|
--- |
|
language: ko |
|
datasets: |
|
- zeroth_korean |
|
tags: |
|
- speech |
|
- audio |
|
- automatic-speech-recognition |
|
license: apache-2.0 |
|
--- |
|
|
|
|
|
## Evaluation on Zeroth-Korean ASR corpus |
|
|
|
[Google colab notebook(Korean)](https://colab.research.google.com/github/indra622/tutorials/blob/master/wav2vec2_korean_tutorial.ipynb) |
|
|
|
``` |
|
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor |
|
from datasets import load_dataset |
|
import soundfile as sf |
|
import torch |
|
from jiwer import wer |
|
|
|
processor = Wav2Vec2Processor.from_pretrained("kresnik/wav2vec2-large-xlsr-korean") |
|
|
|
model = Wav2Vec2ForCTC.from_pretrained("kresnik/wav2vec2-large-xlsr-korean").to('cuda') |
|
|
|
ds = load_dataset("kresnik/zeroth_korean", "clean") |
|
|
|
test_ds = ds['test'] |
|
|
|
def map_to_array(batch): |
|
speech, _ = sf.read(batch["file"]) |
|
batch["speech"] = speech |
|
return batch |
|
|
|
test_ds = test_ds.map(map_to_array) |
|
|
|
def map_to_pred(batch): |
|
inputs = processor(batch["speech"], sampling_rate=16000, return_tensors="pt", padding="longest") |
|
input_values = inputs.input_values.to("cuda") |
|
#attention_mask = inputs.attention_mask.to("cuda") |
|
|
|
with torch.no_grad(): |
|
#logits = model(input_values, attention_mask=attention_mask).logits |
|
logits = model(input_values).logits |
|
|
|
predicted_ids = torch.argmax(logits, dim=-1) |
|
transcription = processor.batch_decode(predicted_ids) |
|
batch["transcription"] = transcription |
|
return batch |
|
|
|
result = test_ds.map(map_to_pred, batched=True, batch_size=16, remove_columns=["speech"]) |
|
|
|
print("WER:", wer(result["text"], result["transcription"])) |
|
|
|
``` |
|
|
|
### Expected WER: 7.43% |
|
### Expected CER: 2.5% |