mirror of
https://github.com/dogkeeper886/ollama37.git
synced 2025-12-11 08:17:03 +00:00
IBM granite/granitemoe architecture support (#6760)
* fix(ext_server): Port llama.cpp sampling refactors to ext_server
This was a fairly large changeset. I closely followed the changes here:
df270ef745
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(server.cpp): Refactor server.cpp logging for llama.cpp overhaul
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* feat: Bump llama.cpp to the latest master with `granite` support
This does not yet have granite MoE support, but that can come in a
follow up PR
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(patches): Update all patches (except solar-pro) to work with bumped llama.cpp
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(solar): Update solar patch for llama.cpp bump
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* feat(llama.cpp): Bump llama.cpp for granitemoe support
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* feat(llama.cpp): Bump llama.cpp for granitemoe support
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(solar): Update the solar-pro patch for latest llama.cpp bump
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* feat(llama.cpp): Bump to the latest master of llama.cpp
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(patches): Update all patches for latest bump
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* feat(llama): Always run sync.sh from the right directory
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama/patches): Update llama patches
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* feat(llama)!: Rough sync with llama.cpp submodule
There are a number of changes that will need to be propagated to llama.go
before any of this works!
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama/patches): Add a patch and update for missing ggml-impl.h include
This include is where the ggml_cgraph struct is defined. It is included in
many of the .c files to define the forward declartion in ggml.h. It seems
that with the subset of code included here, the import was somehow lost (or
out-of-order) when building, so adding this include to llama.cpp fixes the
missing definition.
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama/sync): Add missing ggml-cpu-impl.h copy-over in sync.sh
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama): Add missing log.cpp
This was added as part of the logging overhaul done in llama.cpp
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama): Overhaul use of sampling module for llama.cpp changes
The changes here reflect the changes made in the big llama.cpp sampling PR
https://github.com/ggerganov/llama.cpp/pull/9294
The sampling functionality is now broken into the base interface
(llama_sampler) and the generation implementation (gpt_sampler). The
changes here reflect that. Since the sampling.h/sampling.cpp code uses c++
STL headers, the sampling_ext.[h|cpp] wrapper is maintained to allow go to
access a pure-C interface.
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama): Fix the impl of SampleTokenGreedy for new sampling
I don't think this method is currently used, so it could probably just be
removed so that all sampling goes through the GPT interface, but in the
interest of doing no harm, this should keep the method working as expected.
Branch: IBMGraniteArchitectureSupport
* fix(llama): Remove unused SampleTokenGreedy
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(sync): Remove bash-specific change to sync.sh
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* chore(gofumpt): Format on llama.go to pass linting
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llm): Fix missing <thread> include in ext_server
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama): Remove TODO about grammar_first
This feature was not used/needed previously so should be fine without
plumbing it through now.
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama): Better naming for sampling wrapper and args
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama): Fix patch 05 to use new wrapper api and re-sync
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* runner: Flush pending responses before returning
If there are any pending reponses (such as from potential stop
tokens) then we should send them back before ending the sequence.
Otherwise, we can be missing tokens at the end of a response.
Fixes #6707
* fix(llama/sampling): Use gpt_sampler with a forward declaration
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llama): Remove unnecessary patch for gguf impl header
This was caused by an earlier mistake in the embeddings patch that was
dereferencing the pointer instead of using the wrapper API.
