许可证：apache-2.0
语言：德语
库名称：transformers
缩略图：无
标签：

• 自动语音识别
• whisper-event
  数据集：
• mozilla-foundation/common_voice_11_0
  评估指标：
• 词错误率（WER）

模型索引：

• 名称：针对德语ASR微调的whisper-large-v2模型
  结果：
  • 任务：
    名称：自动语音识别
    类型：automatic-speech-recognition
    数据集：
    名称：Common Voice 11.0
    类型：mozilla-foundation/common_voice_11_0
    配置：德语
    拆分：测试集
    参数：德语
    评估指标：
    • 名称：贪婪解码WER
      类型：wer
      值：5.76

针对德语ASR微调的whisper-large-v2模型

本模型是基于openai/whisper-large-v2在mozilla-foundation/common_voice_11_0德语数据集上微调的版本。使用时请确保语音输入采样率为16kHz。该模型可预测大小写和标点符号。

性能表现

下表为预训练模型在Common Voice 9.0的WER结果，数据源自原始论文。

下表为微调模型在Common Voice 11.0的WER结果。

使用方式

使用🤗 Pipeline推理

import torch  
from datasets import load_dataset  
from transformers import pipeline  

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")  

# 加载pipeline  
pipe = pipeline("automatic-speech-recognition", model="bofenghuang/whisper-large-v2-cv11-german", device=device)  

# 注意：为生成工具设置强制解码器ID  
pipe.model.config.forced_decoder_ids = pipe.tokenizer.get_decoder_prompt_ids(language="de", task="transcribe")  

# 加载数据  
ds_mcv_test = load_dataset("mozilla-foundation/common_voice_11_0", "de", split="test", streaming=True)  
test_segment = next(iter(ds_mcv_test))  
waveform = test_segment["audio"]  

# 注意：解码选项  
# 限制生成标记最大数为225  
pipe.model.config.max_length = 225 + 1  
# 采样  
# pipe.model.config.do_sample = True  
# 束搜索  
# pipe.model.config.num_beams = 5  
# 返回  
# pipe.model.config.return_dict_in_generate = True  
# pipe.model.config.output_scores = True  
# pipe.model.config.num_return_sequences = 5  

# 执行  
generated_sentences = pipe(waveform)["text"]  

使用🤗底层API推理

import torch  
import torchaudio  
from datasets import load_dataset  
from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq  

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")  

# 加载模型  
model = AutoModelForSpeechSeq2Seq.from_pretrained("bofenghuang/whisper-large-v2-cv11-german").to(device)  
processor = AutoProcessor.from_pretrained("bofenghuang/whisper-large-v2-cv11-german", language="german", task="transcribe")  

# 注意：为生成工具设置强制解码器ID  
model.config.forced_decoder_ids = processor.get_decoder_prompt_ids(language="de", task="transcribe")  

# 16kHz  
model_sample_rate = processor.feature_extractor.sampling_rate  

# 加载数据  
ds_mcv_test = load_dataset("mozilla-foundation/common_voice_11_0", "de", split="test", streaming=True)  
test_segment = next(iter(ds_mcv_test))  
waveform = torch.from_numpy(test_segment["audio"]["array"])  
sample_rate = test_segment["audio"]["sampling_rate"]  

# 重采样  
if sample_rate != model_sample_rate:  
    resampler = torchaudio.transforms.Resample(sample_rate, model_sample_rate)  
    waveform = resampler(waveform)  

# 提取特征  
inputs = processor(waveform, sampling_rate=model_sample_rate, return_tensors="pt")  
input_features = inputs.input_features  
input_features = input_features.to(device)  

# 生成  
generated_ids = model.generate(inputs=input_features, max_new_tokens=225)  # 贪婪解码  
# generated_ids = model.generate(inputs=input_features, max_new_tokens=225, num_beams=5)  # 束搜索  

# 反标记化  
generated_sentences = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]  

# 必要时对预测句子进行规范化处理