Skip to content

llama : load MTP tensors only if they are really used #26296

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

Merged
fairydreaming merged 2 commits into from
Jul 31, 2026
Merged

llama : load MTP tensors only if they are really used #26296

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
fb6a9cc
llama : load MTP tensors only if they are really used
sszymczy Jul 29, 2026
4499d9f
llama : skip loading MTP (if not used) in remaining models that suppo…
sszymczy Jul 30, 2026
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
1 change: 1 addition & 0 deletions common/common.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1606,6 +1606,7 @@ struct llama_model_params common_model_params_to_llama(common_params & params) {
mparams.progress_callback = params.load_progress_callback;
mparams.progress_callback_user_data = params.load_progress_callback_user_data;
mparams.no_alloc = params.no_alloc;
mparams.load_mtp = std::find(params.speculative.types.begin(), params.speculative.types.end(), COMMON_SPECULATIVE_TYPE_DRAFT_MTP) != params.speculative.types.end();

return mparams;
}
Expand Down
1 change: 1 addition & 0 deletions include/llama.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -337,6 +337,7 @@ extern "C" {
bool use_extra_bufts; // use extra buffer types (used for weight repacking)
bool no_host; // bypass host buffer allowing extra buffers to be used
bool no_alloc; // only load metadata and simulate memory allocations
bool load_mtp; // whether to load MTP layers
};

struct llama_sampler_seq_config {
Expand Down
2 changes: 2 additions & 0 deletions src/llama-model-loader.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -526,6 +526,7 @@ llama_model_loader::llama_model_loader(
llama_load_mode load_mode,
bool check_tensors,
bool no_alloc,
bool load_mtp,
const llama_model_kv_override * param_overrides_p,
const llama_model_tensor_buft_override * param_tensor_buft_overrides_p)
: metadata(meta), set_tensor_data(set_tensor_data), set_tensor_data_ud(set_tensor_data_ud) {
Expand Down Expand Up @@ -812,6 +813,7 @@ llama_model_loader::llama_model_loader(

this->check_tensors = check_tensors;
this->no_alloc = no_alloc;
this->load_mtp = load_mtp;
}

std::string llama_model_loader::get_arch_name() const {
Expand Down
2 changes: 2 additions & 0 deletions src/llama-model-loader.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,6 +79,7 @@ struct llama_model_loader {
bool use_direct_io = false;
bool check_tensors;
bool no_alloc;
bool load_mtp;

llama_files files;
llama_ftype ftype;
Expand Down Expand Up @@ -129,6 +130,7 @@ struct llama_model_loader {
llama_load_mode load_mode,
bool check_tensors,
bool no_alloc,
bool load_mtp,
const llama_model_kv_override * param_overrides_p,
const llama_model_tensor_buft_override * param_tensor_buft_overrides_p);

Expand Down
1 change: 1 addition & 0 deletions src/llama-model.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2388,6 +2388,7 @@ llama_model_params llama_model_default_params() {
/*.use_extra_bufts =*/ true,
/*.no_host =*/ false,
/*.no_alloc =*/ false,
/*.load_mtp =*/ false,
};

return result;
Expand Down
2 changes: 1 addition & 1 deletion src/llama-quant.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -893,7 +893,7 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
const llama_model_kv_override * kv_overrides = params->kv_overrides;
std::vector<std::string> splits = {};
llama_model_loader ml(/*metadata*/ nullptr, /*set_tensor_data*/ nullptr, /*set_tensor_data_ud*/ nullptr,
fname_inp, splits, /*file*/ nullptr, /*load_mode*/ load_mode, /*check_tensors*/ true, /*no_alloc*/ false, kv_overrides, nullptr);
fname_inp, splits, /*file*/ nullptr, /*load_mode*/ load_mode, /*check_tensors*/ true, /*no_alloc*/ false, /*load_mtp*/ true, kv_overrides, nullptr);
Comment on lines 895 to +896