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
* fix(llm): Remove use of deprecated --log-disable flag
Branch: IBMGraniteArchitectureSupport
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
---------
Signed-off-by: Gabe Goodhart <ghart@us.ibm.com>
This commit is contained in:
Gabe Goodhart
211
llama/ggml.h
vendored
211
llama/ggml.h
vendored
@@ -1,5 +1,5 @@
|
||||
/**
|
||||
* llama.cpp - commit 8962422b1c6f9b8b15f5aeaea42600bcc2d44177 - do not edit this file
|
||||
* llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
|
||||
*
|
||||
* MIT License
|
||||
*
|
||||
@@ -255,14 +255,16 @@
|
||||
#define GGML_MAX_PARAMS 2048
|
||||
#define GGML_MAX_CONTEXTS 64
|
||||
#define GGML_MAX_SRC 10
|
||||
#ifndef GGML_MAX_NAME
|
||||
#define GGML_MAX_NAME 64
|
||||
#define GGML_MAX_N_THREADS 512
|
||||
|
||||
#endif
|
||||
#define GGML_MAX_OP_PARAMS 64
|
||||
|
||||
#ifndef GGML_MAX_NAME
|
||||
# define GGML_MAX_NAME 64
|
||||
#endif
|
||||
|
||||
#define GGML_DEFAULT_N_THREADS 4
|
||||
#define GGML_DEFAULT_GRAPH_SIZE 2048
|
||||
|
||||
#if UINTPTR_MAX == 0xFFFFFFFF
|
||||
#define GGML_MEM_ALIGN 4
|
||||
#else
|
||||
@@ -285,21 +287,21 @@
|
||||
#define GGML_PAD(x, n) (((x) + (n) - 1) & ~((n) - 1))
|
||||
|
||||
#ifndef NDEBUG
|
||||
#define GGML_UNREACHABLE() do { fprintf(stderr, "statement should be unreachable\n"); abort(); } while(0)
|
||||
# define GGML_UNREACHABLE() do { fprintf(stderr, "statement should be unreachable\n"); abort(); } while(0)
|
||||
#elif defined(__GNUC__)
|
||||
#define GGML_UNREACHABLE() __builtin_unreachable()
|
||||
# define GGML_UNREACHABLE() __builtin_unreachable()
|
||||
#elif defined(_MSC_VER)
|
||||
#define GGML_UNREACHABLE() __assume(0)
|
||||
# define GGML_UNREACHABLE() __assume(0)
|
||||
#else
|
||||
#define GGML_UNREACHABLE() ((void) 0)
|
||||
# define GGML_UNREACHABLE() ((void) 0)
|
||||
#endif
|
||||
|
||||
#ifdef __cplusplus
|
||||
#define GGML_NORETURN [[noreturn]]
|
||||
# define GGML_NORETURN [[noreturn]]
|
||||
#elif defined(_MSC_VER)
|
||||
#define GGML_NORETURN __declspec(noreturn)
|
||||
# define GGML_NORETURN __declspec(noreturn)
|
||||
#else
|
||||
#define GGML_NORETURN _Noreturn
|
||||
# define GGML_NORETURN _Noreturn
|
||||
#endif
|
||||
|
||||
#define GGML_ABORT(...) ggml_abort(__FILE__, __LINE__, __VA_ARGS__)
|
||||
@@ -384,6 +386,7 @@ extern "C" {
|
||||
|
||||
struct ggml_object;
|
||||
struct ggml_context;
|
||||
struct ggml_cgraph;
|
||||
|
||||
// NOTE: always add types at the end of the enum to keep backward compatibility
|
||||
enum ggml_type {
|
||||
@@ -421,6 +424,8 @@ extern "C" {
|
||||
GGML_TYPE_Q4_0_4_4 = 31,
|
||||
GGML_TYPE_Q4_0_4_8 = 32,
|
||||
GGML_TYPE_Q4_0_8_8 = 33,
|
||||
GGML_TYPE_TQ1_0 = 34,
|
||||
GGML_TYPE_TQ2_0 = 35,
|
||||
GGML_TYPE_COUNT,
|
||||
};
|
||||
|
||||
@@ -557,6 +562,7 @@ extern "C" {
|
||||
|
||||
GGML_OP_CROSS_ENTROPY_LOSS,
|
||||
GGML_OP_CROSS_ENTROPY_LOSS_BACK,
|
||||
GGML_OP_OPT_STEP_ADAMW,
|
||||
|
||||
GGML_OP_COUNT,
|
||||
};
|
||||
@@ -587,35 +593,25 @@ extern "C" {
|
||||
};
|
||||
|
||||
enum ggml_log_level {
|
||||
GGML_LOG_LEVEL_ERROR = 2,
|
||||
GGML_LOG_LEVEL_WARN = 3,
|
||||
GGML_LOG_LEVEL_INFO = 4,
|
||||
GGML_LOG_LEVEL_DEBUG = 5
|
||||
GGML_LOG_LEVEL_NONE = 0,
|
||||
GGML_LOG_LEVEL_INFO = 1,
|
||||
GGML_LOG_LEVEL_WARN = 2,
|
||||
GGML_LOG_LEVEL_ERROR = 3,
|
||||
GGML_LOG_LEVEL_DEBUG = 4,
|
||||
GGML_LOG_LEVEL_CONT = 5, // continue previous log
|
||||
};
|
||||
|
||||
// this tensor...
