Wav2vec2 Large Xlsr Hindi

🚀 Wav2Vec2-Large-XLSR-53-印地語

本項目基於 [Multilingual and code - switching ASR challenges for low resource Indian languages](https://navana - tech.github.io/IS21SS - indicASRchallenge/data.html) 對 [facebook/wav2vec2 - large - xlsr - 53](https://huggingface.co/facebook/wav2vec2 - large - xlsr - 53) 模型進行印地語微調。使用該模型時，請確保語音輸入採樣率為 16kHz。

🔍 模型信息

🚀 快速開始

本模型可直接使用（無需語言模型），以下是使用示例：

💻 使用示例

基礎用法

import torch
import torchaudio
from datasets import load_dataset
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor

test_dataset = load_dataset("common_voice", "hi", split="test[:2%]")
processor = Wav2Vec2Processor.from_pretrained("theainerd/Wav2Vec2-large-xlsr-hindi")
model = Wav2Vec2ForCTC.from_pretrained("theainerd/Wav2Vec2-large-xlsr-hindi")
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])

📊 評估

可按以下方式在 Common Voice 的印地語測試數據上對模型進行評估：

import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re

test_dataset = load_dataset("common_voice", "hi", split="test")
wer = load_metric("wer")

processor = Wav2Vec2Processor.from_pretrained("theainerd/Wav2Vec2-large-xlsr-hindi")
model = Wav2Vec2ForCTC.from_pretrained("theainerd/Wav2Vec2-large-xlsr-hindi")
model.to("cuda")

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

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

# 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"])))

測試結果：72.62 %

🛠️ 訓練

用於訓練的腳本可在 [印地語自動語音識別微調 Wav2Vec2](https://colab.research.google.com/drive/1m - F7et3CHT_kpFqg7UffTIwnUV9AKgrg?usp=sharing) 找到。