🚀 Wav2Vec2-Large-XLSR-53-Chuvash
本项目基于 Common Voice 数据集,对楚瓦什语(Chuvash)微调了 facebook/wav2vec2-large-xlsr-53 模型。使用该模型时,请确保输入的语音采样率为 16kHz。
🚀 快速开始
本模型可直接使用(无需语言模型),使用方法如下:
💻 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", "cv", split="test[:2%]")
processor = Wav2Vec2Processor.from_pretrained("anton-l/wav2vec2-large-xlsr-53-chuvash")
model = Wav2Vec2ForCTC.from_pretrained("anton-l/wav2vec2-large-xlsr-53-chuvash")
resampler = torchaudio.transforms.Resample(48_000, 16_000)
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])
模型评估
可以按照以下方式在 Common Voice 的楚瓦什语测试数据上对模型进行评估:
高级用法
import torch
import torchaudio
import urllib.request
import tarfile
import pandas as pd
from tqdm.auto import tqdm
from datasets import load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
data_url = "https://voice-prod-bundler-ee1969a6ce8178826482b88e843c335139bd3fb4.s3.amazonaws.com/cv-corpus-6.1-2020-12-11/cv.tar.gz"
filestream = urllib.request.urlopen(data_url)
data_file = tarfile.open(fileobj=filestream, mode="r|gz")
data_file.extractall()
wer = load_metric("wer")
processor = Wav2Vec2Processor.from_pretrained("anton-l/wav2vec2-large-xlsr-53-chuvash")
model = Wav2Vec2ForCTC.from_pretrained("anton-l/wav2vec2-large-xlsr-53-chuvash")
model.to("cuda")
cv_test = pd.read_csv("cv-corpus-6.1-2020-12-11/cv/test.tsv", sep='\t')
clips_path = "cv-corpus-6.1-2020-12-11/cv/clips/"
def clean_sentence(sent):
sent = sent.lower()
sent = "".join(ch if ch.isalpha() else " " for ch in sent)
sent = " ".join(sent.split())
return sent
targets = []
preds = []
for i, row in tqdm(cv_test.iterrows(), total=cv_test.shape[0]):
row["sentence"] = clean_sentence(row["sentence"])
speech_array, sampling_rate = torchaudio.load(clips_path + row["path"])
resampler = torchaudio.transforms.Resample(sampling_rate, 16_000)
row["speech"] = resampler(speech_array).squeeze().numpy()
inputs = processor(row["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)
targets.append(row["sentence"])
preds.append(processor.batch_decode(pred_ids)[0])
print("WER: {:2f}".format(100 * wer.compute(predictions=preds, references=targets)))
测试结果:40.01 %
📚 详细文档
训练信息
训练使用了 Common Voice 的 train
和 validation
数据集。训练脚本可在 此处 找到。
📄 许可证
本项目采用 Apache-2.0 许可证。
📦 模型信息
属性 |
详情 |
模型类型 |
基于 Wav2Vec2-Large-XLSR-53 微调的楚瓦什语语音识别模型 |
训练数据 |
Common Voice 数据集的楚瓦什语部分 |
评估指标 |
词错误率(WER) |
测试 WER |
40.01% |