From 5aa513e33c1641690f59d6444ba2ca779948c9d1 Mon Sep 17 00:00:00 2001 From: Steffy-zxf <48793257+Steffy-zxf@users.noreply.github.com> Date: Mon, 18 May 2020 10:15:04 +0800 Subject: [PATCH] Update release.md --- docs/release.md | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/docs/release.md b/docs/release.md index 18491918..8401c93f 100644 --- a/docs/release.md +++ b/docs/release.md @@ -1,5 +1,28 @@ # 更新历史 +## `v1.6.2` + +* 修复图像分类在windows下运行错误 + +## `v1.6.1` + +* 修复windows下安装PaddleHub缺失config.json文件 + +## `v1.6.0` + +* NLP Module全面升级,提升应用性和灵活性 + * lac、senta系列(bow、cnn、bilstm、gru、lstm)、simnet_bow、porn_detection系列(cnn、gru、lstm)升级高性能预测,性能提升高达50% + * ERNIE、BERT、RoBERTa等Transformer类语义模型新增获取预训练embedding接口get_embedding,方便接入下游任务,提升应用性 + * 新增RoBERTa通过模型结构压缩得到的3层Transformer模型[rbt3](https://www.paddlepaddle.org.cn/hubdetail?name=rbt3&en_category=SemanticModel)、[rbtl3](https://www.paddlepaddle.org.cn/hubdetail?name=rbtl3&en_category=SemanticModel) + +* Task predict接口增加高性能预测模式accelerate_mode,性能提升高达90% + +* PaddleHub Module创建流程开放,支持Fine-tune模型转化,全面提升应用性和灵活性 + * [预训练模型转化为PaddleHub Module教程](https://github.com/PaddlePaddle/PaddleHub/blob/release/v1.6/docs/contribution/contri_pretrained_model.md) + * [Fine-tune模型转化为PaddleHub Module教程](https://github.com/PaddlePaddle/PaddleHub/blob/release/v1.6/docs/tutorial/finetuned_model_to_module.md) + +* [PaddleHub Serving](https://github.com/PaddlePaddle/PaddleHub/blob/release/v1.6/docs/tutorial/serving.md)优化启动方式,支持更加灵活的参数配置 + ## `v1.5.2` * 优化pyramidbox_lite_server_mask、pyramidbox_lite_mobile_mask模型的服务化部署性能 -- GitLab