root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-27 03:44:35 +00:00
Code Issues Actions 4 Packages Projects Releases Wiki Activity

allocators wip

Browse Source 
renamed ggml_backend functions
changed ggml_buffer and ggml_backend to always be used as pointers
rename ggml_tensor::params -> op_params
This commit is contained in:
slaren 2023-07-17 19:03:51 +02:00
parent 1102ff56db
commit 295f85654a
8 changed files with 640 additions and 365 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
examples/common.cpp
View File

@ -327,24 +327,24 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
params.n_gpu_layers = std::stoi(argv[i]); params.n_gpu_layers = std::stoi(argv[i]);
#else #else
fprintf(stderr, "warning: not compiled with GPU offload support, --n-gpu-layers option will be ignored\n"); fprintf(stderr, "warning: not compiled with GPU offload support, --n-gpu-layers option will be ignored\n");
fprintf(stderr, "warning: see main README.md for information on enabling GPU BLAS support\n"); fprintf(stderr, "warning: see main README.md for information on enabling GPU support\n");
#endif #endif
} else if (arg == "--main-gpu" || arg == "-mg") { } else if (arg == "--main-gpu" || arg == "-mg") {
if (++i >= argc) { if (++i >= argc) {
invalid_param = true; invalid_param = true;
break; break;
} }
#ifdef GGML_USE_CUBLAS #ifdef GGML_USE_CUDA
params.main_gpu = std::stoi(argv[i]); params.main_gpu = std::stoi(argv[i]);
#else #else
fprintf(stderr, "warning: llama.cpp was compiled without cuBLAS. It is not possible to set a main GPU.\n"); fprintf(stderr, "warning: llama.cpp was compiled without CUDA. It is not possible to set a main GPU.\n");
#endif #endif
} else if (arg == "--tensor-split" || arg == "-ts") { } else if (arg == "--tensor-split" || arg == "-ts") {
if (++i >= argc) { if (++i >= argc) {
invalid_param = true; invalid_param = true;
break; break;
} }
#ifdef GGML_USE_CUBLAS #ifdef GGML_USE_CUDA
std::string arg_next = argv[i]; std::string arg_next = argv[i];
// split string by , and / // split string by , and /
@ -361,14 +361,14 @@ bool gpt_params_parse(int argc, char ** argv, gpt_params & params) {
} }
} }
#else #else
fprintf(stderr, "warning: llama.cpp was compiled without cuBLAS. It is not possible to set a tensor split.\n"); fprintf(stderr, "warning: llama.cpp was compiled without CUDA. It is not possible to set a tensor split.\n");
#endif // GGML_USE_CUBLAS #endif // GGML_USE_CUDA
} else if (arg == "--low-vram" || arg == "-lv") { } else if (arg == "--low-vram" || arg == "-lv") {
#ifdef GGML_USE_CUBLAS #ifdef GGML_USE_CUDA
params.low_vram = true; params.low_vram = true;
#else #else
fprintf(stderr, "warning: llama.cpp was compiled without cuBLAS. It is not possible to set lower vram usage.\n"); fprintf(stderr, "warning: llama.cpp was compiled without CUDA. It is not possible to set lower vram usage.\n");
#endif // GGML_USE_CUBLAS #endif // GGML_USE_CUDA
} else if (arg == "--no-mmap") { } else if (arg == "--no-mmap") {
params.use_mmap = false; params.use_mmap = false;
} else if (arg == "--mtest") { } else if (arg == "--mtest") {

459
ggml-backend.c
View File

@ -7,22 +7,114 @@
#define UNUSED(x) (void)(x) #define UNUSED(x) (void)(x)
// backend buffer // allocator
struct ggml_buffer ggml_backend_alloc_buffer(struct ggml_backend * backend, size_t size, size_t max_tensors) { static size_t aligned_offset(const void * buffer, size_t offset, size_t alignment) {
struct ggml_buffer buffer; assert(alignment && !(alignment & (alignment - 1))); // power of 2
