Skip to content

Remove hardcoded number of output groups and rank #347

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
LifeDJIK wants to merge 1 commit into
base: main
Choose a base branch
from
+6 −6
Open

Remove hardcoded number of output groups and rank #347

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
12 changes: 6 additions & 6 deletions ds4.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6911,10 +6911,10 @@ static void layer_grouped_out_one(
const ds4_model * model,
const ds4_layer_weights * layer,
const float * heads) {
const uint32_t n_groups = 8;
const uint32_t n_groups = DS4_N_OUT_GROUP;
const uint32_t group_heads = DS4_N_HEAD / n_groups;
const uint32_t group_dim = DS4_N_HEAD_DIM * group_heads;
const uint32_t rank = 1024;
const uint32_t rank = DS4_N_LORA_O;

float *low = xcalloc((size_t)n_groups * rank, sizeof(low[0]));

Expand All @@ -6930,10 +6930,10 @@ static void layer_grouped_out_one_decode_scratch(
const ds4_layer_weights * layer,
const float * heads,
ds4_cpu_decode_scratch * scratch) {
const uint32_t n_groups = 8;
const uint32_t n_groups = DS4_N_OUT_GROUP;
const uint32_t group_heads = DS4_N_HEAD / n_groups;
const uint32_t group_dim = DS4_N_HEAD_DIM * group_heads;
const uint32_t rank = 1024;
const uint32_t rank = DS4_N_LORA_O;

memset(scratch->attn_low, 0, (size_t)n_groups * rank * sizeof(scratch->attn_low[0]));
matvec_q8_0_grouped_rows_decode_scratch(scratch->attn_low, model, layer->attn_output_a,
Expand All @@ -6947,10 +6947,10 @@ static void layer_grouped_out_batch(
const ds4_layer_weights * layer,
const float * heads,
uint32_t n_tok) {
const uint32_t n_groups = 8;
const uint32_t n_groups = DS4_N_OUT_GROUP;
const uint32_t group_heads = DS4_N_HEAD / n_groups;
const uint32_t group_dim = DS4_N_HEAD_DIM * group_heads;
const uint32_t rank = 1024;
const uint32_t rank = DS4_N_LORA_O;

float *low = xcalloc((size_t)n_tok * n_groups * rank, sizeof(low[0]));

Expand Down