@ggerganov ggerganov Jul 31, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

for next PRs: we should refactor these params in a helper struct

ml.init_mappings(false); // no prefetching

auto mparams = llama_model_default_params();
Expand Down
2 changes: 1 addition & 1 deletion src/llama.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -305,7 +305,7 @@ static std::pair<int, llama_model *> llama_model_load(struct gguf_context * meta
const std::string & fname, std::vector<std::string> & splits, FILE * file, llama_model_params & params) {
try {
llama_model_loader ml(metadata, set_tensor_data, set_tensor_data_ud, fname, splits, file, params.load_mode,
params.check_tensors, params.no_alloc, params.kv_overrides, params.tensor_buft_overrides);
params.check_tensors, params.no_alloc, params.load_mtp, params.kv_overrides, params.tensor_buft_overrides);

ml.print_info();
std::unique_ptr<llama_model> model_ptr(llama_model_create(ml, params));
Expand Down
6 changes: 5 additions & 1 deletion src/models/cohere2moe.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,7 +55,11 @@ void llama_model_cohere2moe::load_arch_tensors(llama_model_loader & ml) {
const std::string mtp_probe = "blk." + std::to_string(n_layer) + ".nextn.eh_proj.weight";
const bool trunk_only = (hparams.n_layer_nextn > 0) && (ml.get_weight(mtp_probe.c_str()) == nullptr);
const int trunk_flags = mtp_only ? TENSOR_NOT_REQUIRED : 0;
const int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;
int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;

if (!ml.load_mtp) {
mtp_flags |= TENSOR_SKIP;
}

tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), { n_embd, n_vocab }, 0);

Expand Down
6 changes: 5 additions & 1 deletion src/models/glm-dsa.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,7 +91,11 @@ void llama_model_glm_dsa::load_arch_tensors(llama_model_loader & ml) {
const std::string mtp_probe = "blk." + std::to_string(n_layer) + ".nextn.eh_proj.weight";
const bool trunk_only = (hparams.n_layer_nextn > 0) && (ml.get_weight(mtp_probe.c_str()) == nullptr);
const int trunk_flags = mtp_only ? TENSOR_NOT_REQUIRED : 0;
const int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;
int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;

if (!ml.load_mtp) {
mtp_flags |= TENSOR_SKIP;
}

const bool is_mla = hparams.is_mla();
if (!is_mla) {
Expand Down
6 changes: 5 additions & 1 deletion src/models/hy-v3.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,7 +33,11 @@ void llama_model_hy_v3::load_arch_tensors(llama_model_loader & ml) {
const std::string mtp_probe = "blk." + std::to_string(n_layer) + ".nextn.eh_proj.weight";
const bool trunk_only = (hparams.n_layer_nextn > 0) && (ml.get_weight(mtp_probe.c_str()) == nullptr);
const int trunk_flags = mtp_only ? TENSOR_NOT_REQUIRED : 0;
const int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;
int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;

if (!ml.load_mtp) {
mtp_flags |= TENSOR_SKIP;
}

tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, 0);

Expand Down
31 changes: 16 additions & 15 deletions src/models/qwen35.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,7 @@ void llama_model_qwen35::load_arch_tensors(llama_model_loader & ml) {

const bool mtp_only = (hparams.n_layer_nextn > 0) && (ml.get_weight("blk.0.attn_norm.weight") == nullptr);
const int trunk_flags = mtp_only ? TENSOR_NOT_REQUIRED : 0;
int mtp_flags = !ml.load_mtp ? TENSOR_SKIP : 0;

tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), { n_embd, n_vocab }, 0);

Expand Down Expand Up @@ -97,25 +98,25 @@ void llama_model_qwen35::load_arch_tensors(llama_model_loader & ml) {
auto & layer = layers[il];

