Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / 666DZY666/micronet issues and pull requests
#113 - histogramobserver是否缺少了对min的统计
Issue -
State: closed - Opened by Lamkx 3 months ago
#112 - modify weight/activation
Issue -
State: open - Opened by PariaDarbani 12 months ago
#111 - 请求帮助,模型量化时是一定要重新手动创建一个量化的模型类脚本吗。
Issue -
State: open - Opened by Thzny about 1 year ago
#110 - zero_point 求法
Issue -
State: open - Opened by cqray1990 almost 2 years ago
#109 - wbwtab下的bn_fuse 设置问题
Issue -
State: open - Opened by cqray1990 almost 2 years ago
#108 - 模型压缩数据对比,没有指出是wbwtab 还是wqaq训练的?
Issue -
State: open - Opened by cqray1990 almost 2 years ago
#107 - Can you add an English version please?
Issue -
State: open - Opened by WJ-Huang almost 2 years ago
#106 - 修改成自己模型后遇到的问题
Issue -
State: open - Opened by maziyi123 over 2 years ago
#105 - 关于是否有必要添加输出量化的问题?
Issue -
State: open - Opened by Joejwu over 2 years ago
- 1 comment
#104 - WQ的max_val为什么是 (1 << self.bits) - 2)?
Issue -
State: open - Opened by xingyueye almost 3 years ago
#103 - 关于反卷积和BN层融合的疑惑
Issue -
State: open - Opened by xiaoluomi almost 3 years ago
- 1 comment
#102 - 关于后训练量化的疑惑,真真诚诚希望得到解答
Issue -
State: open - Opened by JoeYounger almost 3 years ago
- 1 comment
#101 - weight量化问题
Issue -
State: open - Opened by paranoidai almost 3 years ago
#100 - yolov5量化实验
Issue -
State: open - Opened by Joejwu almost 3 years ago
- 1 comment
#99 - 模型·大小对比
Issue -
State: open - Opened by dan123yi almost 3 years ago
- 8 comments
#98 - 采用迁移的dorefa方法的量化训练问题
Issue -
State: closed - Opened by CLC2019 about 3 years ago
#97 - nin_gc训练一段step后,loss变很大很大
Issue -
State: open - Opened by EdwardVincentMa about 3 years ago
- 2 comments
#96 - Dorefa训练的反向传播
Issue -
State: open - Opened by csorujian about 3 years ago
- 1 comment
#95 - 量化导入自己pth模型时候出现 KeyError: 'state_dict'
Issue -
State: open - Opened by maziyi234 about 3 years ago
- 1 comment
#94 - 请问压缩之后的网络可以直接拿来在pytorch上做inference嘛?
Issue -
State: open - Opened by z02070183 about 3 years ago
- 1 comment
#93 - nin_gc网络,在add_quant_op()函数中为什么没有通道拆分并transpose,再reshape?
Issue -
State: closed - Opened by EdwardVincentMa about 3 years ago
- 1 comment
#92 - 合并BN层操作
Issue -
State: open - Opened by muboyu1 about 3 years ago
- 1 comment
#91 - TensorRT demo 什么时候更新
Issue -
State: open - Opened by zbs881314 about 3 years ago
- 1 comment
#90 - AQ中的数据范围缩小问题
Issue -
State: open - Opened by xingyueye about 3 years ago
#89 - Concat量化问题
Issue -
State: open - Opened by xingyueye about 3 years ago
- 1 comment
#88 - 剪枝中的layer是只对设定的layer进行剪枝吗?不是整体吗?
Issue -
State: open - Opened by GeneralJing about 3 years ago
#87 - qat输出模型如何支持后续的部署工作?
Issue -
State: open - Opened by xingyueye about 3 years ago
- 2 comments
#86 - 请问pytorch qat中如何保存float模型呢
Issue -
State: closed - Opened by liuqunzhong about 3 years ago
#85 - Typo
Issue -
State: open - Opened by zzh132 over 3 years ago
#84 - 请问其中QAT和QAFT的区别是什么呢
Issue -
State: closed - Opened by wmkai over 3 years ago
- 2 comments
#83 - 怎么样正确将量化后模型转换成onnx格式,并用onnxruntime加载呢?
Issue -
State: open - Opened by dlml over 3 years ago
- 1 comment
#82 - torch round error
Issue -
State: open - Opened by Jzz24 over 3 years ago
- 1 comment
#81 - 请问可以支持rnn网络的压缩吗?比如LSTM或者GRU,谢谢!
Issue -
State: open - Opened by NewEricWang over 3 years ago
- 1 comment
#80 - Update quantize.py
Pull Request -
State: open - Opened by RiccardoRuggiero over 3 years ago
#78 - How to put the quantized model in .jit version?
