Import from Zenodo: v1
Browse files- README.md +77 -0
- pytorch_model.bin +3 -0
README.md
ADDED
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- asteroid
|
4 |
+
- audio
|
5 |
+
- ConvTasNet
|
6 |
+
datasets:
|
7 |
+
- LibriMix
|
8 |
+
- sep_noisy
|
9 |
+
license: cc-by-sa-3.0
|
10 |
+
inference: false
|
11 |
+
---
|
12 |
+
|
13 |
+
## Asteroid model `mpariente/ConvTasNet_Libri3Mix_sep_noisy`
|
14 |
+
Imported from [Zenodo](https://zenodo.org/record/4020529)
|
15 |
+
|
16 |
+
## Description:
|
17 |
+
This model was trained by Takhir Mirzaev using the Librimix/ConvTasNet recipe in Asteroid.
|
18 |
+
It was trained on the `sep_noisy` task of the Libri3Mix dataset.
|
19 |
+
|
20 |
+
|
21 |
+
## Training config:
|
22 |
+
```yaml
|
23 |
+
data:
|
24 |
+
n_src: 3
|
25 |
+
sample_rate: 8000
|
26 |
+
segment: 3
|
27 |
+
task: sep_noisy
|
28 |
+
train_dir: data/wav8k/min/train-360
|
29 |
+
valid_dir: data/wav8k/min/dev
|
30 |
+
filterbank:
|
31 |
+
kernel_size: 16
|
32 |
+
n_filters: 512
|
33 |
+
stride: 8
|
34 |
+
masknet:
|
35 |
+
bn_chan: 128
|
36 |
+
hid_chan: 512
|
37 |
+
mask_act: relu
|
38 |
+
n_blocks: 8
|
39 |
+
n_repeats: 3
|
40 |
+
skip_chan: 128
|
41 |
+
optim:
|
42 |
+
lr: 0.001
|
43 |
+
optimizer: adam
|
44 |
+
weight_decay: 0.0
|
45 |
+
positional arguments:
|
46 |
+
training:
|
47 |
+
batch_size: 4
|
48 |
+
early_stop: True
|
49 |
+
epochs: 200
|
50 |
+
half_lr: True
|
51 |
+
num_workers: 4
|
52 |
+
```
|
53 |
+
|
54 |
+
|
55 |
+
## Results:
|
56 |
+
```yaml
|
57 |
+
si_sdr: 6.824750632456865
|
58 |
+
si_sdr_imp: 11.234803761803752
|
59 |
+
sdr: 7.715799858488098
|
60 |
+
sdr_imp: 11.778681386239114
|
61 |
+
sir: 16.442141130818637
|
62 |
+
sir_imp: 19.527535070051055
|
63 |
+
sar: 8.757864265661263
|
64 |
+
sar_imp: -0.15657258049670303
|
65 |
+
stoi: 0.7854554136619554
|
66 |
+
stoi_imp: 0.22267957718163015
|
67 |
+
```
|
68 |
+
|
69 |
+
|
70 |
+
## License notice:
|
71 |
+
This work "ConvTasNet_Libri3Mix_sepnoisy"
|
72 |
+
is a derivative of [LibriSpeech ASR corpus](http://www.openslr.org/12) by
|
73 |
+
[Vassil Panayotov](https://github.com/vdp),
|
74 |
+
used under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).
|
75 |
+
"ConvTasNet_Libri3Mix_sepnoisy"
|
76 |
+
is licensed under [Attribution-ShareAlike 3.0 Unported](https://creativecommons.org/licenses/by-sa/3.0/)
|
77 |
+
by Manuel Pariente.
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:0627aec1fd07b941dfdf9242e2be46332d047af50b35e5dacb52b44097b699a9
|
3 |
+
size 20592186
|