airockchip/rknn-toolkit2

Issues

• Cannot load RKNN in jupyter-notebook

  #160 opened 2 months ago by gerth2
  1

• 在运行exSoftmax13算子推理的时候出现failed to submit!的报错。

  #199 opened a month ago by goudaqiang
  8

• onnx转换双目立体深度预测模型总是报错

  #202 opened a month ago by jieli1990
  1

• arm64 librknnrt.so(version 2.2), matmul result all zero

  #154 opened 2 months ago by ishine
  2

• toolkit2.2转换yolov8s/v11s onnx 模型时报错"E build: Unsupport file #!"

  #201 opened a month ago by rainote2020
  3

• rknn_run接口推理时cpu占用很高，不管用通用API接口还是零拷贝接口，占用都很高，是什么原因呢

  #174 opened a month ago by hhd-shuai
  2

• onnx转rknn遇到opset版本问题，rknn-toolkit2 v2.2.0要求opset<=15

  #157 opened 2 months ago by EvW1998
  7

• 使用rknnlite2在板端推理lightGLue结果正常，使用c++代码推理时结果差异较大

  #200 opened a month ago by liuqinglong110
  7

• E RKNN: [16:50:26.190] invalid tensor malloc size, tensor name: , target: CPU, size: 0 段错误 (核心已转储)

  #193 opened a month ago by MiningIrving
  4

• 多个input时，运行到rknn_run()报错Segmentation fault

  #192 opened a month ago by liuqinglong110
  14

• 关于主流深度学习框架的转换，是否考虑Mindspore和PaddlePaddle？

  #189 opened a month ago by Suqier
  3

• 我在安装RKNN Toolkit2有问题

  #164 opened 2 months ago by abcdetab
  9

• rknn_yolov5_android_apk_demo modifying the flip parameter does not take effect

  #198 opened a month ago by Starkguava
  0

• rk3588s是否支持bf16?

  #197 opened a month ago by AAA-Fei
  1

• RKNN 量化， 非图像模型如何指定量化数据，比如音频数据， 或模型的输入就是一个多维向量

  #196 opened a month ago by walletiger
  2

• License to use for commercial applications

  #171 opened a month ago by cao-nv
  2

• rknntoolkit2使用mmse量化报错run: Can not find tensor value info for '30 26_gemm_conv'!

  #195 opened a month ago by blueWatermelonFri
  0

• How to inference video with RTSP with RKNN?

  #181 opened a month ago by Keele520
  0

• pruning用的是什么算法呢

  #194 opened a month ago by xiaowanzi-tju
  0

• 安装提示：is not a supported wheel on this platform.

  #191 opened a month ago by mrtian
  4

• 连板推理精度太差，模拟运行一切正常

  #190 opened a month ago by shaoyie
  4

• rknpu2/examples 无法编译为 Debug

  #188 opened a month ago by songziming
  0

• android armv8a librknnrt-2.2 layernorm结果不正确

  #186 opened a month ago by ishine
  2

• Where is the 32Bit rknn-toolkit2-lite for Linux (Ubuntu and BuildRoot) ?

  #187 opened a month ago by nassimus26
  0

• 模型转化问题，使用rknn-toolkit2:2.1.0工具进行导出，时找不到onnx的输出头输出头

  #184 opened a month ago by suse-chen
  2

• DETR模型fp16精度正常，量化精度几乎为0

  #176 opened a month ago by blueWatermelonFri
  5

• 能否提供一份适用于RK3588的librknnrt.a静态链接文件

  #185 opened a month ago by Suqier
  0

• 连板测试精度发现有算子的输出值异常

  #183 opened a month ago by lwcode
  9

• RWKV-6 模型转换报错 `The input tensor cannot be reshaped to the requested shape. Input shape:{1,2048,1,1}, requested shape:{1,32,64,64}` (rknn-toolkit2 v2.2.0)

  #169 opened 2 months ago by happyme531
  2

• 使用onnx_edit插入新维度后得到的模型无法转换

  #170 opened 2 months ago by happyme531
  2

• 模拟器精度正常, 连板精度异常如何解决?

  #182 opened a month ago by lwcode
  2

• rknn.accuracy_analysis连板推理报错

  #180 opened a month ago by sangsasi
  1

• 板子的rknn_server需要和rknn-toolkit2一致吗？哪里有1.5.2的rknn_server？

  #175 opened a month ago by sangsasi
  1

• InputOperator, OutputOperator

  #179 opened a month ago by hhd-shuai
  4

• onnx转rknn失败

  #177 opened a month ago by ishine
  2

• TFLite conversion failed

  #178 opened a month ago by nassimus26
  0

• RK3588 seem not support 3-dim input like (n,c,h*w)

  #173 opened a month ago by lsq117
  2

• 量化报错ValueError: Unsupport file -0.886949! 不做量化可以完成编译

  #172 opened 2 months ago by cwr250
  2

• Yolo v8 compile error

  #163 opened 2 months ago by St-Viacheslavik
  7

• RV1103板端能用rknn-toolkit-lite2吗

  #166 opened 2 months ago by hzr12
  1

• RWKV-6 模型转换报错 `invalid tensor malloc size, tensor name: , target: CPU, size: 0` (rknn-toolkit2 v2.1.0)

  #168 opened 2 months ago by happyme531
  1

• RK3588一个NPU的核可以同时跑多个yolov5模型么？

  #155 opened 2 months ago by sunian6183
  2

• disabling warnings and info to stdout

  #167 opened 2 months ago by lzace817
  1

• RV1103用哪个呀

  #153 opened 2 months ago by hzr12
  1

• Too slow when using RoiAlign

  #165 opened 2 months ago by ph0ngg
  0

• onnx模型转rknn模型后尺寸异常

  #162 opened 2 months ago by 770768405
  12

• 不支持float32输入?Is float32 not supported as an input?

  #158 opened 2 months ago by sbssh
  1

• 环境配置报错

  #156 opened 2 months ago by WN-Wolf
  2

• 连板精度分析报错 exportDataSize large then model size

  #161 opened 2 months ago by happyme531
  2

• TopK算子输出结果偏差大

  #159 opened 2 months ago by happyme531
  1