huawei-noah/bolt

Issues

• C API ducument don't exsit

  #138 opened 10 months ago by Nicooo-Wang
  0

• binary conv arm中input bit-packing像素顺序问题

  #131 opened 2 years ago by yanghaojin
  6

• Can't convert model to int8 precision with post_training_quantization

  #137 opened 2 years ago by jerrydyc
  3

• 请问支持mips架构的板子吗

  #136 opened 2 years ago by Henryplay
  1

• 请问bolt有1.3.1的发布计划吗？

  #126 opened 2 years ago by chillingche
  2

• BNN 只支持conv算子么？

  #129 opened 2 years ago by LitLeo
  1

• Kotlin MultiPlatform Library

  #132 opened 2 years ago by Doomsdayrs
  1

• 声音克隆的demo可以参考一下吗？

  #133 opened 2 years ago by sunzhe09
  1

• Unable to compile jpeg on Windows

  #134 opened 2 years ago by PengYu-Shi
  1

• x2bolt转化工具报错

  #124 opened 2 years ago by kuzuodaoqingchen
  3

• I have a problem. Does bolt quantization support x86?

  #123 opened 2 years ago by avan-debug
  2

• x86_64编译报错

  #121 opened 2 years ago by muranran
  5

• version 1.2.1 and 1.3.0 issues

  #120 opened 2 years ago by xvyaward
  7

• TinyBert模型经过post_training_quantization进行INT8量化后，在Linux_X86-64平台推理报错

  #119 opened 2 years ago by zxzlogic
  4

• 是否支持BGEMM？

  #116 opened 2 years ago by ThisisBillhe
  3

• arm cpu dilated conv遇到nchw类型的输入会出错

  #115 opened 2 years ago by chillingche
  0

• 建议每个模型的gpu kernel binary保存为独立的so

  #104 opened 3 years ago by chillingche
  1

• 展开OCL kernel中的标量dot操作可以获得更高的GFLOPs

  #113 opened 3 years ago by chillingche
  2

• 如何设置运行时浮点精度为fp16

  #108 opened 3 years ago by likefxz2009
  4

• android平台编译失败

  #107 opened 3 years ago by likefxz2009
  6

• Elewise算子耗时较大

  #106 opened 3 years ago by ThisisBillhe
  2

• How to build on Raspberry?

  #68 opened 3 years ago by yuxianzhi
  0

• benchmark issue

  #90 opened 3 years ago by yy6163773
  8

• using netron to visualise bolt model

  #97 opened 3 years ago by yuxianzhi
  1

• Bolt doesn't support bilstm of tensorflow2.6? ?

  #99 opened 3 years ago by xuezhongfei2008
  1

• 在macos上交叉编译ios出现错误

  #100 opened 3 years ago by lixcli
  1

• 如何在mac m1平台上编译？

  #98 opened 3 years ago by lixcli
  9

• tensorflow to caffe model, use caffe to infer, caffe uses that version ???

  #96 opened 3 years ago by xuezhongfei2008
  1

• opencl的一个bug

  #91 opened 3 years ago by peyer
  8

• bolt build failed on windows, here's the error report

  #89 opened 3 years ago by tcwangbuaa
  3

• 支持TorchScript格式的模型做为输入吗 ?

  #59 opened 3 years ago by shoveller86
  3

• [ERROR] thread 0 file C:\Users\Blank\Desktop\bolt-master\model_tools\src\onnx\onnx_adaptee.h line 264 operator name:Less_8 type:Less not supported.

  #84 opened 3 years ago by tcwangbuaa
  2

• Does bolt pad not support dynamic input? ??

  #86 opened 3 years ago by xuezhongfei2008
  1

• model_tools/tools/tensorflow2caffe/tts/transform_fastspeech2.py Is the model source code open source?

  #85 opened 3 years ago by xuezhongfei2008
  1

• Raspberry Pi 4 compile error

  #83 opened 3 years ago by MyraBaba
  1

• onnx_adaptee.h line 2128 can not process operator name:Pad__82 type:Pad attributes

  #80 opened 3 years ago by xuezhongfei2008
  6

• TFLite not install success

  #82 opened 3 years ago by xuezhongfei2008
  0

• Android平台是否有考虑改成动态加载手机/system/vendor/lib(64)下的libOpenCL.so

  #81 opened 3 years ago by paulwgw88
  3

• bolt的未来

  #75 opened 3 years ago by bxkdstar123
  1

• batch inference

  #73 opened 3 years ago by lishaojun412
  1

• llvm-ranlib问题

  #72 opened 3 years ago by DekkerXu
  2

• STORE_OUTPUT存在typo，Adreno GPU计算反卷积f2s2出错

  #69 opened 3 years ago by chillingche
  1

• 树莓派上可以使用吗

  #65 opened 3 years ago by seacj
  4

• conv_direct_sh1_fn_spe.cl存在typo，导致GPU kernel编译失败

  #61 opened 3 years ago by chillingche
  1

• 使用1.2.1的X2bolt转换tflite失败（1.2.0没有问题）

  #63 opened 3 years ago by chillingche
  1

• bolt编译后能在visual studio导入使用吗？

  #58 opened 3 years ago by carter54
  6

• onnx squezee在opset13改了定义

  #56 opened 3 years ago by Bil17t
  1

• 模型功耗分析

  #54 opened 3 years ago by happy-ngh
  5

• 使用GPU算法选择文件加速模型初始化，存在corner case未被加速

  #55 opened 3 years ago by chillingche
  1

• 模型名中包含特殊字符时，build_preprocess_ocl.sh失败

  #53 opened 3 years ago by chillingche
  1