Issue -
State: open - Opened by RiccardoRuggiero over 3 years ago
#77 - 请教关于剪枝的问题
Issue -
State: open - Opened by Kayterancer over 3 years ago
- 1 comment
#76 - local variable 'conv_child_temp' referenced before assignment
Issue -
State: closed - Opened by sunbin1993 over 3 years ago
- 1 comment
#75 - 想问一下wbwtab和wqaq的区别是什么?
Issue -
State: open - Opened by myzhuang over 3 years ago
- 2 comments
#74 - bn融合训练时就没有bn了吗相当于?
Issue -
State: open - Opened by ArtyZe over 3 years ago
- 1 comment
#72 - 模型量化的问题
Issue -
State: open - Opened by huangshenneng almost 4 years ago
- 1 comment
#71 - fix dorefa wqaq QuantConv2d forward
Pull Request -
State: closed - Opened by imonace almost 4 years ago
- 1 comment
#70 - 关于TensorRT
Issue -
State: open - Opened by zhangfeixiang222 almost 4 years ago
- 1 comment
#69 - 请问这个对显卡显存有要求么,我用的1660ti,6G显存,CUDA9.2,显示CUDA out of memory
Issue -
State: open - Opened by lxy21319 almost 4 years ago
- 1 comment
#68 - 关于修复ioa量化训练会出现nan的bug
Issue -
State: open - Opened by Racha1992 almost 4 years ago
- 2 comments
#67 - 已修复。
Issue -
State: open - Opened by yanfang12 almost 4 years ago
- 2 comments
#66 - yolov5s int8量化不出框?
Issue -
State: closed - Opened by Wulingtian almost 4 years ago
#65 - 将Conv2d全部替换为QuantizedConv2d后,是需要重新训练么?
Issue -
State: open - Opened by KiedaTamashi almost 4 years ago
- 2 comments
#64 - 使用多卡训练时的bug
Issue -
State: open - Opened by coderhss almost 4 years ago
- 4 comments
#63 - bn层单独量化和卷积与bn融合量化的问题
Issue -
State: open - Opened by WangQiangItachi almost 4 years ago
- 1 comment
#62 - dorefa 的梯度量化问题
Issue -
State: open - Opened by prio158 almost 4 years ago
- 2 comments
#61 - 如何进行定点化测试
Issue -
State: open - Opened by Baymax-chen almost 4 years ago
- 1 comment
#60 - bn_fold的量化代码问题
Issue -
State: closed - Opened by Racha1992 almost 4 years ago
- 2 comments
#59 - bn_fold的量化代码问题
Issue -
State: closed - Opened by Racha1992 almost 4 years ago
#58 - 请问为什么要在BN层中进行剪枝的相关操作呢
Issue -
State: open - Opened by weilinapple almost 4 years ago
- 1 comment
#57 - Cooperation Proposal: 666DZY666/model-compression & PaddlePaddle
Issue -
State: open - Opened by saxon-zh about 4 years ago
#56 - 如何压缩自己的模型?
Issue -
State: open - Opened by charles-str about 4 years ago
- 2 comments
#55 - resume无法加载的问题
Issue -
State: closed - Opened by zhangfeixiang222 about 4 years ago
#54 - 全精度训练只能达到88%
Issue -
State: closed - Opened by zhangfeixiang222 about 4 years ago
#53 - 关于resnet的add和inception的concatenation如何量化
Issue -
State: open - Opened by zhangfeixiang222 about 4 years ago
- 1 comment
#52 - 剪枝部分:请问如果我的模型结构较为复杂,分支较多那么我如何修改剪枝的代码来遍历模型获取每个weight呢?
Issue -
State: open - Opened by zhishao about 4 years ago
- 1 comment
#51 - 单独对bn层进行量化,不进行bn_fold的量化实施方案
Issue -
State: open - Opened by ghost about 4 years ago
- 1 comment
#50 - 关于BN_fold的问题
Issue -
State: closed - Opened by ghost about 4 years ago
#49 - 量化的IAO中用BN融合训练出现梯度爆炸
Issue -
State: open - Opened by yhl1010 about 4 years ago
- 12 comments
#48 - 稀疏训练(对BN层γ进行约束)的程式碼看不懂
Issue -
State: open - Opened by lunasdejavu about 4 years ago
- 1 comment
#47 - 关于预训练模型的问题
Issue -
State: open - Opened by Yaoxingtian about 4 years ago
- 1 comment
#46 - 移植到目标检测模型效果不佳
Issue -
State: open - Opened by lovepan1 over 4 years ago
- 2 comments
#45 - GFLOPs如何计算
Issue -
State: open - Opened by clttyou over 4 years ago
#44 - BnFoldConv question
Issue -
State: open - Opened by lovepan1 over 4 years ago
- 3 comments
#43 - 入门小白求问XNORnet
Issue -
State: open - Opened by b-bread over 4 years ago
- 2 comments
#42 - 可否将CONV3D,以及GRU量化一下?
