Digital2Slave

1. 由于**U版的YoloV5**，有多个不同的版本`yolov5s, yolov5m, yolov5s, yolov5l, yolov5x, yolov5n`等，可以将现有的`yolov5s`接口调整为统一调用接口。比如 **[EfficientNet-PyTorch](https://github.com/lukemelas/EfficientNet-PyTorch)** 2. 另外，项目[README.md](https://github.com/Tencent/ncnn/blob/master/README.md)文件中提到了支持`YoloV7`，也同时新增一个`yolov7`统一调用接口。 3. 针对**U版的YoloV5**，参考[yolov5.py](https://github.com/Tencent/ncnn/blob/master/python/ncnn/model_zoo/yolov5.py)新增一个针对`yolov5l`的类。 - `import torch` 为了处理单个类别的检测。 - 类初始化新增`param_path`，`model_path`，`class_names`。 - 修改模型加载代码，`self.net.load_param(param_path)` `self.net.load_model(model_path)` - 通过查看[Netron](https://netron.app/)查看`*.param`文件对应的模型结构，修改了模型输出处理代码。 ``` ret1, mat_out1 = ex.extract("output") ret2, mat_out2...

how to release model data after work done?

2

Hi sh1r0! I modified the demo to meet the needs of my own app. The Caffe model I used is about 400+MB, but the sum memory of android phone is...

How to convert yolov5s-seg.pt to ncnn?

2

0. environment ``` Ubuntu 18.04.6 LTS torch 1.8.0+cpu torchaudio 0.8.0 torchvision 0.9.0+cpu cmake 3.25.0 ninja 1.11.1 gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0 ``` Build `pnnx` refer https://zhuanlan.zhihu.com/p/444022507. 2. **yolov5s-seg.pt** https://github.com/ultralytics/yolov5/releases/download/v7.0/yolov5s-seg.pt 3. **export.py**...

add CMakeLists.txt for ncnn-yolov8s-seg and fix default parameter for yolov8.load_param yolov8.load_model.

ik_smart分析器如何做数学公式搜索？

1

es索引类型映射的时候，对title field用来ik_smart分析器。怎么实现title field的数学公式搜索呢，例如：Latex格式的公式，我想把两个$符直接的数学公式不进行拆分~ Anyone could you help out? Thanks very much.