vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-05-11 13:11:49 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

Merge branch 'upstream' into concedo_experimental

Browse source 
# Conflicts:
#	.devops/vulkan.Dockerfile
#	.github/workflows/build.yml
#	.github/workflows/server.yml
#	common/common.cpp
#	examples/batched/README.md
#	ggml/CMakeLists.txt
#	ggml/src/CMakeLists.txt
#	ggml/src/ggml-cann/ggml-cann.cpp
#	ggml/src/ggml-cpu/CMakeLists.txt
#	ggml/src/ggml-cpu/arch-fallback.h
#	ggml/src/ggml-opencl/ggml-opencl.cpp
#	scripts/sync-ggml.last
#	src/CMakeLists.txt
#	tests/test-backend-ops.cpp
#	tools/server/CMakeLists.txt
This commit is contained in:
Concedo 2025-11-25 16:38:07 +08:00
commit 724763fdec
55 changed files with 5248 additions and 3753 deletions
Download patch file Download diff file Expand all files Collapse all files

60
src/llama-arch.cpp
View file

@ -108,24 +108,37 @@ static const std::map<llm_arch, const char *> LLM_ARCH_NAMES = {
{ LLM_ARCH_APERTUS, "apertus" },
{ LLM_ARCH_MINIMAX_M2, "minimax-m2" },
{ LLM_ARCH_COGVLM, "cogvlm" },
{ LLM_ARCH_RND1, "rnd1" },
{ LLM_ARCH_PANGU_EMBED, "pangu-embedded" },
{ LLM_ARCH_UNKNOWN, "(unknown)" },
};
static const std::map<llm_kv, const char *> LLM_KV_NAMES = {
{ LLM_KV_GENERAL_TYPE, "general.type" },
{ LLM_KV_GENERAL_ARCHITECTURE, "general.architecture" },
{ LLM_KV_GENERAL_QUANTIZATION_VERSION, "general.quantization_version" },
{ LLM_KV_GENERAL_ALIGNMENT, "general.alignment" },
{ LLM_KV_GENERAL_FILE_TYPE, "general.file_type" },
{ LLM_KV_GENERAL_NAME, "general.name" },
{ LLM_KV_GENERAL_AUTHOR, "general.author" },
{ LLM_KV_GENERAL_VERSION, "general.version" },
{ LLM_KV_GENERAL_URL, "general.url" },
{ LLM_KV_GENERAL_DESCRIPTION, "general.description" },
{ LLM_KV_GENERAL_LICENSE, "general.license" },
{ LLM_KV_GENERAL_SOURCE_URL, "general.source.url" },
{ LLM_KV_GENERAL_SOURCE_HF_REPO, "general.source.huggingface.repository" },
{ LLM_KV_GENERAL_TYPE, "general.type" },
{ LLM_KV_GENERAL_ARCHITECTURE, "general.architecture" },
{ LLM_KV_GENERAL_QUANTIZATION_VERSION, "general.quantization_version" },
{ LLM_KV_GENERAL_ALIGNMENT, "general.alignment" },
{ LLM_KV_GENERAL_FILE_TYPE, "general.file_type" },
{ LLM_KV_GENERAL_SAMPLING_SEQUENCE, "general.sampling.sequence" },
{ LLM_KV_GENERAL_SAMPLING_TOP_K, "general.sampling.top_k" },
