Skip to content

llama : fix K-shift with quantized K (wip) #5653

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

Closed
wants to merge 1 commit into from
Closed

llama : fix K-shift with quantized K (wip) #5653

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
31 changes: 22 additions & 9 deletions llama.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4671,17 +4671,30 @@ static void llm_build_k_shift(
}

for (int il = 0; il < n_layer; ++il) {
struct ggml_tensor * tmp =
// we rotate only the first n_rot dimensions
ggml_rope_custom_inplace(ctx,
ggml_view_3d(ctx, kv.k_l[il],
n_embd_head_k, n_head_kv, n_ctx,
ggml_row_size(kv.k_l[il]->type, n_embd_head_k),
ggml_row_size(kv.k_l[il]->type, n_embd_k_gqa),
0),
struct ggml_tensor * k = ggml_view_3d(ctx, kv.k_l[il],
n_embd_head_k, n_head_kv, n_ctx,
ggml_row_size(kv.k_l[il]->type, n_embd_head_k),
ggml_row_size(kv.k_l[il]->type, n_embd_k_gqa),
0);

struct ggml_tensor * tmp;
if (ggml_is_quantized(k->type)) {
// dequantize to f32 -> RoPE -> quantize back
tmp = ggml_cast(ctx, k, GGML_TYPE_F32);
cb(tmp, "K_f32", il);
tmp = ggml_rope_custom_inplace(ctx, tmp,
K_shift, n_rot, rope_type, 0, n_orig_ctx, freq_base, freq_scale,
ext_factor, attn_factor, beta_fast, beta_slow);
cb(tmp, "K_shifted", il);
cb(tmp, "K_shifted_f32", il);
tmp = ggml_cpy(ctx, tmp, k);
cb(tmp, "K_shifted_q", il);
} else {
// we rotate only the first n_rot dimensions
tmp = ggml_rope_custom_inplace(ctx, k,
K_shift, n_rot, rope_type, 0, n_orig_ctx, freq_base, freq_scale,
ext_factor, attn_factor, beta_fast, beta_slow);
cb(tmp, "K_shifted", il);
}
ggml_build_forward_expand(graph, tmp);
}
}
Expand Down