buffer.mem_size = ggml_tensor_overhead() * max_tensors; size_t align = (alignment - (((uintptr_t)buffer + offset) % alignment)) % alignment;
buffer.mem_buffer = malloc(buffer.mem_size); return offset + align;
buffer.backend = backend; }
static inline size_t ggml_backend_buffer_get_alloc_size(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor) { return alloc->interface.get_alloc_size(alloc, tensor); }
static inline void ggml_backend_buffer_init_tensor(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor) { alloc->interface.init_tensor(alloc, tensor); }
void ggml_backend_buffer_free(struct ggml_backend_buffer * alloc) {
alloc->interface.free_buffer(alloc);
free(alloc);
}
// backend buffer allocator - simple
struct ggml_allocator_simple_context {
void * data;
size_t size;
size_t offset;
size_t alignment;
};
static void ggml_allocator_simple_free_buffer(struct ggml_backend_buffer * alloc) {
struct ggml_allocator_simple_context * context = (struct ggml_allocator_simple_context *)alloc->context;
free(context);
}
static void ggml_allocator_simple_alloc_tensor(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor) {
struct ggml_allocator_simple_context * context = (struct ggml_allocator_simple_context *)alloc->context;
size_t size = ggml_backend_buffer_get_alloc_size(alloc, tensor);
if (context->offset + size > context->size) {
fprintf(stderr, "%s: not enough space in the buffer (needed %zu, available %zu)\n",
__func__, size, context->size - context->offset);
GGML_ASSERT(!"not enough space in the buffer");
return;
}
void * ptr = (char*)context->data + context->offset;
context->offset = aligned_offset(context->data, context->offset + size, context->alignment);
tensor->data = ptr;
if (alloc->interface.init_tensor) {
alloc->interface.init_tensor(alloc, tensor);
}
}
static void ggml_allocator_simple_free_tensor(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor) {
GGML_ASSERT(!"ggml_simple_allocator cannot free individual tensors");
UNUSED(alloc);
UNUSED(tensor);
}
static void ggml_allocator_simple_reset(struct ggml_backend_buffer * alloc) {
struct ggml_allocator_simple_context * context = (struct ggml_allocator_simple_context *)alloc->context;
context->offset = aligned_offset(context->data, 0, context->alignment);
}
size_t ggml_allocator_simple_get_alloc_size(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor) {
return ggml_nbytes(tensor);
UNUSED(alloc);
}
static const struct ggml_backend_buffer_interface ggml_allocator_simple_interface = {
/* .free_buffer = */ ggml_allocator_simple_free_buffer,
/* .alloc_tensor = */ ggml_allocator_simple_alloc_tensor,
/* .free_tensor = */ ggml_allocator_simple_free_tensor,
/* .reset = */ ggml_allocator_simple_reset,
/* .get_alloc_size = */ ggml_allocator_simple_get_alloc_size,
/* .init_tensor = */ NULL,
/* .free_data = */ NULL,
};
struct ggml_backend_buffer * ggml_allocator_simple_init(void * data, size_t size, size_t alignment) {
struct ggml_allocator_simple_context * ctx = malloc(sizeof(struct ggml_allocator_simple_context));
ctx->data = data;
ctx->size = size;
ctx->offset = aligned_offset(data, 0, alignment);
ctx->alignment = alignment;
struct ggml_backend_buffer * allocator = malloc(sizeof(struct ggml_backend_buffer));
*allocator = (struct ggml_backend_buffer){
/* .interface = */ ggml_allocator_simple_interface,
/* .context = */ ctx,
/* .backend_data = */ NULL,
};
return allocator;
}
// buffer
struct ggml_buffer * ggml_buffer_alloc(struct ggml_backend * backend, size_t size, size_t max_tensors) {
struct ggml_buffer * buffer = malloc(sizeof(struct ggml_buffer));
buffer->mem_size = ggml_tensor_overhead() * max_tensors;
