Wav2vec2 Conformer Rel Pos Large 960h Ft

🚀 Wav2Vec2-Conformer-Large-960h 帶相對位置嵌入

Wav2Vec2-Conformer 帶有相對位置嵌入，在16kHz採樣的語音音頻上進行了預訓練，並在960小時的Librispeech數據集上進行了微調。使用該模型時，請確保您的語音輸入也是16kHz採樣的。

📚 詳細文檔

• 論文：fairseq S2T: Fast Speech-to-Text Modeling with fairseq
• 作者：Changhan Wang, Yun Tang, Xutai Ma, Anne Wu, Sravya Popuri, Dmytro Okhonko, Juan Pino

Wav2Vec2-Conformer的實驗結果可在官方論文的表3和表4中找到。

原始模型可在https://github.com/pytorch/fairseq/tree/master/examples/wav2vec#wav2vec-20 找到。

📦 模型信息

📊 評估結果

💻 使用示例

基礎用法

from transformers import Wav2Vec2Processor, Wav2Vec2ConformerForCTC
from datasets import load_dataset
import torch

# 加載模型和處理器
processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-conformer-rel-pos-large-960h-ft")
model = Wav2Vec2ConformerForCTC.from_pretrained("facebook/wav2vec2-conformer-rel-pos-large-960h-ft")

# 加載示例數據集並讀取音頻文件
ds = load_dataset("patrickvonplaten/librispeech_asr_dummy", "clean", split="validation")

# 進行分詞
input_values = processor(ds[0]["audio"]["array"], return_tensors="pt", padding="longest").input_values

# 獲取對數概率
logits = model(input_values).logits

# 取最大值並解碼
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.batch_decode(predicted_ids)

評估示例

此代碼片段展示瞭如何在LibriSpeech的“clean”和“other”測試數據上評估facebook/wav2vec2-conformer-rel-pos-large-960h-ft。

from datasets import load_dataset
from transformers import Wav2Vec2ConformerForCTC, Wav2Vec2Processor
import torch
from jiwer import wer

librispeech_eval = load_dataset("librispeech_asr", "clean", split="test")

model = Wav2Vec2ConformerForCTC.from_pretrained("facebook/wav2vec2-large-960h-lv60-self").to("cuda")
processor = Wav2Vec2Processor.from_pretrained("facebook/wav2vec2-large-960h-lv60-self")

def map_to_pred(batch):
    inputs = processor(batch["audio"]["array"], return_tensors="pt", padding="longest")
    input_values = inputs.input_values.to("cuda")
    attention_mask = inputs.attention_mask.to("cuda")

    with torch.no_grad():
        logits = model(input_values, attention_mask=attention_mask).logits

    predicted_ids = torch.argmax(logits, dim=-1)
    transcription = processor.batch_decode(predicted_ids)
    batch["transcription"] = transcription
    return batch

result = librispeech_eval.map(map_to_pred, remove_columns=["audio"])

print("WER:", wer(result["text"], result["transcription"]))

結果 (字錯率 WER):