{ LLM_KV_GENERAL_SAMPLING_TOP_P, "general.sampling.top_p" },
{ LLM_KV_GENERAL_SAMPLING_MIN_P, "general.sampling.min_p" },
{ LLM_KV_GENERAL_SAMPLING_XTC_PROBABILITY, "general.sampling.xtc_probability" },
{ LLM_KV_GENERAL_SAMPLING_XTC_THRESHOLD, "general.sampling.xtc_threshold" },
{ LLM_KV_GENERAL_SAMPLING_TEMP, "general.sampling.temp" },
{ LLM_KV_GENERAL_SAMPLING_PENALTY_LAST_N, "general.sampling.penalty_last_n" },
{ LLM_KV_GENERAL_SAMPLING_PENALTY_REPEAT, "general.sampling.penalty_repeat" },
{ LLM_KV_GENERAL_SAMPLING_MIROSTAT, "general.sampling.mirostat" },
{ LLM_KV_GENERAL_SAMPLING_MIROSTAT_TAU, "general.sampling.mirostat_tau" },
{ LLM_KV_GENERAL_SAMPLING_MIROSTAT_ETA, "general.sampling.mirostat_eta" },
{ LLM_KV_GENERAL_NAME, "general.name" },
{ LLM_KV_GENERAL_AUTHOR, "general.author" },
{ LLM_KV_GENERAL_VERSION, "general.version" },
{ LLM_KV_GENERAL_URL, "general.url" },
{ LLM_KV_GENERAL_DESCRIPTION, "general.description" },
{ LLM_KV_GENERAL_LICENSE, "general.license" },
{ LLM_KV_GENERAL_SOURCE_URL, "general.source.url" },
{ LLM_KV_GENERAL_SOURCE_HF_REPO, "general.source.huggingface.repository" },
{ LLM_KV_VOCAB_SIZE, "%s.vocab_size" },
{ LLM_KV_CONTEXT_LENGTH, "%s.context_length" },
@ -2446,6 +2459,26 @@ static const std::map<llm_arch, std::map<llm_tensor, const char *>> LLM_TENSOR_N
{ LLM_TENSOR_VISEXP_FFN_UP, "blk.%d.vis_up" },
},
},
{
LLM_ARCH_RND1,
{
{ LLM_TENSOR_TOKEN_EMBD, "token_embd" },
{ LLM_TENSOR_OUTPUT_NORM, "output_norm" },
{ LLM_TENSOR_OUTPUT, "output" },
{ LLM_TENSOR_ATTN_NORM, "blk.%d.attn_norm" },
{ LLM_TENSOR_ATTN_Q, "blk.%d.attn_q" },
{ LLM_TENSOR_ATTN_Q_NORM, "blk.%d.attn_q_norm" },
{ LLM_TENSOR_ATTN_K, "blk.%d.attn_k" },
{ LLM_TENSOR_ATTN_K_NORM, "blk.%d.attn_k_norm" },
{ LLM_TENSOR_ATTN_V, "blk.%d.attn_v" },
{ LLM_TENSOR_ATTN_OUT, "blk.%d.attn_output" },
{ LLM_TENSOR_FFN_NORM, "blk.%d.ffn_norm" },
{ LLM_TENSOR_FFN_GATE_INP, "blk.%d.ffn_gate_inp" },
{ LLM_TENSOR_FFN_GATE_EXPS, "blk.%d.ffn_gate_exps" },
{ LLM_TENSOR_FFN_DOWN_EXPS, "blk.%d.ffn_down_exps" },
{ LLM_TENSOR_FFN_UP_EXPS, "blk.%d.ffn_up_exps" },
},
},
{
LLM_ARCH_UNKNOWN,
{
@ -2722,6 +2755,7 @@ bool llm_arch_is_diffusion(const llm_arch & arch) {
case LLM_ARCH_DREAM:
case LLM_ARCH_LLADA:
case LLM_ARCH_LLADA_MOE:
case LLM_ARCH_RND1:
return true;
default:
return false;

