-
您好，我正在把lightGlue模型移植到RK3588板端，目前使用rknnlite2的Python脚本在板端推理符合预期。但是使用c++代码推理同样的模型时，直接得到的输出结果与rknnlite2的Python脚本结果差异较大。
以下是我的详细操作：
1、从onnx转换成rknn模型后，输入输出维度如下所示：有四个输入，一个输出。
转换rknn模型时，设置了QUANTIZE_ON = …
-
```python
#!/usr/bin/env python
# coding: utf-8
from rknn.api import RKNN
from sys import exit
rknn = RKNN(verbose=True)
ONNX_MODEL="RWKV-x060-World-1B6-v2.1-20240328-ctx4096.onnx"
RKNN_MOD…
-
原模型输出结构：
onnx_edit后结构:
ONNX_MODEL="RWKV-x060-World-1B6-v2.1-20240328-ctx4096.onnx"
R…
-
im trying convert model (cnn+lstm) from onnx to rknn for rk3588
My code:
shape = (1, 7, 3, 608, 184)
import numpy as np
img_means = (np.array((0.19007764876619865, 0.15170388157131237, 0.106594…
-
多尺度的静态输入推理结果正常，代码如下：
```
if scale==0.25:
ret = rknn.load_rknn('./models/dense_64.rknn')
elif scale==0.5:
ret = rknn.load_rknn('./models/dense_128.rknn')
else:
ret = rknn.load_rkn…
-
Hello,
I would like to request the addition of **YOLOv11** model support in the RKNN Model Zoo. It would be great if you could provide a process or script for converting YOLOv11 models (either from…
-
I'm trying to compile the model yolov8 small for RK3588
[model_on_google_disk](https://drive.google.com/file/d/1CRYK9Gmj9sQghd36WcUpidEcTi5zohPV/view?usp=drive_link)
`rknn-toolkit2 2.2.…
-
The only tested computer vision models (to my knowledge) are YOLOv5 models. There seems to be (untested?) support for YOLOv8 in the rknn-jni.
The YOLOv10 repo has some compelling graphs, and there…
-
Extending https://github.com/rockchip-linux/rknn-toolkit2/issues/222
Here's the Proof of Concept (run all cells) - https://colab.research.google.com/drive/1X5KeI0TVZui0teakQantH8CW5OWYA8mx?usp=sha…