File size: 1,604 Bytes
1e8d6bc |
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 49 50 51 52 53 54 55 56 57 58 |
---
language: nl
datasets:
- common_voice
tags:
- audio
- automatic-speech-recognition
- phoneme-recognition
model-index:
- name: wav2vec2-base-960h-phoneme-reco-dutch
results:
- task:
name: Automatic Phoneme Recognition
type: automatic-phoneme-recognition
dataset:
name: CommonVoice (clean)
type: librispeech_asr
config: clean
split: test
args:
language: nl
metrics:
- name: Test PER
type: per
value: 20.83
- name: Val PER
type: per
value: 16.18
---
The Wav2vec2 base model [facebook/wav2vec2-base-960h](https://huggingface.co/facebook/wav2vec2-base-960h) fine tuned on phoneme recognition task for the dutch language.
# Usage
To transcribe in phonemes audio files the model can be used as a standalone acoustic model as follows:
```python
from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
from datasets import load_dataset
import torch
# load model and tokenizer
processor = Wav2Vec2Processor.from_pretrained("Clementapa/wav2vec2-base-960h-phoneme-reco-dutch")
model = Wav2Vec2ForCTC.from_pretrained("Clementapa/wav2vec2-base-960h-phoneme-reco-dutch")
# load dummy dataset and read soundfiles
ds = load_dataset("common_voice", "nl", split="validation")
# tokenize
input_values = processor(ds[0]["audio"]["array"], return_tensors="pt", padding="longest").input_values # Batch size 1
# retrieve logits
logits = model(input_values).logits
# take argmax and decode
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.batch_decode(predicted_ids)
``` |