tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-08-17 13:40:55 -04:00
Code Issues Packages Projects Releases Wiki Activity

llama : add gpt-oss (#15091)

Browse Source 
* oai moe

* compat with new checkpoint

* add attn sink impl

* add rope scaling yarn

* logits match with latest transformers code

* wip chat template

* rm trailing space

* use ggml_scale_bias

* rm redundant is_swa_all

* convert interleaved gate_up

* graph : fix activation function to match reference (#7)

* vocab : handle o200k_harmony special tokens

* ggml : add attention sinks support (#1)

* llama : add attn sinks

* ggml : add attn sinks

* cuda : add attn sinks

* vulkan : add support for sinks in softmax

remove unnecessary return

* ggml : add fused swiglu_oai op (#11)

* ggml : add fused swiglu_oai op

* Update ggml/src/ggml-cpu/ops.cpp

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

* update CUDA impl

* cont : metal impl

* add vulkan impl

* test-backend-ops : more test cases, clean up

* llama : remove unfused impl

* remove extra lines

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>

---------

Co-authored-by: slaren <slarengh@gmail.com>

* repack mxfp4 upon conversion

* clean up a bit

* enable thinking

* add quick hack to render only some special tokens

* fix bf16 conversion

* remove vocab hack

* webui ok

* support chat parsing for gpt-oss

* fix webui

* direct mapping mxfp4, FINALLY

* force using mxfp4

* properly use lazy tensor

* ggml : add mxfp4

ggml : use e8m0 conversion instead of powf

Co-authored-by: Diego Devesa <slarengh@gmail.com>

change kvalues_mxfp4 table to match e2m1 (#6)

metal : remove quantization for now (not used)

cuda : fix disabled CUDA graphs due to ffn moe bias

vulkan : add support for mxfp4

cont : add cm2 dequant

* ggml : add ggml_add_id (#13)

* ggml : add ggml_add_id

* add cuda impl

* llama : add weight support check for add_id

* perf opt

* add vulkan impl

* rename cuda files

* add metal impl

* allow in-place ggml_add_id

* llama : keep biases on CPU with --cpu-moe

* llama : fix compile error

ggml-ci

* cuda : add fallback for __nv_cvt_e8m0_to_bf16raw

ggml-ci

* cleanup

ggml-ci

* sycl : fix supports_op for MXFP4

ggml-ci

* fix Unknown reasoning format

* ggml-cpu : fix AVX build

ggml-ci

* fix hip build

ggml-ci

* cuda : add mxfp4 dequantization support for cuBLAS

ggml-ci

* ggml-cpu : fix mxfp4 fallback definitions for some architectures

ggml-ci

* cuda : fix version required for __nv_cvt_e8m0_to_bf16raw

---------

Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
Co-authored-by: slaren <slarengh@gmail.com>
This commit is contained in:
Georgi Gerganov
2025-08-05 22:10:36 +03:00
committed by GitHub
parent f324a3b715
commit fd1234cb46
83 changed files with 2942 additions and 227 deletions
Download Patch File Download Diff File Expand all files Collapse all files

38
ggml/include/ggml.h
View File

@@ -304,6 +304,16 @@
GGML_TENSOR_LOCALS(int64_t, ne, dst, ne) \
GGML_TENSOR_LOCALS(size_t, nb, dst, nb)
#define GGML_TENSOR_TERNARY_OP_LOCALS \
GGML_TENSOR_LOCALS(int64_t, ne0, src0, ne) \
GGML_TENSOR_LOCALS(size_t, nb0, src0, nb) \
GGML_TENSOR_LOCALS(int64_t, ne1, src1, ne) \
GGML_TENSOR_LOCALS(size_t, nb1, src1, nb) \
GGML_TENSOR_LOCALS(int64_t, ne2, src2, ne) \
GGML_TENSOR_LOCALS(size_t, nb2, src2, nb) \
GGML_TENSOR_LOCALS(int64_t, ne, dst, ne) \
GGML_TENSOR_LOCALS(size_t, nb, dst, nb)
#define GGML_TENSOR_BINARY_OP_LOCALS01 \
GGML_TENSOR_LOCALS(int64_t, ne0, src0, ne) \
GGML_TENSOR_LOCALS(size_t, nb0, src0, nb) \
@@ -395,7 +405,8 @@ extern "C" {
// GGML_TYPE_IQ4_NL_4_4 = 36,
// GGML_TYPE_IQ4_NL_4_8 = 37,
// GGML_TYPE_IQ4_NL_8_8 = 38,
GGML_TYPE_COUNT = 39,
GGML_TYPE_MXFP4 = 39, // MXFP4 (1 block)
GGML_TYPE_COUNT = 40,
};
// precision
@@ -430,6 +441,7 @@ extern "C" {
GGML_FTYPE_MOSTLY_IQ4_XS = 22, // except 1d tensors
GGML_FTYPE_MOSTLY_IQ1_M = 23, // except 1d tensors
GGML_FTYPE_MOSTLY_BF16 = 24, // except 1d tensors
GGML_FTYPE_MOSTLY_MXFP4 = 25, // except 1d tensors
};
// available tensor operations:
@@ -438,6 +450,7 @@ extern "C" {
GGML_OP_DUP,
GGML_OP_ADD,
GGML_OP_ADD_ID,
GGML_OP_ADD1,
GGML_OP_ACC,
GGML_OP_SUB,
@@ -557,6 +570,7 @@ extern "C" {
GGML_GLU_OP_REGLU,
GGML_GLU_OP_GEGLU,
GGML_GLU_OP_SWIGLU,
GGML_GLU_OP_SWIGLU_OAI,
GGML_GLU_OP_GEGLU_ERF,
GGML_GLU_OP_GEGLU_QUICK,
@@ -831,6 +845,13 @@ extern "C" {
struct ggml_tensor * b,
enum ggml_type type);
// dst[i0, i1, i2] = a[i0, i1, i2] + b[i0, ids[i1, i2]]
GGML_API struct ggml_tensor * ggml_add_id(
struct ggml_context * ctx,
struct ggml_tensor * a,
struct ggml_tensor * b,
struct ggml_tensor * ids);
GGML_API struct ggml_tensor * ggml_add1(
struct ggml_context * ctx,
struct ggml_tensor * a,
@@ -1198,6 +1219,13 @@ extern "C" {
struct ggml_tensor * a,
struct ggml_tensor * b);
GGML_API struct ggml_tensor * ggml_swiglu_oai(
struct ggml_context * ctx,
struct ggml_tensor * a,
struct ggml_tensor * b,
float alpha,
float limit);
// normalize along rows
GGML_API struct ggml_tensor * ggml_norm(
struct ggml_context * ctx,
@@ -1570,6 +1598,10 @@ extern "C" {
float scale,
float max_bias);
GGML_API void ggml_soft_max_add_sinks(
struct ggml_tensor * a,
struct ggml_tensor * sinks);
GGML_API struct ggml_tensor * ggml_soft_max_ext_back(
struct ggml_context * ctx,
struct ggml_tensor * a,
@@ -2052,6 +2084,10 @@ extern "C" {
GGML_API enum ggml_prec ggml_flash_attn_ext_get_prec(
const struct ggml_tensor * a);
GGML_API void ggml_flash_attn_ext_add_sinks(
struct ggml_tensor * a,
struct ggml_tensor * sinks);
// TODO: needs to be adapted to ggml_flash_attn_ext
GGML_API struct ggml_tensor * ggml_flash_attn_back(
struct ggml_context * ctx,