Comments (4)
您好,从这个issue 反馈的结果上看,relu2_output.npz 的结果一致,fc1_output.npz 存在 diff,这中间还经过了 pool2、flatten,可否再提供一下这两步的计算结果对比,以此确定是哪个算子的计算存在 diff
from paddle.
您好,从这个issue 反馈的结果上看,relu2_output.npz 的结果一致,fc1_output.npz 存在 diff,这中间还经过了 pool2、flatten,可否再提供一下这两步的计算结果对比,以此确定是哪个算子的计算存在 diff
已经更新
https://github.com/PhyllisJi/MoCoDiff_Bug/tree/paddle-issue%2364537
from paddle.
看日志,是fc1_output结果有 diff,这个就是一个简单的 linear,日志显示输入是 flatten 的输出,都是0,Linear bias 都是空的,理论上计算输出也应该是0吧?
from paddle.
看日志,是fc1_output结果有 diff,这个就是一个简单的 linear,日志显示输入是 flatten 的输出,都是0,Linear bias 都是空的,理论上计算输出也应该是0吧?
我们这里的数据给的是和完全一致的pytorch版本的代码的层输出的差异,如果是0,表示完全一致或者非常接近。现在的情况是在CUDA 11.6上面和pytorch的输出基本上完全一致,但是在CUDA 11.8上的运行结果和pytorch的结果有较大的差异,跳过pytorch,可以看出来直接对比两个CUDA版本上的输出,也是会有巨大差异的。
from paddle.
Related Issues (20)
- cudaErrorInitializationError HOT 6
- nn.Conv2D没有对kernel_size进行检查 HOT 1
- 同一份训练模型的代码在使用CPU和GPU运行时梯度存在明显差异
- 对于形状相同的输入,nn.MaxPool2D的输出形状与PyTorch有差异 HOT 1
- cuda是11.8,但自检程序报错说是10.2,导致cudnn不对,进而导致不能运行。找不到合适的paddle版本的离线下载方式。 HOT 6
- Paddle compile error when -DWITH_CINN=ON on Hygon DCU HOT 3
- 如何训练分类数据模型 HOT 1
- aarch64架构,cmake正常,make报错 HOT 1
- ValueError: (InvalidArgument) 'begin_norm_axis' in Op(LayerNorm) should begreater than zero. But received [0].
- YOLOv5训练自定义SAR舰船尾迹数据集时评估阶段报错:IndexError: list index out of range HOT 3
- paddle inference如何在XPU下推理量化模型 HOT 11
- Conv2DTranspose lacks upper level checks on groups and out_channels constraints
- paddle.nn.Conv2DTranspose代码中规定当group==inchannels时dilations只能为1,但是文档中没有写明
- paddle.nn.Conv2D does not check for constraints between Kernel size and input size
- Conv2D/ConvTranspose2d lacks checking of parameter values for exceptions HOT 2
- nn.Conv2DTranspose does not check for parameter dilation causing crash
- paddle.ones生成的Tensor数据全部是0 HOT 1
- paddlepaddle新版本在cpu下调用出现异常 HOT 2
- 无法配置Paddle多卡环境 HOT 27
- PaddleOCR能正常训练但无法评估 HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from paddle.