// MTP block looks like a full-attention Qwen3.5 decoder block.
layer.attn_norm = create_tensor(tn(LLM_TENSOR_ATTN_NORM, "weight", il), { n_embd }, 0);
layer.attn_post_norm = create_tensor(tn(LLM_TENSOR_ATTN_POST_NORM, "weight", il), { n_embd }, 0);
layer.attn_norm = create_tensor(tn(LLM_TENSOR_ATTN_NORM, "weight", il), { n_embd }, mtp_flags);
layer.attn_post_norm = create_tensor(tn(LLM_TENSOR_ATTN_POST_NORM, "weight", il), { n_embd }, mtp_flags);

create_tensor_qkv(layer, il, n_embd, n_embd_head_k * n_head * 2, n_embd_k_gqa, n_embd_v_gqa, 0);
layer.wo = create_tensor(tn(LLM_TENSOR_ATTN_OUT, "weight", il), { n_embd_head_k * n_head, n_embd }, 0);
layer.attn_q_norm = create_tensor(tn(LLM_TENSOR_ATTN_Q_NORM, "weight", il), { n_embd_head_k }, 0);
layer.attn_k_norm = create_tensor(tn(LLM_TENSOR_ATTN_K_NORM, "weight", il), { n_embd_head_k }, 0);
create_tensor_qkv(layer, il, n_embd, n_embd_head_k * n_head * 2, n_embd_k_gqa, n_embd_v_gqa, mtp_flags);
layer.wo = create_tensor(tn(LLM_TENSOR_ATTN_OUT, "weight", il), { n_embd_head_k * n_head, n_embd }, mtp_flags);
layer.attn_q_norm = create_tensor(tn(LLM_TENSOR_ATTN_Q_NORM, "weight", il), { n_embd_head_k }, mtp_flags);
layer.attn_k_norm = create_tensor(tn(LLM_TENSOR_ATTN_K_NORM, "weight", il), { n_embd_head_k }, mtp_flags);

layer.ffn_gate = create_tensor(tn(LLM_TENSOR_FFN_GATE, "weight", il), {n_embd, n_ff}, 0);
layer.ffn_down = create_tensor(tn(LLM_TENSOR_FFN_DOWN, "weight", il), { n_ff, n_embd}, 0);
layer.ffn_up = create_tensor(tn(LLM_TENSOR_FFN_UP, "weight", il), {n_embd, n_ff}, 0);
layer.ffn_gate = create_tensor(tn(LLM_TENSOR_FFN_GATE, "weight", il), {n_embd, n_ff}, mtp_flags);
layer.ffn_down = create_tensor(tn(LLM_TENSOR_FFN_DOWN, "weight", il), { n_ff, n_embd}, mtp_flags);
layer.ffn_up = create_tensor(tn(LLM_TENSOR_FFN_UP, "weight", il), {n_embd, n_ff}, mtp_flags);

// NextN-specific tensors that define the MTP block.
layer.nextn.eh_proj = create_tensor(tn(LLM_TENSOR_NEXTN_EH_PROJ, "weight", il), { 2 * n_embd, n_embd }, 0);
layer.nextn.enorm = create_tensor(tn(LLM_TENSOR_NEXTN_ENORM, "weight", il), { n_embd }, 0);
layer.nextn.hnorm = create_tensor(tn(LLM_TENSOR_NEXTN_HNORM, "weight", il), { n_embd }, 0);
layer.nextn.embed_tokens = create_tensor(tn(LLM_TENSOR_NEXTN_EMBED_TOKENS, "weight", il), { n_embd, n_vocab }, TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_head = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD, "weight", il), { n_embd, n_vocab }, TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_norm = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_NORM, "weight", il), { n_embd }, TENSOR_NOT_REQUIRED);
layer.nextn.eh_proj = create_tensor(tn(LLM_TENSOR_NEXTN_EH_PROJ, "weight", il), { 2 * n_embd, n_embd }, mtp_flags);
layer.nextn.enorm = create_tensor(tn(LLM_TENSOR_NEXTN_ENORM, "weight", il), { n_embd }, mtp_flags);
layer.nextn.hnorm = create_tensor(tn(LLM_TENSOR_NEXTN_HNORM, "weight", il), { n_embd }, mtp_flags);
layer.nextn.embed_tokens = create_tensor(tn(LLM_TENSOR_NEXTN_EMBED_TOKENS, "weight", il), { n_embd, n_vocab }, mtp_flags|TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_head = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD, "weight", il), { n_embd, n_vocab }, mtp_flags|TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_norm = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_NORM, "weight", il), { n_embd }, mtp_flags|TENSOR_NOT_REQUIRED);
};

