thundersvm

thundersvm copied to clipboard

thundersvm vs sklearn-svm（python）

1

Why is my single dataset thundersvm not as fast as sklearn-svm

yuhangzhu4

Compatibility with cmake version >= 3.10 which deprecated FindCUDA module

darikoneil

The optimality indicator $f_i$ of the instance $x_i$ is updated to $f^{'}_ i$ using the following formula: $f_i^{'}=f_{i} + (\alpha_{u}^{'} - \alpha_{u}) y_{u} K(x_u,x_i) + (\alpha_l^{'} - \alpha_l ) y_{l}...

dongzhiming

Memory limit in prediction?

2

Hello, I wonder if there is a parameter about **max memory set for predict** when using jupyter notebook like this: `x1_test=X_test.drop('Pixel',inplace=False,axis=1)` `predictions1 = clf_svm1.predict(x1_test)` When I run the above prediction,...

Song-Yuqi

My original program is to generate different data sets in a loop, and each data set does svm.(matlab+libsvm+cpu) If I change to python+thundersvm+GPU,Will there be an increase in speed?

yuhangzhu4

问题咨询

2

您好，我最近正在尝试使用ThunderSVM，现在我的程序可以运行，但是现在的程序运行较慢，我排查原因发现一方面我的数据集比较大，在Excel表格有10万行左右，另一方面在程序运行时我的GPU内存利用率比较低，我现在使用的显卡是GTX1650。因此我想询问一下ThunderSVM是否有方法或者相关参数让GPU内存占用率提高，让程序运行变快呢？

liqu12138

GPU info :NVIDIA-SMI 535.171.04 Driver Version: 535.171.04 CUDA Version: 12.2 NVIDIA GeForce RTX 3090 ```text (base) root-wang@yao:~/machine-learn/frequenceFinger$ /usr/bin/env /opt/miniconda/envs/rw_torch/bin/python /home/root-wang/.vscode-server/extensions/ms-python.debugpy-2024.6.0-linux-x64/bundled/libs/debugpy/adapter/../../debugpy/launcher 51551 -- /home/root-wang/machine-learn/frequenceFinger/model/ML/svm.py 2024-05-13 18:02:48,635 INFO [default] #instances = 150,...

root-wang

Ubuntu, Linker error, CPU-only

6

Hi. Im getting an linker error while building thundersvm for working on CPU only. Im doing ``` git clone https://github.com/Xtra-Computing/thundersvm.git cd thundersvm git submodule init eigen && git submodule update...

ptrajdos

poly kernal of thundersvm

1

Can thundersvm use the "poly" kernal ? When i use the poly kernal just like in sklearn-svm, it dispalyed: "The kernel parameter not recognized, please refer to the document."

YYANGZHOU1142221

environment: windows 10 python 3.8 cuda 11.6 thundersvm commit : smalll dataset train log: svm fit begin: (404, 13) (404,) 2024-04-06 20:30:50,603 INFO [default] #instances = 404, #features = 13...

liningbo