buffer->mem_buffer = malloc(buffer->mem_size);
buffer->backend = backend;
size += 128 * max_tensors; // alignment overhead size += 128 * max_tensors; // alignment overhead
buffer.backend_buffer = backend->interface->alloc_buffer(backend->context, size); buffer->backend_buffer = backend->interface.alloc_buffer(backend, size);
return buffer; return buffer;
} }
void ggml_backend_free_buffer(struct ggml_buffer * buffer) { void ggml_buffer_free(struct ggml_buffer * buffer) {
struct ggml_backend * backend = buffer->backend; ggml_backend_buffer_free(buffer->backend_buffer);
backend->interface->free_buffer(backend->context, buffer->backend_buffer);
free(buffer->mem_buffer); free(buffer->mem_buffer);
free(buffer);
} }
// backend copy // backend copy
@ -42,7 +134,7 @@ static bool ggml_are_same_layout(const struct ggml_tensor * a, const struct ggml
return true; return true;
} }
void ggml_backend_cpy_tensor(struct ggml_tensor * dst, struct ggml_tensor * src) { void ggml_backend_tensor_copy(struct ggml_tensor * src, struct ggml_tensor * dst) {
//printf("src: %s ne: [%d %d %d %d] nb: [%d %d %d %d]\n", src->name, (int)src->ne[0], (int)src->ne[1], (int)src->ne[2], (int)src->ne[3], (int)src->nb[0], (int)src->nb[1], (int)src->nb[2], (int)src->nb[3]); //printf("src: %s ne: [%d %d %d %d] nb: [%d %d %d %d]\n", src->name, (int)src->ne[0], (int)src->ne[1], (int)src->ne[2], (int)src->ne[3], (int)src->nb[0], (int)src->nb[1], (int)src->nb[2], (int)src->nb[3]);
//printf("dst: %s ne: [%d %d %d %d] nb: [%d %d %d %d]\n", dst->name, (int)dst->ne[0], (int)dst->ne[1], (int)dst->ne[2], (int)dst->ne[3], (int)dst->nb[0], (int)dst->nb[1], (int)dst->nb[2], (int)dst->nb[3]); //printf("dst: %s ne: [%d %d %d %d] nb: [%d %d %d %d]\n", dst->name, (int)dst->ne[0], (int)dst->ne[1], (int)dst->ne[2], (int)dst->ne[3], (int)dst->nb[0], (int)dst->nb[1], (int)dst->nb[2], (int)dst->nb[3]);
GGML_ASSERT(ggml_are_same_layout(src, dst) && "cannot copy tensors with different layouts"); GGML_ASSERT(ggml_are_same_layout(src, dst) && "cannot copy tensors with different layouts");
@ -53,17 +145,17 @@ void ggml_backend_cpy_tensor(struct ggml_tensor * dst, struct ggml_tensor * src)
return; return;
} }
if (dst->backend->interface->cpy_tensor_from != NULL) { if (dst->backend->interface.cpy_tensor_from != NULL) {
dst->backend->interface->cpy_tensor_from(dst->backend->context, src, dst); dst->backend->interface.cpy_tensor_from(dst->backend->context, src, dst);
} else if (src->backend->interface->cpy_tensor_to != NULL) { } else if (src->backend->interface.cpy_tensor_to != NULL) {
src->backend->interface->cpy_tensor_to(src->backend->context, src, dst); src->backend->interface.cpy_tensor_to(src->backend->context, src, dst);
} else { } else {
// not ideal, but shouldn't be hit when copying from/to CPU // not ideal, but shouldn't be hit when copying from/to CPU
// TODO: print a performance warning in debug builds // TODO: print a performance warning in debug builds
size_t nbytes = ggml_nbytes(src); size_t nbytes = ggml_nbytes(src);
void * data = malloc(nbytes); void * data = malloc(nbytes);
ggml_backend_get_tensor(src, data, 0, nbytes); ggml_backend_tensor_get(src, data, 0, nbytes);
ggml_backend_set_tensor(dst, data, 0, nbytes); ggml_backend_tensor_set(dst, data, 0, nbytes);
free(data); free(data);
} }
} }
@ -76,105 +168,70 @@ struct ggml_backend_cpu_context {
size_t work_size; size_t work_size;
}; };
static const char * ggml_backend_cpu_name(ggml_backend_context_t ctx) { static const char * ggml_backend_cpu_name(struct ggml_backend * backend) {
return "CPU"; return "CPU";
