Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / OpenPPL/ppq issues and pull requests
#575 - 导出的QDQ ONNX模型转为TensorRT后,速度比FP16的ONNX还要慢?
Issue -
State: open - Opened by zhishao 12 days ago
- 1 comment
#574 - ppq支持对lstm算子的QAT或PTQ量化吗?
Issue -
State: open - Opened by 1778813838 16 days ago
#573 - 如何对不同的层设置不同的bit精度呀
Issue -
State: open - Opened by Zeudfish 18 days ago
#572 - snpe 作为后端支持per_channel 量化吗?
Issue -
State: open - Opened by bruce1408 20 days ago
- 1 comment
#571 - Use torch.amax in ReduceMax forward to support multiplie dim
Pull Request -
State: open - Opened by ice-tong about 1 month ago
#570 - How does PPQ perform real quantization and achieve speed up?
Issue -
State: open - Opened by YixuanSeanZhou about 1 month ago
#569 - PPQ量化后,不能导出engine(05_QuantizationAgain.py)
Issue -
State: open - Opened by koihoo about 2 months ago
#568 - PPQ can not process parameter variable(PPQ_Variable_108) with multiple destinations
Issue -
State: open - Opened by Xiaochaosui 2 months ago
- 3 comments
#567 - Quantization for tflite
Issue -
State: open - Opened by liamsun2019 3 months ago
- 1 comment
#566 - RuntimeError: Op Execution Error: /Slice(Type: Slice, Num of Input: 5, Num of Output: 1)
Issue -
State: open - Opened by yehu3d 3 months ago
#565 - ProgramEntrance_2.py读取imagenet图片reshape报错
Issue -
State: open - Opened by huanFigure 3 months ago
#564 - 求助 为什么量化以后模型没有减小?
Issue -
State: open - Opened by stricklandye 4 months ago
#563 - AttributeError: 'NoneType' object has no attribute 'copy'
Issue -
State: open - Opened by undcloud 4 months ago
#562 - PPQ在画质类任务上(如低光降噪)量化效果咋样呢?
Issue -
State: open - Opened by 666DZY666 4 months ago
#561 - PPQ已经支持了不少后端推理框架(TensorRT,ncnn,mnn等),这些后端的量化规格细节都是怎么获取的呢?现在PPQ与后端量化推理可以完全对齐吗?
Issue -
State: open - Opened by 666DZY666 4 months ago
#560 - VIT类模型量化感知训练loss跑飞
Issue -
State: open - Opened by XingchenSong1116 4 months ago
- 1 comment
#559 - 能否支持权重和激活位宽不同的量化?
Issue -
State: open - Opened by frankgt 4 months ago
- 2 comments
#558 - 这个工具链已经收敛了吗?我看已经比较久没有特性更新了,后续还有新特性计划吗?
Issue -
State: open - Opened by 666DZY666 4 months ago
- 1 comment
#557 - ppq使用torch_mlir导出tosa格式的mlir
Issue -
State: open - Opened by zccyman 5 months ago
- 1 comment
#556 - 是否有计划支持高通QNN(比较新的版本)和海思芯片呢?
Issue -
State: open - Opened by 666DZY666 5 months ago
- 1 comment
#555 - 关于QAT中的BN?
Issue -
State: open - Opened by 666DZY666 5 months ago
- 1 comment
#554 - Op Execution Error MMCVModulatedDeformConv2d
Issue -
State: closed - Opened by JXCab 5 months ago
#553 - PPQint8CPU量化bug
Issue -
State: open - Opened by 37ETs 5 months ago
- 1 comment
#552 - [视频资料]使用profiler来分析网络性能,tensorboard无法解析日志文件
Issue -
State: open - Opened by JXCab 5 months ago
- 2 comments
#551 - 请问PPQ支持一维网络模型的量化吗
Issue -
State: open - Opened by JiawenDeng 5 months ago
- 1 comment
#550 - yolov5量化,在ppq/samples/yolo中运行02_Quantization.py文件出现警告和报错,输出模型只有onnx没有engine文件
Issue -
State: open - Opened by zjhleaning 5 months ago
- 1 comment
#549 - Op Execution Error: Einsum
Issue -
State: open - Opened by heqqian 5 months ago
- 1 comment
#548 - 请问如何使用ppq量化模型的部分模块,而不是整个模型呢?
