From 501c92444b3125eb1f34bb132bf3fc032a70e350 Mon Sep 17 00:00:00 2001 From: Mathis Date: Mon, 22 Apr 2024 14:32:25 +0000 Subject: [PATCH] Update content/blog/cuda_net.md --- content/blog/cuda_net.md | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/content/blog/cuda_net.md b/content/blog/cuda_net.md index a679c10..beeb0bc 100644 --- a/content/blog/cuda_net.md +++ b/content/blog/cuda_net.md @@ -3,6 +3,8 @@ title: Writing a Convolutional Neural Network library with CUDA Support draft: true --- +Straightforward project, learned a lot more than I expected. + "Just use cuBLAS, it'll be easier. You don't have to implement custom CUDA kernels.", they said. Actually, noone said that. I just thought that because I didn't do enough research. Why not combine multiple challenging things into 1 (C++, cmake, CUDA, CNN) @@ -18,4 +20,10 @@ Quickly discovering that without writing custom kernels, you can't really progre - naive conv2d - learning 3D memory representation - optimizing conv2d -- softmax sum reduce \ No newline at end of file +- softmax sum reduce +- softmax numerical stability - max reduce +- custom binary weights file - (safetensors - json parser vs csv) values overwritten by header +- tests passing -> implement AlexNet +- AlexNet cmake, opencv +- AlexNet crashing -> add cuda error checking to tests -> test crashing +- compute-sanitizer memecheck \ No newline at end of file