File size: 2,706 Bytes
75005c5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15bce1b
a4a72f4
af6d443
 
15bce1b
7446449
 
 
 
15bce1b
7446449
 
 
 
 
 
 
 
 
 
 
6a2760b
7446449
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
59
60
61
62
63
64
65
66
---
language: sv-SE
datasets:
- common_voice
- NST Swedish ASR Database
- P4
metrics:
- wer
tags:
- audio
- automatic-speech-recognition
- speech
license: cc0
model-index:
- name: Wav2vec 2.0 large VoxRex Swedish
  results:
  - task: 
      name: Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Common Voice
      type: common_voice
      args: sv-SE
    metrics:
       - name: Test WER
         type: wer
         value: 9.914
---
# Wav2vec 2.0 large VoxRex Swedish

Finetuned version of KBs [VoxRex large](https://huggingface.co/KBLab/wav2vec2-large-voxrex) model using Swedish radio broadcasts, NST and Common Voice data. Evalutation without a language model gives the following: WER for NST + Common Voice test set (2% of total sentences) is **3.617%**. WER for Common Voice test set is **9.914%** directly and **7.77%** with a 4-gram language model.

When using this model, make sure that your speech input is sampled at 16kHz.

## Training
This model has additionally pretrained on 3500h of a mix of Swedish local radio broadcasts, audio books and other audio sources. It has been fine-tuned for 120000 updates on NST + CommonVoice and then for an additional 20000 updates on CommonVoice only. The additional fine-tuning on CommonVoice hurts performance on the NST+CommonVoice test set somewhat and, unsurprisingly, improves it on the CommonVoice test set. It seems to perform generally better though [citation needed].

![WER during training](chart_1.svg "WER")

## Usage
The model can be used directly (without a language model) as follows:
```python
import torch
import torchaudio
from datasets import load_dataset
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
test_dataset = load_dataset("common_voice", "sv-SE", split="test[:2%]").
processor = Wav2Vec2Processor.from_pretrained("KBLab/wav2vec2-large-voxrex-swedish")
model = Wav2Vec2ForCTC.from_pretrained("KBLab/wav2vec2-large-voxpopuli-sv-swedish")
resampler = torchaudio.transforms.Resample(48_000, 16_000)
# Preprocessing the datasets.
# We need to read the aduio files as arrays
def speech_file_to_array_fn(batch):
    speech_array, sampling_rate = torchaudio.load(batch["path"])
    batch["speech"] = resampler(speech_array).squeeze().numpy()
    return batch
test_dataset = test_dataset.map(speech_file_to_array_fn)
inputs = processor(test_dataset["speech"][:2], sampling_rate=16_000, return_tensors="pt", padding=True)
with torch.no_grad():
    logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
predicted_ids = torch.argmax(logits, dim=-1)
print("Prediction:", processor.batch_decode(predicted_ids))
print("Reference:", test_dataset["sentence"][:2])
```