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_free_context(ggml_backend_context_t ctx) { static void ggml_backend_cpu_free(struct ggml_backend * backend) {
struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)ctx; struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)backend->context;
free(cpu_ctx->work_data); free(cpu_ctx->work_data);
free(ctx); free(cpu_ctx);
free(backend);
} }
struct cpu_backend_buffer {
void * data;
size_t offset;
size_t size;
};
static const size_t TENSOR_ALIGNMENT = 64; // should be enough for AVX 512 static const size_t TENSOR_ALIGNMENT = 64; // should be enough for AVX 512
static size_t aligned_offset(const void * buffer, size_t offset, size_t alignment) { static void ggml_backend_cpu_free_buffer(struct ggml_backend_buffer * alloc) {
assert(alignment && !(alignment & (alignment - 1))); // power of 2 free(alloc->backend_data);
size_t align = (alignment - (((uintptr_t)buffer + offset) % alignment)) % alignment;
return offset + align;
} }
static ggml_backend_buffer_t ggml_backend_cpu_alloc_buffer(ggml_backend_context_t ctx, size_t size) { static struct ggml_backend_buffer * ggml_backend_cpu_alloc_buffer(struct ggml_backend * backend, size_t size) {
struct cpu_backend_buffer * buffer = malloc(sizeof(struct cpu_backend_buffer)); void * data = malloc(size);
buffer->data = malloc(size);
buffer->offset = aligned_offset(buffer->data, 0, TENSOR_ALIGNMENT); struct ggml_backend_buffer * buffer = ggml_allocator_simple_init(data, size, TENSOR_ALIGNMENT);
buffer->size = size; buffer->interface.free_data = ggml_backend_cpu_free_buffer;
buffer->backend_data = data;
return buffer; return buffer;
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_free_buffer(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer) { static void ggml_backend_cpu_set_tensor_async(struct ggml_backend * backend, struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
struct cpu_backend_buffer * cpu_buffer = (struct cpu_backend_buffer *)buffer;
free(cpu_buffer->data);
free(cpu_buffer);
UNUSED(ctx);
}
static void ggml_backend_cpu_reset_buffer(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer) {
struct cpu_backend_buffer * cpu_buffer = (struct cpu_backend_buffer *)buffer;
cpu_buffer->offset = aligned_offset(cpu_buffer->data, 0, TENSOR_ALIGNMENT);
UNUSED(ctx);
}
static void ggml_backend_cpu_alloc_tensor(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer, struct ggml_tensor * tensor) {
struct cpu_backend_buffer * cpu_buffer = (struct cpu_backend_buffer *)buffer;
// TODO: make this error recoverable
if (cpu_buffer->offset + ggml_nbytes(tensor) > cpu_buffer->size) {
fprintf(stderr, "%s: not enough space in the buffer (needed %zu, available %zu)\n",
__func__, ggml_nbytes(tensor), cpu_buffer->size - cpu_buffer->offset);
GGML_ASSERT(false);
}
tensor->data = (char*)cpu_buffer->data + cpu_buffer->offset;
cpu_buffer->offset = aligned_offset(cpu_buffer->data, cpu_buffer->offset + ggml_nbytes(tensor), TENSOR_ALIGNMENT);
UNUSED(ctx);
}
static void ggml_backend_cpu_set_tensor_async(ggml_backend_context_t ctx, struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor write out of bounds"); GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor write out of bounds");
GGML_ASSERT(tensor->data != NULL && "tensor not allocated"); GGML_ASSERT(tensor->data != NULL && "tensor not allocated");
memcpy((char *)tensor->data + offset, data, size); memcpy((char *)tensor->data + offset, data, size);
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_get_tensor_async(ggml_backend_context_t ctx, const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) { static void ggml_backend_cpu_get_tensor_async(struct ggml_backend * backend, const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) {
GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor read out of bounds"); GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor read out of bounds");
GGML_ASSERT(tensor->data != NULL && "tensor not allocated"); GGML_ASSERT(tensor->data != NULL && "tensor not allocated");
memcpy(data, (const char *)tensor->data + offset, size); memcpy(data, (const char *)tensor->data + offset, size);
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_synchronize(ggml_backend_context_t ctx) { static void ggml_backend_cpu_synchronize(struct ggml_backend * backend) {
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_cpy_tensor_from(ggml_backend_context_t ctx, struct ggml_tensor * src, struct ggml_tensor * dst) { static void ggml_backend_cpu_cpy_tensor_from(struct ggml_backend * backend, struct ggml_tensor * src, struct ggml_tensor * dst) {
ggml_backend_get_tensor(src, dst->data, 0, ggml_nbytes(src)); ggml_backend_tensor_get(src, dst->data, 0, ggml_nbytes(src));
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_cpy_tensor_to(ggml_backend_context_t ctx, struct ggml_tensor * src, struct ggml_tensor * dst) { static void ggml_backend_cpu_cpy_tensor_to(struct ggml_backend * backend, struct ggml_tensor * src, struct ggml_tensor * dst) {
ggml_backend_set_tensor_async(dst, src->data, 0, ggml_nbytes(src)); // for a backend such as CUDA that can queue async calls, it is ok to do this asynchronously, but it may not be the case for other backends
ggml_backend_tensor_set_async(dst, src->data, 0, ggml_nbytes(src));
UNUSED(ctx); UNUSED(backend);
} }
struct ggml_backend_cpu_plan { struct ggml_backend_cpu_plan {
@ -182,8 +239,8 @@ struct ggml_backend_cpu_plan {
struct ggml_cgraph cgraph; struct ggml_cgraph cgraph;
}; };
static ggml_graph_plan_t ggml_backend_cpu_graph_plan_create(ggml_backend_context_t ctx, struct ggml_cgraph * cgraph) { static ggml_graph_plan_t ggml_backend_cpu_graph_plan_create(struct ggml_backend * backend, struct ggml_cgraph * cgraph) {
struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)ctx; struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)backend->context;
struct ggml_backend_cpu_plan * cpu_plan = malloc(sizeof(struct ggml_backend_cpu_plan)); struct ggml_backend_cpu_plan * cpu_plan = malloc(sizeof(struct ggml_backend_cpu_plan));
@ -197,25 +254,25 @@ static ggml_graph_plan_t ggml_backend_cpu_graph_plan_create(ggml_backend_context
return cpu_plan; return cpu_plan;
} }
static void ggml_backend_cpu_graph_plan_free(ggml_backend_context_t ctx, ggml_graph_plan_t plan) { static void ggml_backend_cpu_graph_plan_free(struct ggml_backend * backend, ggml_graph_plan_t plan) {
struct ggml_backend_cpu_plan * cpu_plan = (struct ggml_backend_cpu_plan *)plan; struct ggml_backend_cpu_plan * cpu_plan = (struct ggml_backend_cpu_plan *)plan;
free(cpu_plan->cplan.work_data); free(cpu_plan->cplan.work_data);
free(cpu_plan); free(cpu_plan);
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_graph_plan_compute(ggml_backend_context_t ctx, ggml_graph_plan_t plan) { static void ggml_backend_cpu_graph_plan_compute(struct ggml_backend * backend, ggml_graph_plan_t plan) {
struct ggml_backend_cpu_plan * cpu_plan = (struct ggml_backend_cpu_plan *)plan; struct ggml_backend_cpu_plan * cpu_plan = (struct ggml_backend_cpu_plan *)plan;
ggml_graph_compute(&cpu_plan->cgraph, &cpu_plan->cplan); ggml_graph_compute(&cpu_plan->cgraph, &cpu_plan->cplan);
