Description:
paddlegpuserving 0.8.3
paddle-gpu-serving
paddle-gpu-serving简介是基于Paddle Serving框架的模型预测服务库，通过少量代码即可部署和使用指定的模型进行远程预测。目前支持使用PaddleHub语义理解模型库中的BERT类模型获取文本对应的向量表示。其中serving服务的代码参考bert模型服务demo server端。
安装
server端
环境要求：python3，paddlepaddle>=1.6，paddlehub>=1.4
pip install paddle-gpu-serving
client端
环境要求：ujson，python2或python3
下载bert_service.py脚本使用
使用
server端
from paddle_gpu_serving.run import BertServer
bs = BertServer(with_gpu=True)
bs.with_model('bert_chinese_L-12_H-768_A-12')
bs.run(gpu_index = 0, port = 8010)
client端
bc = BertService(
model_name='bert_chinese_L-12_H-768_A-12',
max_seq_len=20,
show_ids=False,
do_lower_case=True)
bc.add_server('127.0.0.1:8010')
result = bc.encode([["远上寒山石径斜"], ])
print(result[0])
License
For personal and professional use. You cannot resell or redistribute these repositories in their original state.
Customer Reviews
There are no reviews.
