Wav2vec2 Large Xlsr 53 Vietnamese

🚀 Wav2Vec2-Large-XLSR-53-越南语模型

本模型基于 facebook/wav2vec2-large-xlsr-53 在越南语数据集上进行微调，使用的数据集为 Common Voice。使用该模型时，请确保输入语音的采样率为 16kHz。

✨ 主要特性

• 数据集：使用 Common Voice 数据集进行训练和评估。
• 指标：使用字错率（WER）作为评估指标。
• 应用场景：适用于音频处理、自动语音识别等领域。

📦 安装指南

文档未提及具体安装步骤，可参考 Hugging Face 相关库的安装方式，如安装 transformers、datasets、torch、torchaudio 等库：

pip install transformers datasets torch torchaudio

💻 使用示例

基础用法

import torch
import torchaudio
from datasets import load_dataset
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
test_dataset = load_dataset("common_voice", "vi", split="test[:2%]")
processor = Wav2Vec2Processor.from_pretrained("anuragshas/wav2vec2-large-xlsr-53-vietnamese")
model = Wav2Vec2ForCTC.from_pretrained("anuragshas/wav2vec2-large-xlsr-53-vietnamese")
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])

高级用法

以下是对模型进行评估的代码示例：

import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re
test_dataset = load_dataset("common_voice", "vi", split="test")
wer = load_metric("wer")
processor = Wav2Vec2Processor.from_pretrained("anuragshas/wav2vec2-large-xlsr-53-vietnamese")
model = Wav2Vec2ForCTC.from_pretrained("anuragshas/wav2vec2-large-xlsr-53-vietnamese")
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"])))

测试结果：66.78 %

🔧 技术细节

本模型基于 facebook/wav2vec2-large-xlsr-53 进行微调，使用 Common Voice 数据集的 train 和 validation 子集进行训练。评估时使用字错率（WER）作为指标，在 Common Voice 越南语测试集上的测试结果为 66.78%。

📄 许可证

本模型使用 Apache-2.0 许可证。

📦 模型信息