|
||||
enum ggml_tensor_flag {
|
||||
GGML_TENSOR_FLAG_INPUT = 1,
|
||||
GGML_TENSOR_FLAG_OUTPUT = 2,
|
||||
GGML_TENSOR_FLAG_PARAM = 4,
|
||||
GGML_TENSOR_FLAG_INPUT = 1, // ...is an input for the GGML compute graph
|
||||
GGML_TENSOR_FLAG_OUTPUT = 2, // ...is an output for the GGML compute graph
|
||||
GGML_TENSOR_FLAG_PARAM = 4, // ...contains trainable parameters
|
||||
GGML_TENSOR_FLAG_LOSS = 8, // ...defines loss for numerical optimization (multiple loss tensors add up)
|
||||
};
|
||||
|
||||
// ggml object
|
||||
struct ggml_object {
|
||||
size_t offs;
|
||||
size_t size;
|
||||
|
||||
struct ggml_object * next;
|
||||
|
||||
enum ggml_object_type type;
|
||||
|
||||
char padding[4];
|
||||
};
|
||||
|
||||
static const size_t GGML_OBJECT_SIZE = sizeof(struct ggml_object);
|
||||
|
||||
// n-dimensional tensor
|
||||
struct ggml_tensor {
|
||||
enum ggml_type type;
|
||||
enum ggml_type type;
|
||||
|
||||
GGML_DEPRECATED(enum ggml_backend_type backend, "use the buffer type to find the storage location of the tensor");
|
||||
|
||||
@@ -679,7 +675,7 @@ extern "C" {
|
||||
|
||||
struct ggml_threadpool; // forward declaration, see ggml.c
|
||||
|
||||
typedef struct ggml_threadpool * ggml_threadpool_t;
|
||||
typedef struct ggml_threadpool * ggml_threadpool_t;
|
||||
|
||||
// the compute plan that needs to be prepared for ggml_graph_compute()
|
||||
// since https://github.com/ggerganov/ggml/issues/287
|
||||
@@ -695,35 +691,6 @@ extern "C" {
|
||||
void * abort_callback_data;
|
||||
};
|
||||
|
||||
enum ggml_cgraph_eval_order {
|
||||
GGML_CGRAPH_EVAL_ORDER_LEFT_TO_RIGHT = 0,
|
||||
GGML_CGRAPH_EVAL_ORDER_RIGHT_TO_LEFT,
|
||||
GGML_CGRAPH_EVAL_ORDER_COUNT
|
||||
};
|
||||
|
||||
typedef uint32_t ggml_bitset_t;
|
||||
|
||||
struct ggml_hash_set {
|
||||
size_t size;
|
||||
ggml_bitset_t * used;
|
||||
struct ggml_tensor ** keys;
|
||||
};
|
||||
|
||||
// computation graph
|
||||
struct ggml_cgraph {
|
||||
int size;
|
||||
int n_nodes;
|
||||
int n_leafs;
|
||||
|
||||
struct ggml_tensor ** nodes;
|
||||
struct ggml_tensor ** grads;
|
||||
struct ggml_tensor ** leafs;
|
||||
|
||||
struct ggml_hash_set visited_hash_set;
|
||||
|
||||
enum ggml_cgraph_eval_order order;
|
||||
};
|
||||
|
||||
// scratch buffer
|
||||
struct ggml_scratch {
|
||||
size_t offs;
|
||||
@@ -1296,7 +1263,7 @@ extern "C" {
|
||||
size_t nb1,
|
||||
size_t nb2,
|
||||
size_t nb3,
|
||||
size_t offset);
|
||||
size_t offset); // in bytes
|
||||
|
||||
// b -> view(a,offset,nb1,nb2,3), return view(a)
|
||||
GGML_API struct ggml_tensor * ggml_set_inplace(
|
||||
@@ -1306,19 +1273,19 @@ extern "C" {
|
||||
size_t nb1,
|
||||
size_t nb2,
|
||||
size_t nb3,
|
||||
size_t offset);
|
||||
size_t offset); // in bytes
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_set_1d(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
size_t offset);
|
||||
size_t offset); // in bytes
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_set_1d_inplace(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