13
src/llama-arch.h
View file

@ -112,6 +112,7 @@ enum llm_arch {
LLM_ARCH_APERTUS,
LLM_ARCH_MINIMAX_M2,
LLM_ARCH_COGVLM,
LLM_ARCH_RND1,
LLM_ARCH_PANGU_EMBED,
LLM_ARCH_UNKNOWN,
};
@ -122,6 +123,18 @@ enum llm_kv {
LLM_KV_GENERAL_QUANTIZATION_VERSION,
LLM_KV_GENERAL_ALIGNMENT,
LLM_KV_GENERAL_FILE_TYPE,
LLM_KV_GENERAL_SAMPLING_SEQUENCE,
LLM_KV_GENERAL_SAMPLING_TOP_K,
LLM_KV_GENERAL_SAMPLING_TOP_P,
LLM_KV_GENERAL_SAMPLING_MIN_P,
LLM_KV_GENERAL_SAMPLING_XTC_PROBABILITY,
LLM_KV_GENERAL_SAMPLING_XTC_THRESHOLD,
LLM_KV_GENERAL_SAMPLING_TEMP,
LLM_KV_GENERAL_SAMPLING_PENALTY_LAST_N,
LLM_KV_GENERAL_SAMPLING_PENALTY_REPEAT,
LLM_KV_GENERAL_SAMPLING_MIROSTAT,
LLM_KV_GENERAL_SAMPLING_MIROSTAT_TAU,
LLM_KV_GENERAL_SAMPLING_MIROSTAT_ETA,
LLM_KV_GENERAL_NAME,
LLM_KV_GENERAL_AUTHOR,
LLM_KV_GENERAL_VERSION,