Issue -
State: closed - Opened by heqqian 5 months ago
#547 - Fix Resize and Gridsample ops
Pull Request -
State: closed - Opened by liwei109 6 months ago
#546 - 请问目前支持pytorch模型直接量化(而不是转成onnx),并生成量化后pytorch模型吗
Issue -
State: open - Opened by JiliangNi 6 months ago
- 1 comment
#545 - 请问为什么在运行 ppq/ppq/samples/TensorRT/Benchmark_with_onnx.py 文件时会出现Segmentation Fault的可能原因是什么?
Issue -
State: closed - Opened by wmWangmiao 6 months ago
- 1 comment
#544 - How to get a Qoperator format ONNX model after quantization?
Issue -
State: open - Opened by JiliangNi 6 months ago
- 1 comment
#543 - 执行quantize_torch_model.py后,pytorch模型export出的onnx模型并不是QNN模型
Issue -
State: open - Opened by JiliangNi 6 months ago
- 9 comments
#542 - BatchNormalization量化出错
Issue -
State: open - Opened by 18liumin 6 months ago
- 2 comments
#541 - 请问 是否可以通过量化后产生的onnx 推理结果可以直观看出量化的情况?
Issue -
State: open - Opened by fakerhbj 7 months ago
- 2 comments
#540 - 加减类算子类型匹配问题
Issue -
State: open - Opened by ycwnuoyan 7 months ago
- 3 comments
#539 - 关于导出平台的报错信息
Issue -
State: open - Opened by dpf0522 8 months ago
- 2 comments
#538 - 修正了resize op与Prelu op执行时可能出现的问题
Pull Request -
State: closed - Opened by ZhangZhiPku 8 months ago
#537 - 修正了一个 onnx domain 的相关问题
Pull Request -
State: closed - Opened by ZhangZhiPku 8 months ago
#536 - [fix] Remove restriction on calibration dataset length
Pull Request -
State: closed - Opened by liwei109 8 months ago
- 1 comment
#535 - yolov5_sample报错
Issue -
State: open - Opened by uidq2956 8 months ago
- 1 comment
#534 - 量化模型的时候出现TypeError
Issue -
State: open - Opened by Francis235 8 months ago
- 1 comment
#533 - Update README.md
Pull Request -
State: closed - Opened by LJoson 8 months ago
#532 - 关于resize op的一个bug
Issue -
State: open - Opened by Nusselder9 8 months ago
- 2 comments
#531 - Op Execution Error: PRelu_1
Issue -
State: open - Opened by qiankl 9 months ago
- 4 comments
#530 - 调度部分算子至FP32时,生成engine失败
Issue -
State: open - Opened by zhjw0927 9 months ago
- 2 comments
#529 - 请问model zoo的模型在哪里?
Issue -
State: open - Opened by YilanWang 9 months ago
- 1 comment
#528 - NotImplementedError: Graph op: /quantize/ArgMin(ArgMin) has no backend implementation
Issue -
State: open - Opened by YilanWang 9 months ago
- 2 comments
#527 - 量化时出现RandomNormalLike这个算子没有注册跟实现的问题,请问有没有简单的注册实现例子可以参考?