for (int i = 0; i < n_layer; ++i) {
Expand Down
39 changes: 20 additions & 19 deletions src/models/qwen35moe.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,7 @@ void llama_model_qwen35moe::load_arch_tensors(llama_model_loader & ml) {

const bool mtp_only = (hparams.n_layer_nextn > 0) && (ml.get_weight("blk.0.attn_norm.weight") == nullptr);
const int trunk_flags = mtp_only ? TENSOR_NOT_REQUIRED : 0;
int mtp_flags = !ml.load_mtp ? TENSOR_SKIP : 0;

tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), { n_embd, n_vocab }, 0);

Expand Down Expand Up @@ -113,32 +114,32 @@ void llama_model_qwen35moe::load_arch_tensors(llama_model_loader & ml) {
const int64_t n_ff_shexp = hparams.n_ff_shexp ? hparams.n_ff_shexp : n_ff;

// MTP block looks like a full-attention Qwen3.5 decoder block with MoE FFN.
layer.attn_norm = create_tensor(tn(LLM_TENSOR_ATTN_NORM, "weight", il), { n_embd }, 0);
layer.attn_post_norm = create_tensor(tn(LLM_TENSOR_ATTN_POST_NORM, "weight", il), { n_embd }, 0);
layer.attn_norm = create_tensor(tn(LLM_TENSOR_ATTN_NORM, "weight", il), { n_embd }, mtp_flags);
layer.attn_post_norm = create_tensor(tn(LLM_TENSOR_ATTN_POST_NORM, "weight", il), { n_embd }, mtp_flags);

create_tensor_qkv(layer, il, n_embd, n_embd_head_k * n_head * 2, n_embd_k_gqa, n_embd_v_gqa, 0);
layer.wo = create_tensor(tn(LLM_TENSOR_ATTN_OUT, "weight", il), { n_embd_head_k * n_head, n_embd }, 0);
layer.attn_q_norm = create_tensor(tn(LLM_TENSOR_ATTN_Q_NORM, "weight", il), { n_embd_head_k }, 0);
layer.attn_k_norm = create_tensor(tn(LLM_TENSOR_ATTN_K_NORM, "weight", il), { n_embd_head_k }, 0);
create_tensor_qkv(layer, il, n_embd, n_embd_head_k * n_head * 2, n_embd_k_gqa, n_embd_v_gqa, mtp_flags);
layer.wo = create_tensor(tn(LLM_TENSOR_ATTN_OUT, "weight", il), { n_embd_head_k * n_head, n_embd }, mtp_flags);
layer.attn_q_norm = create_tensor(tn(LLM_TENSOR_ATTN_Q_NORM, "weight", il), { n_embd_head_k }, mtp_flags);
layer.attn_k_norm = create_tensor(tn(LLM_TENSOR_ATTN_K_NORM, "weight", il), { n_embd_head_k }, mtp_flags);

// Routed experts
layer.ffn_gate_inp = create_tensor(tn(LLM_TENSOR_FFN_GATE_INP, "weight", il), { n_embd, n_expert }, 0);
layer.ffn_down_exps = create_tensor(tn(LLM_TENSOR_FFN_DOWN_EXPS, "weight", il), { n_ff_exp, n_embd, n_expert }, 0);
create_tensor_gate_up_exps(layer, il, n_embd, n_ff_exp, n_expert, 0);
layer.ffn_gate_inp = create_tensor(tn(LLM_TENSOR_FFN_GATE_INP, "weight", il), { n_embd, n_expert }, mtp_flags);
layer.ffn_down_exps = create_tensor(tn(LLM_TENSOR_FFN_DOWN_EXPS, "weight", il), { n_ff_exp, n_embd, n_expert }, mtp_flags);
create_tensor_gate_up_exps(layer, il, n_embd, n_ff_exp, n_expert, mtp_flags);