2
src/llama-context.cpp
View file

@ -1250,7 +1250,7 @@ int llama_context::decode(const llama_batch & batch_inp) {
// make the outputs have the same order they had in the user-provided batch
// note: this is mostly relevant for recurrent models atm
if (!sorted_output) {
if (!sorted_output && n_outputs > 1) {
GGML_ASSERT((size_t) n_outputs == out_ids.size());
// TODO: is there something more efficient which also minimizes swaps?

47
src/llama-model.cpp
View file

@ -107,6 +107,7 @@
#include "models/qwen3vl-moe.cpp"
#include "models/qwen3moe.cpp"
#include "models/refact.cpp"
#include "models/rnd1.cpp"
#include "models/rwkv6-base.cpp"
#include "models/rwkv6.cpp"
#include "models/rwkv6qwen2.cpp"
@ -1138,6 +1139,18 @@ void llama_model::load_hparams(llama_model_loader & ml) {
default: type = LLM_TYPE_UNKNOWN;
}
} break;
case LLM_ARCH_RND1:
{
ml.get_key(LLM_KV_EXPERT_FEED_FORWARD_LENGTH, hparams.n_ff_exp, false);
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
switch (hparams.n_layer) {
case 48: type = LLM_TYPE_30B_A3B; break;
default: type = LLM_TYPE_UNKNOWN;
}
// Set non-causal attention for diffusion models
hparams.causal_attn = false;
} break;
case LLM_ARCH_QWEN2MOE:
{
ml.get_key(LLM_KV_EXPERT_FEED_FORWARD_LENGTH, hparams.n_ff_exp, false);
@ -1695,7 +1708,8 @@ void llama_model::load_hparams(llama_model_loader & ml) {
} break;
case LLM_ARCH_DEEPSEEK2:
{
bool is_lite = (hparams.n_layer == 27);
// lite variants include DeepSeek-V2-Lite, GigaChat3-10B-A1.8B
bool is_lite = (hparams.n_layer == 27 || hparams.n_layer == 26);
ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps);
ml.get_key(LLM_KV_LEADING_DENSE_BLOCK_COUNT, hparams.n_layer_dense_lead);
if (!is_lite) {
@ -3556,6 +3570,7 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
} break;
case LLM_ARCH_QWEN3MOE:
case LLM_ARCH_QWEN3VLMOE:
case LLM_ARCH_RND1:
{
tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, 0);
@ -4736,7 +4751,8 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
} break;
case LLM_ARCH_DEEPSEEK2:
{
const bool is_lite = (hparams.n_layer == 27);
// lite variants include DeepSeek-V2-Lite, GigaChat3-10B-A1.8B
const bool is_lite = (hparams.n_layer == 27 || hparams.n_layer == 26);
const bool is_mla = (hparams.n_embd_head_k_mla != 0 && hparams.n_embd_head_v_mla != 0);
@ -6874,7 +6890,7 @@ void llama_model::print_info() const {
LLAMA_LOG_INFO("%s: n_ff_shexp = %d\n", __func__, hparams.n_ff_shexp);
}
if (arch == LLM_ARCH_QWEN3MOE || arch == LLM_ARCH_OPENAI_MOE || arch == LLM_ARCH_QWEN3VLMOE) {
if (arch == LLM_ARCH_QWEN3MOE || arch == LLM_ARCH_OPENAI_MOE || arch == LLM_ARCH_QWEN3VLMOE || arch == LLM_ARCH_RND1) {
LLAMA_LOG_INFO("%s: n_ff_exp = %d\n", __func__, hparams.n_ff_exp);
}
@ -7039,6 +7055,7 @@ llama_memory_i * llama_model::create_memory(const llama_memory_params & params,
case LLM_ARCH_DREAM:
case LLM_ARCH_LLADA:
case LLM_ARCH_LLADA_MOE:
case LLM_ARCH_RND1:
{
res = nullptr;
} break;
@ -7232,6 +7249,11 @@ ggml_cgraph * llama_model::build_graph(const llm_graph_params & params) const {
llm = std::make_unique<llm_build_llada_moe>(*this, params);
}
break;
case LLM_ARCH_RND1:
{
llm = std::make_unique<llm_build_rnd1>(*this, params);
}
break;
case LLM_ARCH_QWEN2VL:
{
llm = std::make_unique<llm_build_qwen2vl>(*this, params);
@ -7752,6 +7774,7 @@ llama_rope_type llama_model_rope_type(const llama_model * model) {
case LLM_ARCH_QWEN3:
case LLM_ARCH_QWEN3MOE:
case LLM_ARCH_LLADA_MOE:
case LLM_ARCH_RND1:
case LLM_ARCH_OLMO2:
case LLM_ARCH_OLMOE:
case LLM_ARCH_PHI2:
@ -7824,6 +7847,24 @@ int32_t llama_model_meta_count(const llama_model * model) {
return (int)model->gguf_kv.size();
}
const char * llama_model_meta_key_str(llama_model_meta_key key) {
switch (key) {
case LLAMA_MODEL_META_KEY_SAMPLING_SEQUENCE: return "general.sampling.sequence";
case LLAMA_MODEL_META_KEY_SAMPLING_TOP_K: return "general.sampling.top_k";
case LLAMA_MODEL_META_KEY_SAMPLING_TOP_P: return "general.sampling.top_p";
case LLAMA_MODEL_META_KEY_SAMPLING_MIN_P: return "general.sampling.min_p";
case LLAMA_MODEL_META_KEY_SAMPLING_XTC_PROBABILITY: return "general.sampling.xtc_probability";
case LLAMA_MODEL_META_KEY_SAMPLING_XTC_THRESHOLD: return "general.sampling.xtc_threshold";
case LLAMA_MODEL_META_KEY_SAMPLING_TEMP: return "general.sampling.temp";
case LLAMA_MODEL_META_KEY_SAMPLING_PENALTY_LAST_N: return "general.sampling.penalty_last_n";
case LLAMA_MODEL_META_KEY_SAMPLING_PENALTY_REPEAT: return "general.sampling.penalty_repeat";
case LLAMA_MODEL_META_KEY_SAMPLING_MIROSTAT: return "general.sampling.mirostat";
case LLAMA_MODEL_META_KEY_SAMPLING_MIROSTAT_TAU: return "general.sampling.mirostat_tau";
case LLAMA_MODEL_META_KEY_SAMPLING_MIROSTAT_ETA: return "general.sampling.mirostat_eta";
default: return nullptr;
}
}
int32_t llama_model_meta_key_by_index(const llama_model * model, int i, char * buf, size_t buf_size) {
if (i < 0 || i >= (int)model->gguf_kv.size()) {
if (buf_size > 0) {

