ikawrakow/ik_llama.cpp
1
0
Fork 0
mirror of https://github.com/ikawrakow/ik_llama.cpp.git synced 2026-01-26 09:09:50 +00:00
Code Issues Packages Projects Releases Wiki Activity

Faster hybrid inference when shared experts

Browse Source
This commit is contained in:
Kawrakow
2026-01-25 14:38:54 +00:00
parent 28f8320f3a
commit 109686af6f
1 changed files with 18 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
src/llama-build-context.cpp
View File

@@ -1207,6 +1207,19 @@ llm_expert_gating_func_type gating_op,
GGML_ASSERT(!split_up_b_shexp || split_up_b_shexp->n_device == split_up_shexp->n_device);
GGML_ASSERT(!split_gate_b_shexp || split_gate_b_shexp->n_device == split_up_shexp->n_device);
GGML_ASSERT(!split_down_b_shexp || split_down_b_shexp->n_device == split_up_shexp->n_device);
bool down_bias_added = false;
int id_add_routed = -1;
if (split_up_shexp->splits[lctx.model.main_gpu]) {
id_add_routed = lctx.model.main_gpu;
} else {
for (int id = 0; id < split_up_shexp->n_device; ++id) {
if (split_up_shexp->splits[id]) {
id_add_routed = id;
break;
}
}
}
GGML_ASSERT(id_add_routed >= 0);
for (int id = 0; id < split_up_shexp->n_device; ++id) {
int il_cb = 1000*id + il;
GGML_ASSERT((split_up_shexp->splits[id] && split_gate_shexp->splits[id] && split_down_shexp->splits[id]) ||
@@ -1216,32 +1229,18 @@ llm_expert_gating_func_type gating_op,
auto shared_out = llm_build_ffn(ctx, lctx, the_ffn_norm, input,
split_up_shexp->splits[id], split_up_b_shexp ? split_up_b_shexp->splits[id] : nullptr, nullptr,
split_gate_shexp->splits[id], split_gate_b_shexp ? split_gate_b_shexp->splits[id] : nullptr, nullptr,
split_down_shexp->splits[id], split_down_b_shexp ? split_down_b_shexp->splits[id] : nullptr, nullptr,
nullptr, type_op_shexp, LLM_FFN_PAR, cb, il);
split_down_shexp->splits[id], !down_bias_added && split_down_b_shexp ? split_down_b_shexp->splits[id] : nullptr, nullptr,
nullptr, type_op_shexp, LLM_FFN_PAR, cb, il, graph, false, false,
id == id_add_routed ? routed_out : nullptr);
cb(shared_out, "ffn_shexp_out", il_cb);
if (shared_out->ne[1] > 32 && lctx.cparams.reduce_type != GGML_TYPE_F32) {
shared_out = ggml_cast(ctx, shared_out, lctx.cparams.reduce_type);
}
down_bias_added = true;
results.push_back(shared_out);
}
GGML_ASSERT(!results.empty());
if (results.size() == 1) {
cur = results.front();
} else {
cur = ggml_add(ctx, results[0], results[1]);
cur->op_params[0] = 0xff;
cb(cur, "ffn_shared_combined", il);
for (int id = 2; id < int(results.size()); ++id) {
cur = ggml_add(ctx, cur, results[id]);
cb(cur, "ffn_shared_combined", il);
}
}
if (routed_out->ne[1] > 32 && lctx.cparams.reduce_type != GGML_TYPE_F32) {
auto routed_out_f16 = ggml_cast(ctx, routed_out, lctx.cparams.reduce_type);
cur = ggml_add(ctx, routed_out_f16, cur);
} else {
cur = ggml_add(ctx, routed_out, cur);
}
cur = ggml_reduce(ctx, results.data(), split_up_shexp->n_device, GGML_OP_ADD);
cb(cur, "ffn_out", il);
} else {
auto shared_out = llm_build_ffn(ctx, lctx, nullptr, cur,