size_t offset);
|
||||
size_t offset); // in bytes
|
||||
|
||||
// b -> view(a,offset,nb1,nb2,3), return modified a
|
||||
GGML_API struct ggml_tensor * ggml_set_2d(
|
||||
@@ -1326,7 +1293,7 @@ extern "C" {
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
size_t nb1,
|
||||
size_t offset);
|
||||
size_t offset); // in bytes
|
||||
|
||||
// b -> view(a,offset,nb1,nb2,3), return view(a)
|
||||
GGML_API struct ggml_tensor * ggml_set_2d_inplace(
|
||||
@@ -1334,7 +1301,7 @@ extern "C" {
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
size_t nb1,
|
||||
size_t offset);
|
||||
size_t offset); // in bytes
|
||||
|
||||
// a -> b, return view(b)
|
||||
GGML_API struct ggml_tensor * ggml_cpy(
|
||||
@@ -1469,14 +1436,14 @@ extern "C" {
|
||||
// supports 3D: a->ne[2] == b->ne[1]
|
||||
GGML_API struct ggml_tensor * ggml_get_rows(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
struct ggml_tensor * a, // data
|
||||
struct ggml_tensor * b); // row indices
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_get_rows_back(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
struct ggml_tensor * c);
|
||||
struct ggml_tensor * a, // gradients of ggml_get_rows result
|
||||
struct ggml_tensor * b, // row indices
|
||||
struct ggml_tensor * c); // data for ggml_get_rows, only used for its shape
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_diag(
|
||||
struct ggml_context * ctx,
|
||||
@@ -1627,9 +1594,9 @@ extern "C" {
|
||||
// a - dy
|
||||
GGML_API struct ggml_tensor * ggml_rope_back(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
struct ggml_tensor * c,
|
||||
struct ggml_tensor * a, // gradients of ggml_rope result
|
||||
struct ggml_tensor * b, // positions
|
||||
struct ggml_tensor * c, // freq factors
|
||||
int n_dims,
|
||||
int mode,
|
||||
int n_ctx_orig,
|
||||
@@ -2041,7 +2008,8 @@ extern "C" {
|
||||
typedef void (*ggml_custom2_op_t)(struct ggml_tensor * dst , const struct ggml_tensor * a, const struct ggml_tensor * b, int ith, int nth, void * userdata);
|
||||
typedef void (*ggml_custom3_op_t)(struct ggml_tensor * dst , const struct ggml_tensor * a, const struct ggml_tensor * b, const struct ggml_tensor * c, int ith, int nth, void * userdata);
|
||||
|
||||
#define GGML_N_TASKS_MAX -1
|
||||
#define GGML_N_TASKS_MAX (-1)
|
||||
// n_tasks == GGML_N_TASKS_MAX means to use max number of tasks
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_map_custom1(
|
||||
struct ggml_context * ctx,
|
||||
@@ -2094,48 +2062,75 @@ extern "C" {
|
||||
// loss function
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_cross_entropy_loss(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a, // logits
|
||||
struct ggml_tensor * b); // labels
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_cross_entropy_loss_back(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
struct ggml_tensor * c);
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a, // logits
