语言: ka

数据集:

• common_voice

标签:

• 音频

• 自动语音识别

• 语音

• xlsr微调周

许可证: apache-2.0

模型索引:

• 名称: Mehdi Hosseini Moghadam的wav2vec2-large-xlsr-53-Georgian

  结果:

  • 任务:

    名称: 语音识别

    类型: 自动语音识别

    数据集:

    名称: Common Voice ka

    类型: common_voice

    参数: ka

    指标:

    • 名称: 测试WER

      类型: wer

      值: 60.504024

wav2vec2-large-xlsr-53-Georgian

使用Common Voice对facebook/wav2vec2-large-xlsr-53进行格鲁吉亚语微调

使用此模型时，请确保语音输入采样率为16kHz。

使用方法

该模型可以直接使用（无需语言模型）如下：

import torch

import torchaudio

from datasets import load_dataset

from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor

test_dataset = load_dataset("common_voice", "ka", split="test[:2%]")

processor = Wav2Vec2Processor.from_pretrained("MehdiHosseiniMoghadam/wav2vec2-large-xlsr-53-Georgian")

model = Wav2Vec2ForCTC.from_pretrained("MehdiHosseiniMoghadam/wav2vec2-large-xlsr-53-Georgian")

resampler = torchaudio.transforms.Resample(48_000, 16_000)

# 数据预处理

# 我们需要将音频文件读取为数组

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("预测结果:", processor.batch_decode(predicted_ids))

print("参考文本:", test_dataset["sentence"][:2])

评估

可以在Common Voice的格鲁吉亚语测试数据上如下评估该模型。

import torch

import torchaudio

from datasets import load_dataset, load_metric

from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor

import re

test_dataset = load_dataset("common_voice", "ka", split="test")

wer = load_metric("wer")

processor = Wav2Vec2Processor.from_pretrained("MehdiHosseiniMoghadam/wav2vec2-large-xlsr-53-Georgian")

model = Wav2Vec2ForCTC.from_pretrained("MehdiHosseiniMoghadam/wav2vec2-large-xlsr-53-Georgian")

model.to("cuda")

chars_to_ignore_regex = '[\\,\\?\\.\\!\\-\\;\\:\\"\\“]'

resampler = torchaudio.transforms.Resample(48_000, 16_000)

# 数据预处理

# 我们需要将音频文件读取为数组

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)

# 评估函数

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: 60.504024 %

训练

使用了Common Voice的train和validation数据集进行训练。