File size: 4,768 Bytes
54bcd54 08421c5 0d953a4 08421c5 4a8b27a 08421c5 4a8b27a 08421c5 4a8b27a 08421c5 4a8b27a 08421c5 4a8b27a 08421c5 4a8b27a 08421c5 4a8b27a 08421c5 0d953a4 08421c5 5a9f5d9 1e05833 75a4460 5a9f5d9 75a4460 5a9f5d9 75a4460 5a9f5d9 75a4460 5a9f5d9 75a4460 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 |
---
language: dv
datasets:
- common_voice
metrics:
- wer
tags:
- audio
- automatic-speech-recognition
- speech
- xlsr-fine-tuning-week
license: apache-2.0
model-index:
- name: Shahu Kareem XLSR Wav2Vec2 Large 53 Dhivehi
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice dv
type: common_voice
args: dv
metrics:
- name: Test WER
type: wer
value: 32.85
---
# Wav2Vec2-Large-XLSR-53-Dhivehi
Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on Dhivehi using the [Common Voice](https://huggingface.co/datasets/common_voice).
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
test_dataset = load_dataset("common_voice", "dv", split="test[:2%]")
processor = Wav2Vec2Processor.from_pretrained("shahukareem/wav2vec2-large-xlsr-53-dhivehi")
model = Wav2Vec2ForCTC.from_pretrained("shahukareem/wav2vec2-large-xlsr-53-dhivehi")
# 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])
```
## Evaluation
The model can be evaluated as follows on the Dhivehi test data of Common Voice.
```python
import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re
test_dataset = load_dataset("common_voice", "dv", split="test")
wer = load_metric("wer")
processor = Wav2Vec2Processor.from_pretrained("shahukareem/wav2vec2-large-xlsr-53-dhivehi")
model = Wav2Vec2ForCTC.from_pretrained("shahukareem/wav2vec2-large-xlsr-53-dhivehi")
model.to("cuda")
chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“\%\‘\”\�\،\.\؟\!\'\"\–\’]'
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):
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
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)
# Preprocessing the datasets.
# We need to read the aduio files as arrays
def evaluate(batch):
inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
with torch.no_grad():
logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["pred_strings"] = processor.batch_decode(pred_ids)
return 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**: 32.85%
## Training
The Common Voice `train` and `validation` datasets were used for training.
## Example predictions
```--
reference: ކަރަންޓް ވައިރުކޮށް ބޮކި ހަރުކުރުން
predicted: ކަރަންޓް ވައިރުކޮށް ބޮކި ހަރުކުރުން
--
reference: ދެން އެކުދިންނާ ދިމާއަށް އަތް ދިށްކޮށްލެވެ
predicted: ދެން އެކުދިންނާ ދިމާއަށް އަތް ދިއްކޮށްލެވެ ް
--
reference: ރަކި ހިނިތުންވުމަކާއެކު އޭނާ އަމިއްލައަށް ތައާރަފްވި
predicted: ރަކި ހިނިތުންވުމަކާއެކު އޭނާ އަމިއްލައަށް ތައަރަފްވި
--
reference: ކޮޓަރީގެ ކުޑަދޮރުން ބޭރު ބަލަހައްޓައިގެން އިން ރޫނާގެ މޫނުމަތިން ފާޅުވަމުން ދިޔައީ ކަންބޮޑުވުމުގެ އަސަރުތައް
predicted: ކޮޓަރީގެ ކުޑަދޮރުން ބޭރު ބަލަހައްޓައިގެން އިން ރނާގެ މޫނުމަތިން ފާޅުވަމުން ދިޔައީ ކަންބޮޑުވުމުގެ އަސަރުތައް
--
``` |