Issue -
State: open - Opened by Uwenisme 9 months ago
- 1 comment
#526 - QLinearMatmul : weight zero point must be a scalar, 1D tensor of size 1, or last to second dimension is 1
Issue -
State: open - Opened by fhahaha 9 months ago
#525 - yolov5 模型量化后的精度损失
Issue -
State: open - Opened by zhjw0927 10 months ago
- 2 comments
#524 - 内存泄露
Issue -
State: open - Opened by KompressorSC 10 months ago
- 3 comments
#523 - 能不能量化自己的模型啊?有教程吗
Issue -
State: open - Opened by couakua 10 months ago
- 1 comment
#522 - YOLOv6量化后仅能预测一类
Issue -
State: open - Opened by zhujiajian98 10 months ago
- 3 comments
#521 - yolov8量化后无结果
Issue -
State: open - Opened by kjgfcdb 10 months ago
- 3 comments
#520 - PPQ yolov5-7.0 :return torch.cat(concat_view, axis=axis) RuntimeError: Sizes of tensors must match except in dimension 1. Expected size 28 but got size 27 for tensor number 1 in the list.
Issue -
State: open - Opened by YellenYeason 10 months ago
- 1 comment
#519 - PPQ installtion error: `TypeError: Descriptors cannot be created directly. If this call came from a _pb2.py file, your generated code is out of date and must be regenerated with protoc >= 3.19.0.`
Issue -
State: open - Opened by kyrie2to11 10 months ago
- 5 comments
#518 - LSQ算法微调后,输出的scale量化参数存在负值,请问如何修改源码,使其输出的scale为正值?
Issue -
State: open - Opened by fengyanWang 10 months ago
- 1 comment
#517 - AttributeError: ORT_OOS_INT8
Issue -
State: closed - Opened by Rickustc 10 months ago
- 1 comment
#516 - 回归模型相关咨询
Issue -
State: open - Opened by yeasoon 11 months ago
- 2 comments
#515 - 如何将yolov8n-seg进行int8量化. onnx2trtint8
Issue -
State: closed - Opened by lucy3589 11 months ago
#514 - vision transformer量化模型
Issue -
State: open - Opened by hufangjian 11 months ago
- 1 comment
#513 - statistical_analyse分析时,会异常退出
Issue -
State: open - Opened by BlueSkyB 11 months ago
- 1 comment
#512 - [视频资料]使用profiler来分析网络性能失败
Issue -
State: closed - Opened by Deeperfinder 11 months ago
- 1 comment
#511 - trt onnxParser
Issue -
State: closed - Opened by eezhang123 11 months ago
- 2 comments
#510 - 存在与export_ppq_graph相对应的加载量化模型的接口吗?
Issue -
State: open - Opened by LaoYingBu 11 months ago
- 1 comment
#509 - 自定义量化配置的几个问题
Issue -
State: open - Opened by YueYiHuan 11 months ago
- 2 comments
#508 - PPQ是否支持量化clip模型?
Issue -
State: closed - Opened by shhn1 11 months ago
- 3 comments
#507 - [视频资料:网络性能分析] profiler无法看到GPU信息
Issue -
State: closed - Opened by Deeperfinder 11 months ago
- 4 comments
#506 - 伪量化中的bias处理
Issue -
State: closed - Opened by zhanglei1172 12 months ago
- 1 comment
#505 - 有没有本项目进行量化的中文的视频教程,谢谢大佬们
Issue -
State: closed - Opened by ZJDATY 12 months ago
#504 - DepthToSpace reference 实现问题
Issue -
State: open - Opened by zhiqwang almost 1 year ago
- 2 comments
#503 - 量化onnx模型时报错:cannot convert float NaN to integer
Issue -
State: open - Opened by Lu-junan about 1 year ago
- 2 comments
#502 - 量化策略如何选择?