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cpu_graph_compute(ggml_backend_context_t ctx, struct ggml_cgraph * cgraph) { static void ggml_backend_cpu_graph_compute(struct ggml_backend * backend, struct ggml_cgraph * cgraph) {
struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)ctx; struct ggml_backend_cpu_context * cpu_ctx = (struct ggml_backend_cpu_context *)backend->context;
struct ggml_cplan cplan = ggml_graph_plan(cgraph, cpu_ctx->n_threads); struct ggml_cplan cplan = ggml_graph_plan(cgraph, cpu_ctx->n_threads);
@ -232,11 +289,8 @@ static void ggml_backend_cpu_graph_compute(ggml_backend_context_t ctx, struct gg
static struct ggml_backend_interface cpu_backend_interface = { static struct ggml_backend_interface cpu_backend_interface = {
/* .get_name = */ ggml_backend_cpu_name, /* .get_name = */ ggml_backend_cpu_name,
/* .free_context = */ ggml_backend_cpu_free_context, /* .free = */ ggml_backend_cpu_free,
/* .alloc_buffer = */ ggml_backend_cpu_alloc_buffer, /* .alloc_buffer = */ ggml_backend_cpu_alloc_buffer,
/* .free_buffer = */ ggml_backend_cpu_free_buffer,
/* .reset_buffer = */ ggml_backend_cpu_reset_buffer,
/* .alloc_tensor = */ ggml_backend_cpu_alloc_tensor,
/* .set_tensor_async = */ ggml_backend_cpu_set_tensor_async, /* .set_tensor_async = */ ggml_backend_cpu_set_tensor_async,
/* .get_tensor_async = */ ggml_backend_cpu_get_tensor_async, /* .get_tensor_async = */ ggml_backend_cpu_get_tensor_async,
/* .synchronize = */ ggml_backend_cpu_synchronize, /* .synchronize = */ ggml_backend_cpu_synchronize,
@ -248,14 +302,16 @@ static struct ggml_backend_interface cpu_backend_interface = {
/* .graph_compute = */ ggml_backend_cpu_graph_compute /* .graph_compute = */ ggml_backend_cpu_graph_compute
}; };
struct ggml_backend ggml_backend_cpu_init(void) { struct ggml_backend * ggml_backend_cpu_init(void) {
struct ggml_backend_cpu_context * ctx = malloc(sizeof(struct ggml_backend_cpu_context)); struct ggml_backend_cpu_context * ctx = malloc(sizeof(struct ggml_backend_cpu_context));
ctx->n_threads = GGML_DEFAULT_N_THREADS; ctx->n_threads = GGML_DEFAULT_N_THREADS;
ctx->work_data = NULL; ctx->work_data = NULL;
ctx->work_size = 0; ctx->work_size = 0;
struct ggml_backend cpu_backend = { struct ggml_backend * cpu_backend = malloc(sizeof(struct ggml_backend));
/* .interface = */ &cpu_backend_interface,
*cpu_backend = (struct ggml_backend) {
/* .interface = */ cpu_backend_interface,
/* .context = */ ctx /* .context = */ ctx
}; };
return cpu_backend; return cpu_backend;
@ -287,15 +343,10 @@ void ggml_graph_splits_add_n_va(struct ggml_graph_splits * splits, struct ggml_t
struct ggml_graph_split * split = &splits->splits[splits->n_splits]; struct ggml_graph_split * split = &splits->splits[splits->n_splits];
// check if the split is on the same backend as the previous one
// FIXME: need to check all the inputs
if ((*inputs[0])->backend == ggml_get_ctx_backend(ctx)) { if ((*inputs[0])->backend == ggml_get_ctx_backend(ctx)) {
if (splits->n_splits > 0) { if (splits->n_splits == 0) {
char name[GGML_MAX_NAME];
vsnprintf(name, sizeof(name), fmt, args);
char new_name[GGML_MAX_NAME];
snprintf(new_name, sizeof(new_name), "%s,%s", splits->splits[splits->n_splits - 1].name, name);
strcpy(splits->splits[splits->n_splits - 1].name, new_name);
return;
}
// always add the first split // always add the first split
int i = 0; int i = 0;
while (inputs[i] != NULL) { while (inputs[i] != NULL) {
@ -307,6 +358,16 @@ void ggml_graph_splits_add_n_va(struct ggml_graph_splits * splits, struct ggml_t
split->src_inputs[i] = NULL; split->src_inputs[i] = NULL;
split->dst_inputs[i] = NULL; split->dst_inputs[i] = NULL;
