Can you tell me more about conv2d ( ggml_conv_2d )? #103
Replies: 1 comment 5 replies
-
|
From a lazy perspective, we can inherit third-party libraries for testing |
Beta Was this translation helpful? Give feedback.
-
|
At present, I have found a more feasible reference: https://github.com/Tencent/ncnn/blob/master/src/layer/x86/convolution_x86.cpp#L280 |
Beta Was this translation helpful? Give feedback.
-
|
@Cyberhan123 Thank you a lot!!, best research!.
I will try to implement this code to ggml_conv2d to test memory usage and speed |
Beta Was this translation helpful? Give feedback.
-
|
According to my research, the implementation process on the CPU needs to be completely customized instead of using the ggml implementation. |
Beta Was this translation helpful? Give feedback.
-
|
In CUDA, we are actually not in a hurry to implement it because it is more complicated. |
Beta Was this translation helpful? Give feedback.
-
|
No matter what, this is also my learning process, and I am willing to spend time studying these details with you. This is also part of our growth. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I am not a professional, and I would like to know whether there is room for optimization of ggml_conv_2d from an industry perspective, because it is too slow at 512x512.
I searched for relevant informationοΌI'll list themοΌ
ggml_conv_2d: ggml-org/ggml#564 (comment)
pytorch: pytorch/pytorch#49582
Beta Was this translation helpful? Give feedback.
All reactions