Skip to content

Fix non-parallel residual for GPT-NeoX #45

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
eyusupov wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Fix non-parallel residual for GPT-NeoX #45

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 5 additions & 5 deletions cformers/cpp/main.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2324,13 +2324,13 @@ bool gptneox_eval(
struct ggml_tensor * inpFF;

if (hparams.use_parallel_residual == 0) {
printf("use_parallel_residual == 0\n");
//printf("use_parallel_residual == 0\n");
// This takes the self-attention residual output as input to Feedforward
inpFF = ggml_add(ctx0, cur, inpL);
cur = ggml_add(ctx0, cur, inpL);

// post attention layer norm
{
inpFF = ggml_norm(ctx0, inpFF);
inpFF = ggml_norm(ctx0, cur);

// inpFF = input_layernorm_weight*inpFF + input_layernorm_bias
inpFF = ggml_add(ctx0,
Expand All @@ -2354,8 +2354,8 @@ bool gptneox_eval(
inpFF = ggml_add(ctx0, ggml_repeat(ctx0, model.layers[il].c_mlp_proj_b, inpFF), inpFF);
}

// inpL = inpFF + inpL
inpL = ggml_add(ctx0, inpFF, inpL);
// inpL = inpFF + cur
inpL = ggml_add(ctx0, inpFF, cur);

} else if (hparams.use_parallel_residual == 1) {
// printf("use_parallel_residual == 1\n");
Expand Down