} else { } else {
// add to the previous split
char name[GGML_MAX_NAME - 2];
int n = vsnprintf(name, sizeof(name), fmt, args);
char new_name[GGML_MAX_NAME];
snprintf(new_name, sizeof(new_name), "%.*s,%s", GGML_MAX_NAME - n - 2, splits->splits[splits->n_splits - 1].name, name);
strcpy(splits->splits[splits->n_splits - 1].name, new_name);
return;
}
} else {
// add a new split
int i = 0; int i = 0;
while (inputs[i] != NULL) { while (inputs[i] != NULL) {
GGML_ASSERT(i < GGML_MAX_SPLIT_INPUTS); GGML_ASSERT(i < GGML_MAX_SPLIT_INPUTS);
@ -360,8 +421,6 @@ void ggml_graph_splits_build_forward(struct ggml_graph_splits * splits, struct g
// TODO: allocate graphs in context // TODO: allocate graphs in context
split->graph = (struct ggml_cgraph *) malloc(sizeof(struct ggml_cgraph)); split->graph = (struct ggml_cgraph *) malloc(sizeof(struct ggml_cgraph));
memset(split->graph, 0, sizeof(struct ggml_cgraph)); memset(split->graph, 0, sizeof(struct ggml_cgraph));
// *split->graph = ggml_build_forward_range(output, split->input);
// *split->graph = ggml_build_forward(output);
for (int j = 0; outputs[j] != NULL; j++) { for (int j = 0; outputs[j] != NULL; j++) {
ggml_build_forward_expand(split->graph, outputs[j]); ggml_build_forward_expand(split->graph, outputs[j]);
} }
@ -404,10 +463,8 @@ void ggml_graph_splits_compute(struct ggml_graph_splits * splits) {
// copy the input tensor to the backend // copy the input tensor to the backend
uint64_t copy_start_us = ggml_time_us(); uint64_t copy_start_us = ggml_time_us();
for (int j = 0; split->src_inputs[j] != NULL; j++) { for (int j = 0; split->src_inputs[j] != NULL; j++) {
if (split->src_inputs[j] != split->dst_inputs[j]) {
//printf("\tcopying tensor %d (%s) (%lu bytes)\n", j, split->src_inputs[j]->name, ggml_nbytes(split->src_inputs[j])); //printf("\tcopying tensor %d (%s) (%lu bytes)\n", j, split->src_inputs[j]->name, ggml_nbytes(split->src_inputs[j]));
ggml_backend_cpy_tensor(split->dst_inputs[j], split->src_inputs[j]); ggml_backend_tensor_copy(split->src_inputs[j], split->dst_inputs[j]);
}
} }
// ggml_backend_synchronize(split->dst_inputs[0]->backend); // ggml_backend_synchronize(split->dst_inputs[0]->backend);
copy_us += ggml_time_us() - copy_start_us; copy_us += ggml_time_us() - copy_start_us;
@ -433,3 +490,187 @@ void ggml_graph_splits_compute(struct ggml_graph_splits * splits) {
//printf("splits: %d, nodes: %d, copy: %.2fms, compute_cpu: %.2fms, compute_gpu: %.2fms\n", splits->n_splits, n_nodes, copy_us / 1000.0, compute_cpu_us / 1000.0, compute_gpu_us / 1000.0); //printf("splits: %d, nodes: %d, copy: %.2fms, compute_cpu: %.2fms, compute_gpu: %.2fms\n", splits->n_splits, n_nodes, copy_us / 1000.0, compute_cpu_us / 1000.0, compute_gpu_us / 1000.0);
//exit(0); //exit(0);
} }
#if 0
// default allocator
struct free_block {
void * addr;
size_t size;
};
struct ggml_backend_default_allocator_context {
void * data;
size_t alignment;
int n_free_blocks;
struct free_block free_blocks[];
};
void ggml_backend_default_allocator_free_context(ggml_allocator_context_t ctx) {
struct ggml_backend_default_allocator_context * allocator_ctx = ctx;
free(allocator_ctx);
}
ggml_allocator_context_t ggml_backend_default_allocator_context(void * data, size_t size, size_t alignment, int n_free_blocks) {
struct ggml_backend_default_allocator_context * ctx = malloc(sizeof(struct ggml_backend_default_allocator_context) + n_free_blocks * sizeof(struct free_block));
ctx->data = data;
ctx->alignment = alignment;
ctx->n_free_blocks = 1;
size_t align_offset = align_offset(data, alignment);
