compute_features: !new:speechbrain.lobes.features.Fbank
|
| 16 |
+
n_mels: !ref <n_mels>
|
| 17 |
+
|
| 18 |
+
mean_var_norm: !new:speechbrain.processing.features.InputNormalization
|
| 19 |
+
norm_type: sentence
|
| 20 |
+
std_norm: False
|
| 21 |
+
|
| 22 |
+
embedding_model: !new:speechbrain.lobes.models.Xvector.Xvector
|
| 23 |
+
in_channels: !ref <n_mels>
|
| 24 |
+
tdnn_blocks: 5
|
| 25 |
+
tdnn_channels:
|
| 26 |
+
- !ref <tdnn_channels>
|
| 27 |
+
- !ref <tdnn_channels>
|
| 28 |
+
- !ref <tdnn_channels>
|
| 29 |
+
- !ref <tdnn_channels>
|
| 30 |
+
- !ref <tdnn_channels_out>
|
| 31 |
+
tdnn_kernel_sizes: [5, 3, 3, 1, 1]
|
| 32 |
+
tdnn_dilations: [1, 2, 3, 1, 1]
|
| 33 |
+
lin_neurons: !ref <emb_dim>
|
| 34 |
+
|
| 35 |
+
classifier: !new:speechbrain.lobes.models.Xvector.Classifier
|
| 36 |
+
input_shape: [null, null, !ref <emb_dim>]
|
| 37 |
+
activation: !name:torch.nn.LeakyReLU
|
| 38 |
+
lin_blocks: 1
|
| 39 |
+
lin_neurons: !ref <emb_dim>
|
| 40 |
+
out_neurons: !ref <n_classes>
|
| 41 |
+
|
| 42 |
+
modules:
|
| 43 |
+
compute_features: !ref <compute_features>
|
| 44 |
+
embedding_model: !ref <embedding_model>
|
| 45 |
+
classifier: !ref <classifier>
|
| 46 |
+
mean_var_norm: !ref <mean_var_norm>
|
| 47 |
+
|
| 48 |
+
pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
|
| 49 |
+
loadables:
|
| 50 |
+
embedding_model: !ref <embedding_model>
|
| 51 |
+
classifier: !ref <classifier>
|
normalizer.ckpt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eea3e3b1a844a31238e27615361a12a68bfd7b703fabeaf4478b6c61dfa8aea0
|
| 3 |
+
size 1578
|
optimizer.ckpt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:616b45a74926ef7c03865e94a027f35e5c4b41afb8b6fc05d4bc5d2359694c17
|
| 3 |
+
size 836228
|