Issue -
State: open - Opened by fengyanWang about 1 year ago
- 2 comments
#501 - 在读取pt模型为gragh时,会先将模型转换为onnx,再读取onnx为gragh,会对gragh进行格式化,其中会将constant转换为parameter,但若constant的value为None,导致转换为parameter后出现再次读取gragh时出现空值报错,该如何处理该问题
Issue -
State: open - Opened by KuaiQuXueX1 about 1 year ago
- 1 comment
#500 - 量化表有几层的scale值完全相等
Issue -
State: closed - Opened by Nusselder9 about 1 year ago
- 2 comments
#499 - 量化后模型导出为onnx时,强制在每个op前插入quantize/dequantize
Issue -
State: closed - Opened by SunCrazy about 1 year ago
- 2 comments
#498 - 想利用PPQ做大模型量化的一些疑问
Issue -
State: open - Opened by ehuaa about 1 year ago
- 3 comments
#497 - 关于QuantizationProperty代码的疑问
Issue -
State: closed - Opened by weihChen about 1 year ago
- 2 comments
#496 - RuntimeError: Op Execution Error: BatchNormalization_27(Type: Conv, Num of Input: 3, Num of Output: 1)
Issue -
State: closed - Opened by zyt1024 about 1 year ago
- 2 comments
#495 - 关于量化感知训练的问题
Issue -
State: open - Opened by zhaojinkui about 1 year ago
- 7 comments
#494 - RuntimeError: Op Execution Error
Issue -
State: open - Opened by Pedestrian-Z about 1 year ago
- 3 comments
#493 - yolov5和yolov8模型量化时错误
Issue -
State: closed - Opened by zranguai about 1 year ago
- 2 comments
#492 - 请问dispatching_table是如何影响trt build_engine时各层precision的?
Issue -
State: closed - Opened by Nusselder9 about 1 year ago
- 1 comment
#491 - 关于VIT量化
Issue -
State: closed - Opened by Nusselder9 about 1 year ago
- 2 comments
#487 - dispatch.py run error
Issue -
State: closed - Opened by KKWY0909 about 1 year ago
- 1 comment
#485 - [BUG]run ppq shell python ProgramEntrance_1.py quant openvino int8 model meet the problem
Issue -
State: closed - Opened by PancakeAwesome over 1 year ago
- 3 comments
#482 - QDQ+onnxruntime 和 导出后的 trt+int8 模型结果不能对齐
Issue -
State: closed - Opened by zt706 over 1 year ago
- 3 comments
#481 - 使用 onnxruntime 加速网络出错
Issue -
State: closed - Opened by lDarryll over 1 year ago
- 3 comments
#479 - 修正了一个反卷积与Batchnorm进行合并时可能遇到的问题
Pull Request -
State: closed - Opened by ZhangZhiPku over 1 year ago
#478 - ProgramEntrance_2.py 执行报错
Issue -
State: open - Opened by zt706 over 1 year ago
- 6 comments
#465 - ONNX如何在ARM架构的NVIDIA计算平台上生成engine?
Issue -
State: closed - Opened by zhjw0927 over 1 year ago
- 6 comments
#446 - mnn量化过程中报错
Issue -
State: open - Opened by CYYAI over 1 year ago
- 5 comments
#393 - 运行问题
Issue -
State: closed - Opened by zack-rgb over 1 year ago
- 5 comments
#374 - 如何导出QDQ Onnx(Tensorrt使用)?
Issue -
State: closed - Opened by Noobzm over 1 year ago
- 6 comments
#370 - 当前mnn应该不支持量化后的onnx模型的转换,或者输入量化校正json方式
Issue -
State: open - Opened by zjd1988 over 1 year ago
- 12 comments
#316 - RKNN平台量化使用怎么样选择量化配置?
Issue -
State: closed - Opened by masommao almost 2 years ago
- 3 comments
#275 - TypeError: quantize_onnx_model() missing 1 required positional argument: 'platform'
Issue -
State: closed - Opened by Jianchun-Huo almost 2 years ago
- 6 comments
#100 - 修正了一个tensorRT导出时 offset dtype 是fp32的问题,这个问题不影响导出
Pull Request -
State: closed - Opened by ZhangZhiPku over 2 years ago
#99 - 修复了一个tensorRT Exporter没有正确删掉device switcher的问题
Pull Request -
State: closed - Opened by ZhangZhiPku over 2 years ago