Upload 2 files
Browse files- generator.ckpt +3 -0
- hyperparams.yaml +39 -0
generator.ckpt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5b8a36701e4f51bddf76fb19e8c43bdc0989c8a236329bc6271670cc59d53df0
|
3 |
+
size 53719290
|
hyperparams.yaml
ADDED
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
num_embeddings: 1001
|
2 |
+
embedding_dim: 1024
|
3 |
+
|
4 |
+
in_channels: 1024
|
5 |
+
out_channels: 1
|
6 |
+
resblock_type: "1"
|
7 |
+
resblock_dilation_sizes: [[1, 3, 5], [1, 3, 5], [1, 3, 5]]
|
8 |
+
resblock_kernel_sizes: [3, 7, 11]
|
9 |
+
upsample_kernel_sizes: [11, 8, 8, 4, 4]
|
10 |
+
upsample_initial_channel: 512
|
11 |
+
upsample_factors: [5, 4, 4, 2, 2]
|
12 |
+
inference_padding: 5
|
13 |
+
cond_channels: 0
|
14 |
+
conv_post_bias: True
|
15 |
+
|
16 |
+
generator: !new:speechbrain.lobes.models.HifiGAN.UnitHifiganGenerator
|
17 |
+
in_channels: !ref <in_channels>
|
18 |
+
out_channels: !ref <out_channels>
|
19 |
+
resblock_type: !ref <resblock_type>
|
20 |
+
resblock_dilation_sizes: !ref <resblock_dilation_sizes>
|
21 |
+
resblock_kernel_sizes: !ref <resblock_kernel_sizes>
|
22 |
+
upsample_kernel_sizes: !ref <upsample_kernel_sizes>
|
23 |
+
upsample_initial_channel: !ref <upsample_initial_channel>
|
24 |
+
upsample_factors: !ref <upsample_factors>
|
25 |
+
inference_padding: !ref <inference_padding>
|
26 |
+
cond_channels: !ref <cond_channels>
|
27 |
+
conv_post_bias: !ref <conv_post_bias>
|
28 |
+
num_embeddings: !ref <num_embeddings>
|
29 |
+
embedding_dim: !ref <embedding_dim>
|
30 |
+
duration_predictor: False
|
31 |
+
multi_speaker: False
|
32 |
+
skip_token_embedding: True
|
33 |
+
|
34 |
+
modules:
|
35 |
+
generator: !ref <generator>
|
36 |
+
|
37 |
+
pretrainer: !new:speechbrain.utils.parameter_transfer.Pretrainer
|
38 |
+
loadables:
|
39 |
+
generator: !ref <generator>
|