IA704
Programmation GPU pour l'apprentissage
Goran Frehse, ENSTA Paris
26/02/2024 - 11/03/2024: La première partie du cours est ici :
https://www-inf.telecom-sudparis.eu/COURS/IA704/IA704.html
Rappel :
TP noté : implémenter une multiplication de matrices sur GPU avec tiling en utilisant le modèle ici.
à envoyer par mail à goran.frehse@ensta-paris.fr avant le 29/03/2024 17h.
22/03/2024:
Course: Linear Classification
Lab: Implementing Essential Matrix Operations (pas noté)
For parallelizing sums (optional) : see slides 37ff on dot product here (simply omit the product part)
29/03/2024:
Course: Using fmatrix in Python
Lab: Linear Classification (à soumettre avant le 08/04/2024 à 17h, *TP noté!*)
compare with results on Tensorflow here.
05/04/2024:
Course : From linear classifcation to neural networks (link to Caltech CS 179)
Quizz Linear Classification (due before the end of the session)
08/04/2024:
Lab: Neural Networks (pas noté)
En cas de problème :
Erratum (13/03/2023): In file fmatrix.cu, please deactivate the following line (by adding // at the beginning):
24 : // fmatrix_assert(mat);
Erratum (13/03/2023): COLAB has changed the directory structure. If you get an error message
fatal error: helper_cuda.h: No such file or directory
then you need to update the include directory in the command line. Instead of
!nvcc -g -G -I /usr/local/cuda/samples/common/inc/ ... (broken)
you need to download the files with
!git clone https://github.com/NVIDIA/cuda-samples.git
and then use
!nvcc -g -G -I cuda-samples/Common/ ... (should work)