Skip to content

ggml: allow split-mode tensor to use different quantization types. #23225

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
RedToasty wants to merge 1 commit into from
+11 −7
Closed

ggml: allow split-mode tensor to use different quantization types. #23225

Show file tree
Hide file tree
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
6 changes: 6 additions & 0 deletions ggml/src/ggml-backend-meta.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -536,6 +536,12 @@ static struct ggml_backend_meta_split_state ggml_backend_meta_get_split_state(co
if (src_ss[0].axis == GGML_BACKEND_SPLIT_AXIS_MIRRORED && src_ss[1].axis == GGML_BACKEND_SPLIT_AXIS_MIRRORED) {
return {GGML_BACKEND_SPLIT_AXIS_MIRRORED, {0}, 1};
}
if (src_ss[0].axis == GGML_BACKEND_SPLIT_AXIS_MIRRORED &&
(src_ss[1].axis == GGML_BACKEND_SPLIT_AXIS_2 || src_ss[1].axis == GGML_BACKEND_SPLIT_AXIS_3)) {
ggml_backend_meta_split_state ret = src_ss[1];
ret.n_segments = 1;
return ret;
}
if (src_ss[0].axis == GGML_BACKEND_SPLIT_AXIS_1 && src_ss[1].axis == GGML_BACKEND_SPLIT_AXIS_MIRRORED) {
ggml_backend_meta_split_state ret = src_ss[0];
ret.axis = GGML_BACKEND_SPLIT_AXIS_0;
Expand Down
4 changes: 0 additions & 4 deletions src/llama-context.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3366,10 +3366,6 @@ llama_context * llama_init_from_model(
LLAMA_LOG_ERROR("%s: SPLIT_MODE_TENSOR requires flash_attn to be enabled\n", __func__);
return nullptr;
}
if (ggml_is_quantized(params.type_k) || ggml_is_quantized(params.type_v)) {
LLAMA_LOG_ERROR("%s: simultaneous use of SPLIT_MODE_TENSOR and KV cache quantization not implemented\n", __func__);
return nullptr;
}
}

if (params.flash_attn_type != LLAMA_FLASH_ATTN_TYPE_DISABLED && ggml_is_quantized(params.type_k)) {
Expand Down
5 changes: 3 additions & 2 deletions src/llama-graph.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,10 +65,11 @@ static ggml_tensor * ggml_mul_mat_aux(

ggml_tensor * res;

GGML_ASSERT(cur->ne[0] % n == 0);
if (!ggml_is_contiguous(cur)) {
res = ggml_cont_2d (ctx, cur, n, ggml_nelements(cur)/n);
res = ggml_cont_4d(ctx, cur, n, cur->ne[0]/n, cur->ne[1], cur->ne[2]*cur->ne[3]);
} else {
res = ggml_reshape_2d(ctx, cur, n, ggml_nelements(cur)/n);
res = ggml_reshape_4d(ctx, cur, n, cur->ne[0]/n, cur->ne[1], cur->ne[2]*cur->ne[3]);
}
res = ggml_mul_mat (ctx, rot, res);
ggml_mul_mat_set_hint(res, GGML_HINT_SRC0_IS_HADAMARD);
Expand Down
3 changes: 2 additions & 1 deletion src/llama-kv-cache.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,7 +65,8 @@ static ggml_tensor * ggml_mul_mat_aux(

ggml_tensor * res;

res = ggml_reshape_2d(ctx, cur, n, ggml_nelements(cur)/n);
GGML_ASSERT(cur->ne[0] % n == 0);
res = ggml_reshape_4d(ctx, cur, n, cur->ne[0]/n, cur->ne[1], cur->ne[2]*cur->ne[3]);
res = ggml_mul_mat (ctx, rot, res);
ggml_mul_mat_set_hint(res, GGML_HINT_SRC0_IS_HADAMARD);
res = ggml_reshape_4d(ctx, res, cur->ne[0], cur->ne[1], cur->ne[2], cur->ne[3]);
Expand Down