// Shared experts
layer.ffn_gate_inp_shexp = create_tensor(tn(LLM_TENSOR_FFN_GATE_INP_SHEXP, "weight", il), { n_embd }, 0);
layer.ffn_gate_shexp = create_tensor(tn(LLM_TENSOR_FFN_GATE_SHEXP, "weight", il), { n_embd, n_ff_shexp }, 0);
layer.ffn_up_shexp = create_tensor(tn(LLM_TENSOR_FFN_UP_SHEXP, "weight", il), { n_embd, n_ff_shexp }, 0);
layer.ffn_down_shexp = create_tensor(tn(LLM_TENSOR_FFN_DOWN_SHEXP, "weight", il), { n_ff_shexp, n_embd }, 0);
layer.ffn_gate_inp_shexp = create_tensor(tn(LLM_TENSOR_FFN_GATE_INP_SHEXP, "weight", il), { n_embd }, mtp_flags);
layer.ffn_gate_shexp = create_tensor(tn(LLM_TENSOR_FFN_GATE_SHEXP, "weight", il), { n_embd, n_ff_shexp }, mtp_flags);
layer.ffn_up_shexp = create_tensor(tn(LLM_TENSOR_FFN_UP_SHEXP, "weight", il), { n_embd, n_ff_shexp }, mtp_flags);
layer.ffn_down_shexp = create_tensor(tn(LLM_TENSOR_FFN_DOWN_SHEXP, "weight", il), { n_ff_shexp, n_embd }, mtp_flags);

// NextN-specific tensors that define the MTP block.
layer.nextn.eh_proj = create_tensor(tn(LLM_TENSOR_NEXTN_EH_PROJ, "weight", il), { 2 * n_embd, n_embd }, 0);
layer.nextn.enorm = create_tensor(tn(LLM_TENSOR_NEXTN_ENORM, "weight", il), { n_embd }, 0);
layer.nextn.hnorm = create_tensor(tn(LLM_TENSOR_NEXTN_HNORM, "weight", il), { n_embd }, 0);
layer.nextn.embed_tokens = create_tensor(tn(LLM_TENSOR_NEXTN_EMBED_TOKENS, "weight", il), { n_embd, n_vocab }, TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_head = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD, "weight", il), { n_embd, n_vocab }, TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_norm = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_NORM, "weight", il), { n_embd }, TENSOR_NOT_REQUIRED);
layer.nextn.eh_proj = create_tensor(tn(LLM_TENSOR_NEXTN_EH_PROJ, "weight", il), { 2 * n_embd, n_embd }, mtp_flags);
layer.nextn.enorm = create_tensor(tn(LLM_TENSOR_NEXTN_ENORM, "weight", il), { n_embd }, mtp_flags);
layer.nextn.hnorm = create_tensor(tn(LLM_TENSOR_NEXTN_HNORM, "weight", il), { n_embd }, mtp_flags);
layer.nextn.embed_tokens = create_tensor(tn(LLM_TENSOR_NEXTN_EMBED_TOKENS, "weight", il), { n_embd, n_vocab }, mtp_flags|TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_head = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_HEAD, "weight", il), { n_embd, n_vocab }, mtp_flags|TENSOR_NOT_REQUIRED);
layer.nextn.shared_head_norm = create_tensor(tn(LLM_TENSOR_NEXTN_SHARED_HEAD_NORM, "weight", il), { n_embd }, mtp_flags|TENSOR_NOT_REQUIRED);
};

for (int i = 0; i < n_layer; ++i) {
Expand Down
6 changes: 5 additions & 1 deletion src/models/step35.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,11 @@ void llama_model_step35::load_arch_tensors(llama_model_loader & ml) {
const std::string mtp_probe = "blk." + std::to_string(n_layer) + ".nextn.eh_proj.weight";
const bool trunk_only = (hparams.n_layer_nextn > 0) && (ml.get_weight(mtp_probe.c_str()) == nullptr);
const int trunk_flags = mtp_only ? TENSOR_NOT_REQUIRED : 0;
const int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;
int mtp_flags = trunk_only ? TENSOR_NOT_REQUIRED : 0;

if (!ml.load_mtp) {
mtp_flags |= TENSOR_SKIP;
}

tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, 0);

Expand Down
Loading