zhanjun commited on
Commit
3db6555
1 Parent(s): 729edd4

Upload folder using huggingface_hub

Browse files
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ speechtokenizer/ckpt.dev filter=lfs diff=lfs merge=lfs -text
speechtokenizer/ckpt.dev ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5687870a9c6d44c3defef4d376c64e6c0b75281e868a95256a14e9b577098eb
3
+ size 481944941
speechtokenizer/config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "resblock": "1",
3
+ "num_gpus": 3,
4
+ "batch_size": 60,
5
+ "learning_rate": 0.0001,
6
+ "adam_b1": 0.5,
7
+ "adam_b2": 0.9,
8
+ "lr_decay": 0.98,
9
+ "seed": 1234,
10
+ "lambda_distill": 0.15,
11
+
12
+ "n_filters": 64,
13
+ "strides": [8,5,4,2],
14
+ "dimension": 1024,
15
+ "semantic_dimension": 768,
16
+ "bidirectional": true,
17
+ "dilation_base": 2,
18
+ "residual_kernel_size": 3,
19
+ "n_residual_layers": 1,
20
+ "lstm_layers": 2,
21
+ "activation": "Snake",
22
+
23
+
24
+ "segment_size": 48000,
25
+ "num_mels": 80,
26
+ "num_freq": 1025,
27
+ "n_fft": 1024,
28
+ "hop_size": 240,
29
+ "win_size": 1024,
30
+
31
+ "sampling_rate": 16000,
32
+ "sample_rate": 16000,
33
+
34
+ "codebook_size": 1024,
35
+ "n_q": 8,
36
+ "codebook_dim": 512,
37
+ "codebook_loss_lambda": 1.0,
38
+ "commitment_loss_lambda": 0.25,
39
+
40
+ "fmin": 0,
41
+ "fmax": 8000,
42
+ "fmax_for_loss": null,
43
+
44
+ "num_workers": 12,
45
+
46
+ "dist_config": {
47
+ "dist_backend": "nccl",
48
+ "dist_url": "tcp://localhost:54321",
49
+ "world_size": 1
50
+ }
51
+ }