Issue -
State: open - Opened by gogo03 over 4 years ago
- 1 comment
#41 - Pull
Pull Request -
State: closed - Opened by BBuf over 4 years ago
#40 - 关于WbWtAb中二值化实现的问题
Issue -
State: closed - Opened by LumingSun over 4 years ago
- 1 comment
#39 - 关于 移动端部署 量化模型
Issue -
State: open - Opened by NewCoderQ over 4 years ago
- 2 comments
#38 - 关于量化后模型的问题
Issue -
State: open - Opened by mjanddy over 4 years ago
- 7 comments
#37 - pruning方案存在问题
Issue -
State: open - Opened by euminds over 4 years ago
- 1 comment
#36 - 关于分组卷积结构剪枝的模型剪枝
Issue -
State: closed - Opened by Danee-wawawa over 4 years ago
#35 - IAO BN融合量化测试
Issue -
State: open - Opened by ghost over 4 years ago
- 5 comments
#34 - 缩放操作的位置
Issue -
State: open - Opened by ztianlin over 4 years ago
- 1 comment
#33 - 楼主造的8bit量化的轮子和Pytorch官方量化的API比较
Issue -
State: open - Opened by johnzhou1996 over 4 years ago
#32 - 分组卷积中变量含义,如何针对depthwise卷积进行剪枝?
Issue -
State: open - Opened by AIprogrammer over 4 years ago
#31 - 如何进行Dorefa的32位浮点型计算
Issue -
State: open - Opened by ghost over 4 years ago
- 1 comment
#30 - 二值量化中心化并截断部分发现一处不知是bug还是我没有理解
Issue -
State: open - Opened by onion-rain over 4 years ago
#29 - 可以Post-training quantization 训练后量化压缩吗?
Issue -
State: closed - Opened by tianhaoyue over 4 years ago
- 2 comments
#28 - 萌新刚接触,请问剪枝normal_regular_prune这个文件中的import util 在哪下载呀。万分感谢!!
Issue -
State: open - Opened by 739982423 over 4 years ago
- 1 comment
#27 - 量化流程细节的疑问?
Issue -
State: open - Opened by VisionZQ over 4 years ago
- 5 comments
#26 - 量化后模型参数并未变成指定位数
Issue -
State: open - Opened by lvsen94 over 4 years ago
- 4 comments
#25 - 如何获取压缩后的模型?以及压缩率的计算公式?
Issue -
State: open - Opened by atrimage over 4 years ago
- 1 comment
#24 - ModuleNotFoundError: No module named 'util_wt_bab'
Issue -
State: open - Opened by nce3xin over 4 years ago
- 6 comments
#23 - 想请教一下剪枝方面的问题
Issue -
State: open - Opened by onion-rain over 4 years ago
- 1 comment
#22 - 向大佬请教量化模型保存的问题
Issue -
State: open - Opened by Rishiber over 4 years ago
- 9 comments
#21 - 若可以的话大佬开个模型压缩的课程(付费)
Issue -
State: open - Opened by gongdalinux over 4 years ago
#20 - 大佬可以多在知乎上发表模型压缩优化的文章
Issue -
State: open - Opened by 205418367 over 4 years ago
- 3 comments
#19 - 这个工程能否用于目标检测模型的量化么
Issue -
State: open - Opened by Blue-Eagle-10 almost 5 years ago
- 7 comments
#18 - 你好,gc_prune.py里面对分组卷积后的model参数保存,可否提供下代码
Issue -
State: open - Opened by huangzicheng almost 5 years ago
- 3 comments
#17 - 请问一下要怎么量化linear层
Issue -
State: open - Opened by Guyume almost 5 years ago
- 1 comment
#14 - 复现后发现模型大小并没有减小
Issue -
State: open - Opened by chenxiansen-coder almost 5 years ago
- 8 comments
#12 - 关于量化后权值的精度
Issue -
State: open - Opened by jingdonglin almost 5 years ago
- 7 comments
#11 - hi, 参数:last|activation_mp|activation_nor|last_relu —— 模型结构调整标志位,这些参数都是什么意思呢?🙏
Issue -
State: open - Opened by HaoMyWorld almost 5 years ago
- 3 comments