|
||||
struct ggml_tensor * b, // labels
|
||||
struct ggml_tensor * c); // gradients of cross_entropy_loss result
|
||||
|
||||
// AdamW optimizer step
|
||||
// Paper: https://arxiv.org/pdf/1711.05101v3.pdf
|
||||
// PyTorch: https://pytorch.org/docs/stable/generated/torch.optim.AdamW.html
|
||||
GGML_API struct ggml_tensor * ggml_opt_step_adamw(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * grad,
|
||||
float alpha,
|
||||
float beta1,
|
||||
float beta2,
|
||||
float eps,
|
||||
float wd); // weight decay
|
||||
|
||||
//
|
||||
// automatic differentiation
|
||||
//
|
||||
|
||||
GGML_API void ggml_set_param(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * tensor);
|
||||
|
||||
GGML_API void ggml_set_param(struct ggml_context * ctx, struct ggml_tensor * tensor);
|
||||
GGML_API void ggml_set_loss(struct ggml_tensor * tensor);
|
||||
|
||||
GGML_API void ggml_build_forward_expand (struct ggml_cgraph * cgraph, struct ggml_tensor * tensor);
|
||||
GGML_API void ggml_build_backward_expand(struct ggml_context * ctx, struct ggml_cgraph * gf, struct ggml_cgraph * gb, bool keep);
|
||||
GGML_API void ggml_build_backward_expand(struct ggml_context * ctx, struct ggml_cgraph * gf, struct ggml_cgraph * gb, bool accumulate);
|
||||
|
||||
GGML_API void ggml_build_opt_adamw(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_cgraph * gf,
|
||||
struct ggml_cgraph * gb,
|
||||
float alpha,
|
||||
float beta1,
|
||||
float beta2,
|
||||
float eps,
|
||||
float wd); // weight decay
|
||||
|
||||
// graph allocation in a context
|
||||
GGML_API struct ggml_cgraph * ggml_new_graph (struct ggml_context * ctx); // size = GGML_DEFAULT_GRAPH_SIZE, grads = false
|
||||
GGML_API struct ggml_cgraph * ggml_new_graph_custom (struct ggml_context * ctx, size_t size, bool grads);
|
||||
GGML_API struct ggml_cgraph * ggml_graph_dup (struct ggml_context * ctx, struct ggml_cgraph * cgraph);
|
||||
GGML_API struct ggml_cgraph ggml_graph_view (struct ggml_cgraph * cgraph, int i0, int i1);
|
||||
GGML_API void ggml_graph_cpy (struct ggml_cgraph * src, struct ggml_cgraph * dst);
|
||||
GGML_API void ggml_graph_reset (struct ggml_cgraph * cgraph); // zero grads
|
||||
GGML_API void ggml_graph_clear (struct ggml_cgraph * cgraph);
|
||||
GGML_API struct ggml_cgraph * ggml_new_graph (struct ggml_context * ctx); // size = GGML_DEFAULT_GRAPH_SIZE, grads = false
|
||||
GGML_API struct ggml_cgraph * ggml_new_graph_custom(struct ggml_context * ctx, size_t size, bool grads);
|
||||
GGML_API struct ggml_cgraph * ggml_graph_dup (struct ggml_context * ctx, struct ggml_cgraph * cgraph);
|
||||
GGML_API void ggml_graph_cpy (struct ggml_cgraph * src, struct ggml_cgraph * dst);
|
||||
GGML_API void ggml_graph_reset (struct ggml_cgraph * cgraph); // set regular grads + optimizer momenta to 0, set loss grad to 1
|
||||
GGML_API void ggml_graph_clear (struct ggml_cgraph * cgraph);
|
||||
|
||||
GGML_API int ggml_graph_size (struct ggml_cgraph * cgraph);
