speaker diarization precision 2
简介
核心亮点
- 精准识别多发言人切换时间点
- 显著提升会议转写文本的角色标注准确度
- 开源 Apache-2.0 协议，部署灵活且无商业限制
- 完美适配 Whisper 等 ASR 工具构建完整对话流
使用方法
# 安装 Hugging Face transformers
pip install transformers torch# 使用 transformers 加载模型
from transformers import AutoModel, AutoTokenizer
model = AutoModel.from_pretrained("pyannote/speaker-diarization-precision-2")
tokenizer = AutoTokenizer.from_pretrained("pyannote/speaker-diarization-precision-2")Hugging Face 下载
我们推荐使用命令行或者 Hugging Face Hub SDK 来进行模型的下载。
操作指引：在下载前，请先通过如下命令安装 huggingface_hub：
pip install -U huggingface_hub命令行下载
下载完整模型库
huggingface-cli download pyannote/speaker-diarization-precision-2下载单个文件到指定本地文件夹（以下载 config.json 到当前路径下 ./dir 目录为例）
huggingface-cli download pyannote/speaker-diarization-precision-2 config.json --local-dir ./dirSDK 下载
# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('pyannote/speaker-diarization-precision-2')Git 下载
请确保 lfs 已经被正确安装
git lfs install
git clone https://huggingface.co/pyannote/speaker-diarization-precision-2如果您希望跳过 lfs 大文件下载，可以使用如下命令
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/pyannote/speaker-diarization-precision-2模型文件托管在 Hugging Face Hub，使用 HF CLI / SDK / Git 直接下载，不经过本站。
PyTorch / Transformers 使用
安装 Transformers
pip install -U transformers torch模型加载和推理
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained('pyannote/speaker-diarization-precision-2')
tokenizer = AutoTokenizer.from_pretrained('pyannote/speaker-diarization-precision-2')模型下载
我们推荐使用命令行或者 ModelScope SDK 来进行模型的下载。
操作指引：在下载前，请先通过如下命令安装 ModelScope：
pip install modelscope命令行下载
下载完整模型库
modelscope download --model pyannote/speaker-diarization-precision-2下载单个文件到指定本地文件夹（以下载 README.md 到当前路径下 dir 目录为例）
modelscope download --model pyannote/speaker-diarization-precision-2 README.md --local_dir ./dirSDK 下载
# 模型下载
from modelscope import snapshot_download
model_dir = snapshot_download('pyannote/speaker-diarization-precision-2')Git 下载
请确保 lfs 已经被正确安装
git lfs install
git clone https://www.modelscope.cn/pyannote/speaker-diarization-precision-2.git如果您希望跳过 lfs 大文件下载，可以使用如下命令
GIT_LFS_SKIP_SMUDGE=1 git clone https://www.modelscope.cn/pyannote/speaker-diarization-precision-2.gitModelScope 模型页直接下载模型文件；无需将模型文件放在本站服务器。
Notebook 快速开发
下载并安装 ModelScope library
pip install "modelscope[audio,cv,nlp,multi-modal,science]" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html模型加载和推理
from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks
p = pipeline('text-generation', 'pyannote/speaker-diarization-precision-2')完整文档
---
tags:
- pyannote
- pyannote-audio
- pyannote-audio-pipeline
- audio
- voice
- speech
- speaker
- speaker-diarization
- speaker-change-detection
- voice-activity-detection
- overlapped-speech-detection
---
Precision-2 speaker diarization
This pipeline runs Precision-2 speaker diarization on pyannoteAI cloud.
Read the announcement blog post.
This pipeline is a stripped down version of pyannoteAI SDK that provides much more features:
- speaker diarization optimized for speech-to-text
- speaker voiceprinting and identification
- confidence scores
- and more...
A self-hosted version of Precision-2 is also available for enterprise customers.
Setup
1. pip install pyannote.audio
2. Create an API key on pyannoteAI dashboard (free credits included)
Usage
# initialize speaker diarization pipeline
from pyannote.audio import Pipeline
pipeline = Pipeline.from_pretrained(
'pyannote/speaker-diarization-precision-2',
token="{pyannoteAI-api-key}")
run speaker diarization on pyannoteAI cloud
output = pipeline("/path/to/audio.wav")
enjoy state-of-the-art speaker diarization
for turn, speaker in output.speaker_diarization:
print(f"start={turn.start:.1f}s stop={turn.end:.1f}s speaker_{speaker}")