Skip to content
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

Jump to bottom

sync : llama.cpp (ggml_scale, ggml_row_size, ggml_mul_mat_set_prec) #662

Merged
merged 6 commits into from
Dec 22, 2023
Merged

sync : llama.cpp (ggml_scale, ggml_row_size, ggml_mul_mat_set_prec) #662

Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
caff3db
sync : llama.cpp (ggml_scale, ggml_row_size, ggml_mul_mat_set_prec)
ggerganov Dec 22, 2023
8624146
ggml : add comment about backward GGML_OP_DIAG_MASK_INF (#4203)
GermanAizek Dec 22, 2023
1a58054
llama : fix platforms without mmap (#4578)
slaren Dec 22, 2023
b9e5491
ggml-alloc : fix ggml_tallocr_is_own
ggerganov Dec 22, 2023
fec35d6
whisper : minor
ggerganov Dec 22, 2023
f7b92db
ggml : cuda jetson + arm quants warnings
ggerganov Dec 22, 2023
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
Prev Previous commit
Next Next commit
ggml : add comment about backward GGML_OP_DIAG_MASK_INF (#4203)
  • Loading branch information
@GermanAizek @ggerganov
GermanAizek authored and ggerganov committed Dec 22, 2023
commit 8624146bcf7b8da3ace96be13db1085d79cd63f6
2 changes: 2 additions & 0 deletions src/ggml.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15335,6 +15335,8 @@ static void ggml_compute_backward(struct ggml_context * ctx, struct ggml_tensor
const int n_past = ((int32_t *) tensor->op_params)[0];
src0->grad =
ggml_add_or_set(ctx, src0->grad,
/* ggml_diag_mask_inf_impl() shouldn't be here */
/* ref: https://github.com/ggerganov/llama.cpp/pull/4203#discussion_r1412377992 */
ggml_diag_mask_zero_impl(ctx, tensor->grad, n_past, false),
zero_table);
}
Expand Down