|
||||
GGML_API struct ggml_tensor * ggml_graph_node (struct ggml_cgraph * cgraph, int i); // if i < 0, returns nodes[n_nodes + i]
|
||||
GGML_API struct ggml_tensor ** ggml_graph_nodes (struct ggml_cgraph * cgraph);
|
||||
GGML_API int ggml_graph_n_nodes(struct ggml_cgraph * cgraph);
|
||||
|
||||
GGML_API void ggml_graph_add_node(struct ggml_cgraph * cgraph, struct ggml_tensor * tensor);
|
||||
|
||||
GGML_API size_t ggml_graph_overhead(void);
|
||||
GGML_API size_t ggml_graph_overhead_custom(size_t size, bool grads);
|
||||
|
||||
GGML_API struct ggml_threadpool_params ggml_threadpool_params_default(int n_threads);
|
||||
GGML_API void ggml_threadpool_params_init (struct ggml_threadpool_params *p, int n_threads);
|
||||
GGML_API bool ggml_threadpool_params_match (const struct ggml_threadpool_params *p0, const struct ggml_threadpool_params *p1);
|
||||
GGML_API struct ggml_threadpool* ggml_threadpool_new (struct ggml_threadpool_params * params);
|
||||
GGML_API void ggml_threadpool_free (struct ggml_threadpool * threadpool);
|
||||
GGML_API int ggml_threadpool_get_n_threads(struct ggml_threadpool * threadpool);
|
||||
GGML_API void ggml_threadpool_pause (struct ggml_threadpool * threadpool);
|
||||
GGML_API void ggml_threadpool_resume (struct ggml_threadpool * threadpool);
|
||||
GGML_API struct ggml_threadpool_params ggml_threadpool_params_default(int n_threads);
|
||||
GGML_API void ggml_threadpool_params_init (struct ggml_threadpool_params * p, int n_threads);
|
||||
GGML_API bool ggml_threadpool_params_match (const struct ggml_threadpool_params * p0, const struct ggml_threadpool_params * p1);
|
||||
GGML_API struct ggml_threadpool * ggml_threadpool_new (struct ggml_threadpool_params * params);
|
||||
GGML_API void ggml_threadpool_free (struct ggml_threadpool * threadpool);
|
||||
GGML_API int ggml_threadpool_get_n_threads(struct ggml_threadpool * threadpool);
|
||||
GGML_API void ggml_threadpool_pause (struct ggml_threadpool * threadpool);
|
||||
GGML_API void ggml_threadpool_resume (struct ggml_threadpool * threadpool);
|
||||
|
||||
// ggml_graph_plan() has to be called before ggml_graph_compute()
|
||||
// when plan.work_size > 0, caller must allocate memory for plan.work_data
|
||||
@@ -2533,6 +2528,7 @@ extern "C" {
|
||||
GGML_API int ggml_cpu_has_gpublas (void);
|
||||
GGML_API int ggml_cpu_has_sse3 (void);
|
||||
GGML_API int ggml_cpu_has_ssse3 (void);
|
||||
GGML_API int ggml_cpu_has_riscv_v (void);
|
||||
GGML_API int ggml_cpu_has_sycl (void);
|
||||
GGML_API int ggml_cpu_has_rpc (void);
|
||||
GGML_API int ggml_cpu_has_vsx (void);
|
||||
@@ -2540,6 +2536,9 @@ extern "C" {
|
||||
GGML_API int ggml_cpu_has_cann (void);
|
||||
GGML_API int ggml_cpu_has_llamafile (void);
|
||||
|
||||
// get the sve vector length in bytes
|
||||
GGML_API int ggml_cpu_get_sve_cnt(void);
|
||||
|
||||
//
|
||||
// Internal types and functions exposed for tests and benchmarks
|
||||
//
|
||||
|
||||
Reference in New Issue
Block a user