FastDeploy

FastDeploy copied to clipboard

有计划支持Multi-object tracking PP-Tracking吗？

3

对应PaddleDetection：https://github.com/PaddlePaddle/PaddleDetection/tree/release/2.5/configs/mot

jerryandjune

[Backend] Enable TensorRT BatchedNMSDynamic_TRT plugin

1

### PR types(PR类型) TensorRT后端 ### Describe - 移除原有PaddleDetection模型部署的Trick逻辑，改为使用TensorRT BatchedNMSDynamic_TRT插件(EfficientNMS_TRT无法对齐所有PaddleDetection检测模型结果)

jiangjiajun

### PR types Model ### Describe - **[HeadPose]** Support new model [FSANet](https://github.com/omasaht/headpose-fsanet-pytorch) - Add new result struct ```HeadPoseResult```

wjj19950828

1.完善量化工具, 支持同一模型, 在QAT(train)和PTQ(eval)分别使用不同的图片预处理函数 2.打通PP_LiteSeg模型. 3.更新相关readme. 4.添加PP_LiteSeg量化模型部署example.

yunyaoXYY

可支持推理结果、绘制分开执行吗？？

1

fast deploy 可以支持推理和可视化结果分开吗；预测得到的DetectionResult 结果 ( result = model.predict(im.copy()) ) .可以支持序列化吗？ 然后在反序列化结果执行 vis_im = vision.vis_detection(im, result, score_threshold=0.5) 得到绘制检测对象的图片？ 使用pickle序列化会出现错误 TypeError: can't pickle fastdeploy.libs.fastdeploy_main.vision.DetectionResult objects

fanyuanxiang

[Model] Add text classification task for ernie-3.0

1

### PR types(PR类型) Model ### Describe - Add text classification example for ernie-3.0

joey12300

### PR types(PR类型) Benchmark ### Describe Add benchmark for ernie sequence classification. Statistics include: - Time cost of each stage, including tokenization, runtime and postprocessing - CPU memory cost -...

joey12300

doc error in benchmark/readme.md

1

[文档](https://github.com/PaddlePaddle/FastDeploy/blob/develop/benchmark/README.md) 提供的案例有小错误，漏写了`tgz`后缀 原文： ``` wget https://bj.bcebos.com/paddlehub/fastdeploy/MobileNetV1_x0_25_infer.tgz tar -xvf MobileNetV1_x0_25_infer ``` 应该为： ``` wget https://bj.bcebos.com/paddlehub/fastdeploy/MobileNetV1_x0_25_infer.tgz tar -xvf MobileNetV1_x0_25_infer.tgz ```

czr-gc

有支持PPYOLOE+

2

请问现在能支持PPYOLOE+吗

jo-dean

web_demo ,npm install 为啥找不到包

4


kkkkkkxser