File size: 3,870 Bytes
c94712e b6df156 c94712e b6df156 c94712e b6df156 c94712e b6df156 c94712e b6df156 c94712e b6df156 c94712e b6df156 c94712e b6df156 |
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 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 |
---
language: kk
datasets:
- kazakh_speech_corpus
metrics:
- wer
tags:
- audio
- automatic-speech-recognition
- speech
- xlsr-fine-tuning-week
license: apache-2.0
model-index:
- name: Wav2Vec2-XLSR-53 Kazakh by adilism
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Kazakh Speech Corpus v1.1
type: kazakh_speech_corpus
args: kk
metrics:
- name: Test WER
type: wer
value: 22.84
---
# Wav2Vec2-Large-XLSR-53-Kazakh
Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co./facebook/wav2vec2-large-xlsr-53) on Kazakh using the [Kazakh Speech Corpus v1.1](https://issai.nu.edu.kz/kz-speech-corpus/?version=1.1)
When using this model, make sure that your speech input is sampled at 16kHz.
## 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
from utils import get_test_dataset
test_dataset = get_test_dataset("ISSAI_KSC_335RS_v1.1")
processor = Wav2Vec2Processor.from_pretrained("wav2vec2-large-xlsr-kazakh")
model = Wav2Vec2ForCTC.from_pretrained("wav2vec2-large-xlsr-kazakh")
# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
\tspeech_array, sampling_rate = torchaudio.load(batch["path"])
\tbatch["speech"] = torchaudio.transforms.Resample(sampling_rate, 16_000)(speech_array).squeeze().numpy()
\treturn 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():
\tlogits = 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])
```
## Evaluation
The model can be evaluated as follows on the test data of [Kazakh Speech Corpus v1.1](https://issai.nu.edu.kz/kz-speech-corpus/?version=1.1). To evaluate, download the [archive](https://www.openslr.org/resources/102/ISSAI_KSC_335RS_v1.1_flac.tar.gz), untar and pass the path to data to `get_test_dataset` as below:
```python
import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re
from utils import get_test_dataset
test_dataset = get_test_dataset("ISSAI_KSC_335RS_v1.1")
wer = load_metric("wer")
processor = Wav2Vec2Processor.from_pretrained("adilism/wav2vec2-large-xlsr-kazakh")
model = Wav2Vec2ForCTC.from_pretrained("adilism/wav2vec2-large-xlsr-kazakh")
model.to("cuda")
# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
\tbatch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
\tspeech_array, sampling_rate = torchaudio.load(batch["path"])
\tbatch["speech"] = torchaudio.transforms.Resample(sampling_rate, 16_000)(speech_array).squeeze().numpy()
\treturn batch
test_dataset = test_dataset.map(speech_file_to_array_fn)
def evaluate(batch):
\tinputs = processor(batch["text"], sampling_rate=16_000, return_tensors="pt", padding=True)
\twith torch.no_grad():
\t\tlogits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
\tpred_ids = torch.argmax(logits, dim=-1)
\tbatch["pred_strings"] = processor.batch_decode(pred_ids)
\treturn batch
result = test_dataset.map(evaluate, batched=True, batch_size=8)
print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
```
**Test Result**: 22.84 %
## Training
The Kazakh Speech Corpus v1.1 `train` dataset was used for training, |