rknn-toolkit
rknn-toolkit copied to clipboard
rockchip-linux
lprnet车牌识别模型（有过内部修改），在toolkit1.7.3环境下转换的模型，在3399pro上，每次推理出来结果都不相同，差别很大，在正负几千到几万范围内，但在toolkit1.4环境下转换的模型在3568上是正常的
Bumps [tensorflow](https://github.com/tensorflow/tensorflow) from 1.14.0 to 2.9.3. Release notes Sourced from tensorflow's releases. TensorFlow 2.9.3 Release 2.9.3 This release introduces several vulnerability fixes: Fixes an overflow in tf.keras.losses.poisson (CVE-2022-41887) Fixes a...
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot] avatar"){kind=avatar}
Bumps [tensorflow-gpu](https://github.com/tensorflow/tensorflow) from 1.14.0 to 2.9.3. Release notes Sourced from tensorflow-gpu's releases. TensorFlow 2.9.3 Release 2.9.3 This release introduces several vulnerability fixes: Fixes an overflow in tf.keras.losses.poisson (CVE-2022-41887) Fixes a...
python 3.8 并不支持 onnx=1.6.0 吧？ https://github.com/onnx/onnx/issues/2734 我对 METADATA 做了以下修改才安装通过 ``` Requires-Dist: onnx (>=1.6.0) ```
Hello everyone, Very appreciated of your work. In my device (armvl7 arch), I've installed rknn-toolkit-lite from your repo. I've inspected that the repo does not have rknn-toolkit .whl files for...
源码： ```python import numpy as np import cv2 from rknn.api import RKNN import torchvision.models as models import torch if __name__ == '__main__': model = './hybridnets.pt' input_size_list = [[3, 1280, 720]]...
模型是paddle2onnx 转过来的， 
E Calc node Conv : Conv_12 output shape fail E Catch exception when loading onnx model: testonnx.onnx! E Traceback (most recent call last): E File "rknn/base/RKNNlib/onnx_ir/onnx_numpy_backend/shape_inference.py", line 65, in rknn.base.RKNNlib.onnx_ir.onnx_numpy_backend.shape_inference.infer_shape...
请问用resnet50做backbone训练reid行人特征提取，当模型输出特征没有零值的时候转成rknn全部正常。当某个训练出的模型输出特征数据是下面结构时（特征有一半是0值，但是已经收敛，用来比对也很正常），转成rknn模型后结果会完全错误，现象是转成rknn模型后不同图片会出现同样输出。是因为什么原因？  测试了3种方式， 1： pytorch直接转rknn，不量化do_quantization=False 2: pytorch直接转rknn，不量化do_quantization=True, 300张图片量化 3：pytorch先转onnx，onnx输出结果与pytorch一致，比对相似度也一致。onnx转rknn后，模型输出结果错误，比对相似度全为0，不同图片输出特征一样
