File size: 3,075 Bytes
51a86c4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5f334ec
51a86c4
 
 
3a7a7af
5f334ec
51a86c4
 
5f334ec
 
51a86c4
 
 
 
 
 
 
 
f75eb47
15b4ff5
51a86c4
ec1ce66
51a86c4
 
 
 
 
ec1ce66
550c088
 
51a86c4
550c088
51a86c4
 
ec1ce66
 
 
 
 
 
51a86c4
 
ec1ce66
 
51a86c4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
67
68
69
70
71
72
73
74
75
76
77
78
---
language: hr
datasets:
- parlaspeech-hr
tags:
- audio
- automatic-speech-recognition
- parlaspeech
widget:
- example_title: example 1
  src: https://huggingface.co./classla/wav2vec2-xls-r-parlaspeech-hr/raw/main/1800.m4a
- example_title: example 2
  src: https://huggingface.co./classla/wav2vec2-xls-r-parlaspeech-hr/raw/main/00020578b.flac.wav
- example_title: example 3
  src: https://huggingface.co./classla/wav2vec2-xls-r-parlaspeech-hr/raw/main/00020570a.flac.wav
---

# wav2vec2-large-slavic-parlaspeech-hr-lm

This model for Croatian ASR is based on the [facebook/wav2vec2-large-slavic-voxpopuli-v2 model](https://huggingface.co./facebook/wav2vec2-large-slavic-voxpopuli-v2) and was fine-tuned with 300 hours of recordings and transcripts from the ASR Croatian parliament dataset [ParlaSpeech-HR v1.0](http://hdl.handle.net/11356/1494) and enhanced with a 5-gram language model based on the [ParlaMint dataset](http://hdl.handle.net/11356/1432).

If you use this model, please cite the following paper:

Nikola Ljubešić, Danijel Koržinek, Peter Rupnik, Ivo-Pavao Jazbec. ParlaSpeech-HR -- a freely available ASR dataset for Croatian bootstrapped from the ParlaMint corpus. http://www.lrec-conf.org/proceedings/lrec2022/workshops/ParlaCLARINIII/pdf/2022.parlaclariniii-1.16.pdf

## Metrics

Evaluation is performed on the dev and test portions of the [ParlaSpeech-HR v1.0](http://hdl.handle.net/11356/1494) dataset.

|split|CER|WER|
|---|---|---|
|dev|0.0253|0.0556|
|test|0.0188|0.0430|


## Usage in `transformers`

Tested with `transformers==4.18.0`, `torch==1.11.0`, and `SoundFile==0.10.3.post1`.

```python
from transformers import Wav2Vec2ProcessorWithLM, Wav2Vec2ForCTC
import soundfile as sf
import torch
import os
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
# load model and tokenizer
processor = Wav2Vec2ProcessorWithLM.from_pretrained(
    "classla/wav2vec2-large-slavic-parlaspeech-hr-lm")
model = Wav2Vec2ForCTC.from_pretrained("classla/wav2vec2-large-slavic-parlaspeech-hr-lm")
# download the example wav files:
os.system("wget https://huggingface.co./classla/wav2vec2-large-slavic-parlaspeech-hr-lm/raw/main/00020570a.flac.wav")
# read the wav file 
speech, sample_rate = sf.read("00020570a.flac.wav")
input_values = processor(speech, sampling_rate=sample_rate, return_tensors="pt").input_values.cuda()
inputs = processor(speech, sampling_rate=sample_rate, return_tensors="pt")
with torch.no_grad():
    logits = model(**inputs).logits
transcription = processor.batch_decode(logits.numpy()).text[0]

# remove the raw wav file
os.system("rm 00020570a.flac.wav")

transcription # 'velik broj poslovnih subjekata poslao je sa minusom velik dio'
```



## Training hyperparameters

In fine-tuning, the following arguments were used:

| arg                           | value |
|-------------------------------|-------|
| `per_device_train_batch_size` | 16    |
| `gradient_accumulation_steps` | 4     |
| `num_train_epochs`            | 8     |
| `learning_rate`               | 3e-4  |
| `warmup_steps`                | 500   |