File size: 1,094 Bytes
07ba25b
 
 
3f2064c
07ba25b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3f2064c
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
---
license: apache-2.0
datasets:
- Bingsu/zeroth-korean
language:
- ko
metrics:
- cer
- wer
pipeline_tag: automatic-speech-recognition
---
# Whisper-Medium-KsponSpeech

The Whisper-medium Model finetunned with [KsponSpeech](https://huggingface.co/datasets/Murple/ksponspeech)


### Model Description

<!-- Provide a longer summary of what this model is. -->



- **Developed by :** [yw0nam](https://github.com/yw0nam)
- **Shared by :** [yw0nam](https://github.com/yw0nam)
- **Model type :** ASR
- **License:** [apache-2.0]

## Uses

```

processor = WhisperProcessor.from_pretrained("openai/whisper-medium", language="ko", task="transcribe")
model = WhisperForConditionalGeneration.from_pretrained('spow12/whisper-medium-zeroth_korean').cuda()

data, _ = librosa.load(wav_path, sr=16000)
input_features = processor(data, sampling_rate=16000, return_tensors="pt").input_features.cuda()

predicted_ids = model.generate(input_features)
transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)[0]

```

### Metrics

Metric | result |
--- | --- | 
 WER | 3.96 |
 CER | 1.71 |