upstream-archive/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-07-23 11:10:55 -05:00
Code Issues Packages Projects Releases Wiki Activity

DeepseekV4: Add fused hyper-connection ops (#25585)

Browse Source 
* dsv4 hc-ops

* add missing files;

* add cparams

* update rpc version

* address review comments

* address review comments
This commit is contained in:
Aman Gupta
2026-07-17 00:33:33 +08:00
committed by GitHub
parent b2dd28a3b6
commit 0dc74e332e
16 changed files with 1062 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/include/ggml-rpc.h
View File

@@ -8,10 +8,10 @@ extern "C" {
#define RPC_PROTO_MAJOR_VERSION 4
#define RPC_PROTO_MINOR_VERSION 0
#define RPC_PROTO_PATCH_VERSION 2
#define RPC_PROTO_PATCH_VERSION 3
#ifdef __cplusplus
static_assert(GGML_OP_COUNT == 98, "GGML_OP_COUNT has changed - update RPC_PROTO_PATCH_VERSION");
static_assert(GGML_OP_COUNT == 101, "GGML_OP_COUNT has changed - update RPC_PROTO_PATCH_VERSION");
#endif
#define GGML_RPC_MAX_SERVERS 16

42
ggml/include/ggml.h
View File

@@ -571,6 +571,9 @@ extern "C" {
GGML_OP_SOLVE_TRI,
GGML_OP_GATED_DELTA_NET,
GGML_OP_LIGHTNING_INDEXER,
GGML_OP_DSV4_HC_COMB,
GGML_OP_DSV4_HC_PRE,
GGML_OP_DSV4_HC_POST,
GGML_OP_UNARY,
@@ -2598,6 +2601,45 @@ extern "C" {
struct ggml_tensor * weights,
struct ggml_tensor * mask);
// DeepSeek V4 hyper-connections (ref. https://arxiv.org/pdf/2512.24880)
// In short these operations are replacements for the original residual connection (x = transformer(x) + x)
// using a richer representation through streams.
//
// hc_comb: mixes [(2 + hc)*hc, n_tokens], scale [3], base [(2 + hc)*hc]
// -> [dst_hc, src_hc, n_tokens]
// logits[dst, src, t] = mixes[2*hc + dst + hc*src, t]*scale[2]
// + base[2*hc + dst + hc*src]
// Softmax over dst, add eps, normalize over src, then repeat normalization
// over dst followed by src for iterations 1 through n_iter - 1.
GGML_API struct ggml_tensor * ggml_dsv4_hc_comb(
struct ggml_context * ctx,
struct ggml_tensor * mixes,
struct ggml_tensor * scale,
struct ggml_tensor * base,
float eps,
int32_t n_iter);
// hc_pre: x [n_embd, hc, n_tokens], weights [hc, n_tokens] -> [n_embd, n_tokens]
// result[i, t] = sum_h x[i, h, t]*weights[h, t]
//
GGML_API struct ggml_tensor * ggml_dsv4_hc_pre(
struct ggml_context * ctx,
struct ggml_tensor * x,
struct ggml_tensor * weights);
// hc_post: x [n_embd, n_tokens], residual [n_embd, hc, n_tokens],
// post [hc, n_tokens], comb [dst_hc, src_hc, n_tokens]
// -> [n_embd, hc, n_tokens]
// result[i, dst, t] = x[i, t]*post[dst, t]
// + sum_src residual[i, src, t]*comb[dst, src, t]
//
GGML_API struct ggml_tensor * ggml_dsv4_hc_post(
struct ggml_context * ctx,
struct ggml_tensor * x,
struct ggml_tensor * residual,
struct ggml_tensor * post,
struct ggml_tensor * comb);
// custom operators
typedef void (*ggml_custom1_op_t)(struct ggml_tensor * dst , const struct ggml_tensor * a, int ith, int nth, void * userdata);