Upload fixed README.md
Browse files
README.md
ADDED
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Fixed Speaker Segmentation Model
|
2 |
+
|
3 |
+
์ด ๋ชจ๋ธ์ `jaeyong2/speaker-segmentation-merge`์์ ํค ๋งคํ ๋ฌธ์ ๋ฅผ ํด๊ฒฐํ ๋ฒ์ ์
๋๋ค.
|
4 |
+
|
5 |
+
## ๋ฌธ์ ํด๊ฒฐ
|
6 |
+
- ์๋ณธ ๋ชจ๋ธ: ํค์ `model.` ์ ๋์ฌ ์์
|
7 |
+
- ํ์ฌ ๋ชจ๋ธ: ํค์ `model.` ์ ๋์ฌ ์์
|
8 |
+
- ํด๊ฒฐ: ์ ๋์ฌ ๋งคํ์ผ๋ก 100% ํค ๋งค์นญ ์ฑ๊ณต
|
9 |
+
|
10 |
+
## ์ฌ์ฉ๋ฒ
|
11 |
+
|
12 |
+
```python
|
13 |
+
from diarizers import SegmentationModel
|
14 |
+
import torch
|
15 |
+
|
16 |
+
# ๋ชจ๋ธ ๋ก๋
|
17 |
+
model = SegmentationModel()
|
18 |
+
state_dict = torch.load('pytorch_model.bin', map_location='cpu')
|
19 |
+
model.load_state_dict(state_dict)
|
20 |
+
|
21 |
+
# ์ถ๋ก
|
22 |
+
model.eval()
|
23 |
+
with torch.no_grad():
|
24 |
+
# ์ค๋์ค ์
๋ ฅ: (batch_size, audio_length)
|
25 |
+
audio = torch.randn(1, 16000) # 1์ด ์ค๋์ค ์์
|
26 |
+
output = model(audio)
|
27 |
+
print(f"Output shape: {output.shape}")
|
28 |
+
```
|
29 |
+
|
30 |
+
## ๋ชจ๋ธ ์์ธ
|
31 |
+
- ์ด ํ๋ผ๋ฏธํฐ: 54๊ฐ ๋ ์ด์ด
|
32 |
+
- ์ํคํ
์ฒ: SincNet + LSTM + Linear + Classifier
|
33 |
+
- ์
๋ ฅ: ์์ ์ค๋์ค ํํ
|
34 |
+
- ์ถ๋ ฅ: ํ์ ๋ถํ ๊ฒฐ๊ณผ
|
35 |
+
|
36 |
+
## ์๋ณธ ๋ชจ๋ธ
|
37 |
+
- Repository: jaeyong2/speaker-segmentation-merge
|
38 |
+
- ํค ๋งคํ 100% ์๋ฃ
|
39 |
+
- ๋ชจ๋ ์ฌ์ ํ๋ จ ๊ฐ์ค์น ๋ณด์กด
|