Initial commit
Browse files- .gitattributes +1 -0
- README.md +74 -0
- pytorch_model.bin +3 -0
.gitattributes
CHANGED
@@ -6,3 +6,4 @@
|
|
6 |
*.tar.gz filter=lfs diff=lfs merge=lfs -text
|
7 |
*.ot filter=lfs diff=lfs merge=lfs -text
|
8 |
*.onnx filter=lfs diff=lfs merge=lfs -text
|
|
|
|
6 |
*.tar.gz filter=lfs diff=lfs merge=lfs -text
|
7 |
*.ot filter=lfs diff=lfs merge=lfs -text
|
8 |
*.onnx filter=lfs diff=lfs merge=lfs -text
|
9 |
+
pytorch_model.bin filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
@@ -0,0 +1,74 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- asteroid
|
4 |
+
- audio
|
5 |
+
- ConvTasNet
|
6 |
+
datasets:
|
7 |
+
- Libri2Mix
|
8 |
+
- sep_clean
|
9 |
+
license: CC BY-NC 4.0
|
10 |
+
inference: false
|
11 |
+
---
|
12 |
+
|
13 |
+
## Asteroid model `JorisCos/ConvTasNet_Libri2Mix_sepclean_16k`
|
14 |
+
|
15 |
+
Description:
|
16 |
+
|
17 |
+
This model was trained by Joris Cosentino using the librimix recipe in [Asteroid](https://github.com/asteroid-team/asteroid).
|
18 |
+
It was trained on the `sep_clean` task of the Libri2Mix dataset.
|
19 |
+
|
20 |
+
Training config:
|
21 |
+
```yaml
|
22 |
+
data:
|
23 |
+
n_src: 2
|
24 |
+
sample_rate: 16000
|
25 |
+
segment: 3
|
26 |
+
task: sep_clean
|
27 |
+
train_dir: data/wav16k/min/train-360
|
28 |
+
valid_dir: data/wav16k/min/dev
|
29 |
+
filterbank:
|
30 |
+
kernel_size: 32
|
31 |
+
n_filters: 512
|
32 |
+
stride: 16
|
33 |
+
masknet:
|
34 |
+
bn_chan: 128
|
35 |
+
hid_chan: 512
|
36 |
+
mask_act: relu
|
37 |
+
n_blocks: 8
|
38 |
+
n_repeats: 3
|
39 |
+
skip_chan: 128
|
40 |
+
optim:
|
41 |
+
lr: 0.001
|
42 |
+
optimizer: adam
|
43 |
+
weight_decay: 0.0
|
44 |
+
training:
|
45 |
+
batch_size: 6
|
46 |
+
early_stop: true
|
47 |
+
epochs: 200
|
48 |
+
half_lr: true
|
49 |
+
num_workers: 4
|
50 |
+
```
|
51 |
+
|
52 |
+
|
53 |
+
Results :
|
54 |
+
|
55 |
+
On Libri2Mix min test set :
|
56 |
+
```yaml
|
57 |
+
'si_sdr': 15.243671356901526,
|
58 |
+
'si_sdr_imp': 15.243034178473609,
|
59 |
+
'sdr': 15.668108919568112,
|
60 |
+
'sdr_imp': 15.578229918028036,
|
61 |
+
'sir': 25.295100756629957,
|
62 |
+
'sir_imp': 25.205219921301754,
|
63 |
+
'sar': 16.307682590197313,
|
64 |
+
'sar_imp': -51.64989963759405,
|
65 |
+
'stoi': 0.9394951175291422,
|
66 |
+
'stoi_imp': 0.22640192740016568
|
67 |
+
```
|
68 |
+
|
69 |
+
License notice:
|
70 |
+
|
71 |
+
This work "ConvTasNet_Libri2Mix_sepclean_16k"
|
72 |
+
is a derivative of LibriSpeech ASR corpus by Vassil Panayotov,
|
73 |
+
used under CC BY 4.0. "ConvTasNet_Libri2Mix_sepclean_16k"
|
74 |
+
is licensed under Attribution-ShareAlike 3.0 Unported by Cosentino Joris.
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8d97f012f7b2f22bb79cb0d0983a7ba27a52c1796ee3f63cbf25b4d28630adce
|
3 |
+
size 20394640
|