3
src/models/deepseek2.cpp
View file

@ -4,7 +4,8 @@
llm_build_deepseek2::llm_build_deepseek2(const llama_model & model, const llm_graph_params & params) :
llm_graph_context(params) {
bool is_lite = (hparams.n_layer == 27);
// lite variants include DeepSeek-V2-Lite, GigaChat3-10B-A1.8B
bool is_lite = (hparams.n_layer == 27 || hparams.n_layer == 26);
const bool is_mla = (hparams.n_embd_head_k_mla != 0 && hparams.n_embd_head_v_mla != 0);

4
src/models/models.h
View file

@ -431,6 +431,10 @@ struct llm_build_refact : public llm_graph_context {
llm_build_refact(const llama_model & model, const llm_graph_params & params);
};
struct llm_build_rnd1 : public llm_graph_context {
llm_build_rnd1(const llama_model & model, const llm_graph_params & params);
};
struct llm_build_rwkv6 : public llm_build_rwkv6_base {
llm_build_rwkv6(const llama_model & model, const llm_graph_params & params);
};

126
src/models/rnd1.cpp Normal file
View file

@ -0,0 +1,126 @@
#include "models.h"
// RND1 is a Qwen3Moe AR model converted to diffusion model.
llm_build_rnd1::llm_build_rnd1(const llama_model & model, const llm_graph_params & params) : llm_graph_context(params) {
const int64_t n_embd_head = hparams.n_embd_head_v;
GGML_ASSERT(n_embd_head == hparams.n_embd_head_k);
GGML_ASSERT(n_embd_head == hparams.n_rot);
ggml_tensor * cur;
ggml_tensor * inpL;
inpL = build_inp_embd(model.tok_embd);
// inp_pos - contains the positions
ggml_tensor * inp_pos = build_inp_pos();
// Non-causal attention for diffusion
auto * inp_attn = build_attn_inp_no_cache();
ggml_tensor * inp_out_ids = build_inp_out_ids();
for (int il = 0; il < n_layer; ++il) {
ggml_tensor * inpSA = inpL;
// norm
cur = build_norm(inpL,
model.layers[il].attn_norm, NULL,
LLM_NORM_RMS, il);
cb(cur, "attn_norm", il);
// self_attention
{
// compute Q and K and RoPE them
ggml_tensor * Qcur = build_lora_mm(model.layers[il].wq, cur);
cb(Qcur, "Qcur", il);
ggml_tensor * Kcur = build_lora_mm(model.layers[il].wk, cur);
cb(Kcur, "Kcur", il);
ggml_tensor * Vcur = build_lora_mm(model.layers[il].wv, cur);
cb(Vcur, "Vcur", il);
Qcur = ggml_reshape_3d(ctx0, Qcur, n_embd_head, n_head, n_tokens);
Kcur = ggml_reshape_3d(ctx0, Kcur, n_embd_head, n_head_kv, n_tokens);
Vcur = ggml_reshape_3d(ctx0, Vcur, n_embd_head, n_head_kv, n_tokens);
Qcur = build_norm(Qcur, model.layers[il].attn_q_norm, NULL, LLM_NORM_RMS, il);
cb(Qcur, "Qcur_normed", il);
Qcur = ggml_rope_ext(
ctx0, Qcur, inp_pos, nullptr,
n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,
ext_factor, attn_factor, beta_fast, beta_slow
);
Kcur = build_norm(Kcur, model.layers[il].attn_k_norm, NULL, LLM_NORM_RMS, il);
cb(Kcur, "Kcur_normed", il);
Kcur = ggml_rope_ext(
ctx0, Kcur, inp_pos, nullptr,
n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,
ext_factor, attn_factor, beta_fast, beta_slow
);
cb(Qcur, "Qcur", il);
cb(Kcur, "Kcur", il);
cb(Vcur, "Vcur", il);
cur = build_attn(inp_attn,
model.layers[il].wo, model.layers[il].bo,
Qcur, Kcur, Vcur, nullptr, nullptr, nullptr, 1.0f/sqrtf(float(n_embd_head)), il);
}
if (il == n_layer - 1 && inp_out_ids) {
cur = ggml_get_rows(ctx0, cur, inp_out_ids);
inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);
}
ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);
cb(ffn_inp, "ffn_inp", il);
// MoE branch
cur = build_norm(ffn_inp,
model.layers[il].ffn_norm, NULL,
LLM_NORM_RMS, il);
cb(cur, "ffn_norm", il);
ggml_tensor * moe_out =
build_moe_ffn(cur,
model.layers[il].ffn_gate_inp,
model.layers[il].ffn_up_exps,
model.layers[il].ffn_gate_exps,
model.layers[il].ffn_down_exps,
nullptr,
n_expert, n_expert_used,
LLM_FFN_SILU, true,
false, 0.0,
LLAMA_EXPERT_GATING_FUNC_TYPE_SOFTMAX,
il);
cb(moe_out, "ffn_moe_out", il);
cur = moe_out;
cur = ggml_add(ctx0, cur, ffn_inp);
cur = build_cvec(cur, il);
cb(cur, "l_out", il);
// input for next layer
inpL = cur;
}
cur = inpL;
cur = build_norm(cur,
model.output_norm, NULL,
LLM_NORM_RMS, -1);
cb(cur, "result_norm", -1);
res->t_embd = cur;
// lm_head
cur = build_lora_mm(model.output, cur);
cb(cur, "result_output", -1);
res->t_logits = cur;
ggml_build_forward_expand(gf, cur);
}