ReLU forward and backward implemented in train_gpt2_fp32.cu#1
Open
Hrancheng wants to merge 2 commits intoReaLLMASIC:masterfrom
Open
ReLU forward and backward implemented in train_gpt2_fp32.cu#1Hrancheng wants to merge 2 commits intoReaLLMASIC:masterfrom
Hrancheng wants to merge 2 commits intoReaLLMASIC:masterfrom
Conversation
gkielian
reviewed
Oct 15, 2024
| } | ||
|
|
||
| //ReLU forward kernel | ||
| __global__ void relu_forward_kernel(float* out, const float* inp, int N, int T) { |
Collaborator
There was a problem hiding this comment.
I think we need to add the scale (temperature) as well into the kernel
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add Relu forward and backward kernels and it's fitting into model with validation loss decreasing
Usage:
make train_gpt2fp32cu
Enable relu training:
./train_gpt2fp32cu -r 1