Tencent/ncnn

Issues

• a minor issue in prebuild ncnn-android libs

  #5466 opened 8 days ago by zhouwg
  0

• 龙芯教育派2k1000中运行报错浮点数例外

  #5465 opened 9 days ago by Aoison
  0

• EfficientPhys onnx转ncnn模型转换报错

  #5452 opened 11 days ago by 408550969
  3

• how do I get the fossilize file .foz out of the vulkan driver?

  #5463 opened 12 days ago by cmsanta
  1

• 转换模型报错，如何定位

  #5462 opened 13 days ago by XdpAreKid
  0

• 鲲鹏920环境，yolov8n模型int8量化速度比默认的fp16慢了50%

  #5443 opened 14 days ago by zhenjing
  5

• 是否能够单独使用矩阵乘法的API？

  #5460 opened 14 days ago by xaclincoln
  1

• I convert onnx to ncnn successfully, but all my inference is all nan. Eg, the output of net.extract() is all nan.

  #5442 opened a month ago by Suncheng2022
  15

• pnnx能正常转换模型，模型输出与onnx模型输出不一致

  #5451 opened 17 days ago by Oliverwang
  4

• 用自己的yolov8模型 转成ncnn 在windows下部署后，接口也对上了，但是结果却出不来，不尽人意

  #5453 opened 18 days ago by zeng215217
  0

• [pnnx]:torch.clamp_min convert failed

  #5429 opened 19 days ago by deepage
  2

• rnn(lstm，gru)，解卷积的量化以及增加weight only的量化

  #5423 opened 20 days ago by hugauss
  2

• [ncnn-android-yolov8] How to handle real-time detect when the view set orientation to "landscape" ?

  #5450 opened 20 days ago by chenchen-boop
  1

• 我有3个GPU，但get_gpu_count()=1

  #5440 opened a month ago by 1027663760
  8

• Can simplestl use in RTOS ?

  #5449 opened 21 days ago by Ryuk17
  0

• intrinsic code 没有体现算法的原本的设计，是否有计划升级intrinsic code，并设置vset为ta mu

  #5448 opened 21 days ago by ArthurLiu-jbl
  0

• yolov8训练完成的模型转成ncnn的模型后，推理不出结果，网上查了说要进行前后处理，确实不会，能发个cpp的例子看看么！

  #5420 opened 2 months ago by fengshangren
  4

• Bad performance for int8 inference on XuanTie 906 (RISC-V)

  #5447 opened 22 days ago by vassilijnadarajah
  1

• 在对mtcnn模型第二层进行./ncnn2table过程中出现了段错误

  #5445 opened a month ago by hkdddld
  0

• Android下的build，为什么默认关闭exception呢？

  #5431 opened a month ago by porpalone
  5

• yolov8n模型在鲲鹏ARM机器的检测结果和pytorch结果不一样

  #5439 opened a month ago by zhenjing
  6

• pnnx和ncnn输出不一致

  #5441 opened a month ago by cdliang11
  0

• 手动创建的net，推理慢了很多

  #5438 opened a month ago by w1005444804
  0

• 目前RISCV版本中用到的RVV intrinsic代码已经不是最新版本了，有升级riscv intrinsic代码的计划吗？

  #5430 opened a month ago by ArthurLiu-jbl
  0

• benchmark测试占用率低

  #5434 opened a month ago by jamesdod
  0

• iPhone创建ncnn::net崩溃

  #5433 opened a month ago by xiaozhi003
  2

• openmp冲突引起crash

  #5432 opened a month ago by ljdang
  0

• 用摄像头调用 yolov5 net.extractor时内存增长迅速

  #5415 opened 2 months ago by dc-cheny
  1

• 预编译库中，希望能增加macOS下的动态库版本供下载

  #5428 opened a month ago by maconel
  0

• 可以给生成的pnnx.py中的pytorch model 增加初始化权重默认入参么，方便迁移到其他地方使用

  #5425 opened 2 months ago by hmzjwhmzjw
  0

• 在64核机器上跑ncnn模型，运行时只会调用其中一个核，设置哪些参数能跑指定数量的cpu核心

  #5412 opened 2 months ago by mmdzzh
  1

• ncnn-20240410使用-DNCNN_BF16=OFF编译报错

  #5421 opened 2 months ago by xiaozhi003
  3

• 在Mac M3 芯片下编译ncnn 编译成功后 在优化的时候出现Segmentation fault (core dumped)

  #5401 opened 2 months ago by Elijah-Yi
  1

• MacOS下 new ncnn::Net() 崩溃

  #5417 opened 2 months ago by FyhSky
  2

• Tile 算子似乎没有起作用

  #5419 opened 2 months ago by roachsinai
  0

• mtcnn ncnn模型fp32转fp16为什么会段错误

  #5413 opened 2 months ago by hkdddld
  8

• 有人在鸿蒙系统上使用过ncnn吗，整体加速效果怎么样

  #5404 opened 2 months ago by zcl912
  1

• Missing Python API documentation

  #5408 opened 2 months ago by jakubdolejs
  0

• RVV Illegal Instruction

  #5400 opened 2 months ago by XavierZXY
  1

• pnnx 转 ncnn 得到的 reshape 算子没有参数

  #5394 opened 2 months ago by roachsinai
  3

• ch_PP-OCRv4_rec_infer 轉換問題

  #5395 opened 2 months ago by timlliu
  0

• arm端运行出现-nan的问题

  #5388 opened 2 months ago by sglshanghai
  1

• ncnn::Extractor::extract崩溃

  #5386 opened 2 months ago by xiaozhi003
  1

• pnnx使用出现std::out_of_range

  #5385 opened 2 months ago by Zekegao
  0

• extract遇到Segmentation fault

  #5383 opened 2 months ago by Zekegao
  1

• 编译ncnn时，protobuf要求

  #5384 opened 2 months ago by oym050922021
  0

• I can't get my code to make an inference on an image

  #5374 opened 2 months ago by Vicenbs
  1

• 在PNNX中新增Attention的融合逻辑失败

  #5379 opened 2 months ago by DamonsJ
  5

• extract的时候出现crash

  #5381 opened 2 months ago by ljdang
  0

• layer load_model LSTM failed

  #5378 opened 2 months ago by Haibit
  1