ctx->free_blocks[0].addr = (char *)data + align_offset;
ctx->free_blocks[0].size = size - align_offset;
return ctx;
}
void * ggml_backend_default_allocator_alloc(ggml_allocator_context_t ctx, size_t size) {
struct ggml_backend_default_allocator_context * allocator_ctx = ctx;
size = align_size(size, allocator_ctx->alignment);
// find a free block
for (int i = 0; i < allocator_ctx->n_free_blocks; i++) {
struct free_block * block = &allocator_ctx->free_blocks[i];
if (block->size >= size) {
void * addr = block->addr;
block->addr += size;
block->size -= size;
if (block->size == 0) {
// remove block if empty
allocator_ctx->n_free_blocks--;
for (int j = i; j < allocator_ctx->n_free_blocks; j++) {
allocator_ctx->free_blocks[j] = allocator_ctx->free_blocks[j+1];
}
}
return addr;
}
}
return NULL;
}
// this is a very naive implementation, but for our case the number of free blocks should be very small
void ggml_backend_default_allocator_free(ggml_allocator_context_t ctx, void * ptr, size_t size) {
struct ggml_backend_default_allocator_context * allocator_ctx = ctx;
size = align_size(size, allocator_ctx->alignment);
// see if we can merge with an existing block
for (int i = 0; i < allocator_ctx->n_free_blocks; i++) {
struct free_block * block = &allocator_ctx->free_blocks[i];
// check if ptr is at the end of the block
if (block->addr + block->size == ptr) {
block->size += size;
// check if we can merge with the next block
if (i < allocator_ctx->n_free_blocks - 1 && block->addr + block->size == allocator_ctx->free_blocks[i+1].addr) {
block->size += allocator_ctx->free_blocks[i+1].size;
allocator_ctx->n_free_blocks--;
for (int j = i+1; j < allocator_ctx->n_free_blocks; j++) {
allocator_ctx->free_blocks[j] = allocator_ctx->free_blocks[j+1];
}
}
return;
}
// check if ptr is at the beginning of the block
if (ptr + size == block->addr) {
block->addr = ptr;
block->size += size;
// check if we can merge with the previous block
if (i > 0 && allocator_ctx->free_blocks[i-1].addr + allocator_ctx->free_blocks[i-1].size == block->addr) {
allocator_ctx->free_blocks[i-1].size += block->size;
allocator_ctx->n_free_blocks--;
for (int j = i; j < allocator_ctx->n_free_blocks; j++) {
allocator_ctx->free_blocks[j] = allocator_ctx->free_blocks[j+1];
}
}
return;
}
}
// otherwise, add a new block
if (allocator_ctx->n_free_blocks < MAX_FREE_BLOCKS) {
// insert the new block in the correct position to keep the array sorted
int insert_pos = 0;
while (insert_pos < allocator_ctx->n_free_blocks && allocator_ctx->free_blocks[insert_pos].addr < ptr) {
insert_pos++;
}
// shift all blocks from insert_pos onward to make room for the new block
for (int i = allocator_ctx->n_free_blocks; i > insert_pos; i--) {
allocator_ctx->free_blocks[i] = allocator_ctx->free_blocks[i-1];
}
// insert the new block
allocator_ctx->free_blocks[insert_pos].addr = ptr;
allocator_ctx->free_blocks[insert_pos].size = size;
allocator_ctx->n_free_blocks++;
}
else {
GGML_ASSERT(!"out of free blocks");
}
}
static bool ggml_is_view(struct ggml_tensor * t) {
return t->op == GGML_OP_RESHAPE || t->op == GGML_OP_VIEW || t->op == GGML_OP_TRANSPOSE ||
t->op == GGML_OP_PERMUTE || t->op == GGML_OP_NONE;
}
NOTE: id can be n_leaf OR n_node instead, we can determine the type by checking if the node is a leaf or not
void allocate_graph(struct ggml_cgraph * gf, struct ggml_buffer * buffer) {
int node_children_count[GGML_MAX_NODES*2];
int node_view_count[GGML_MAX_NODES*2];
memset(node_children_count, 0, sizeof(int) * (gf->n_nodes + gf->n_leafs));
memset(node_view_count, 0, sizeof(int) * (gf->n_nodes + gf->n_leafs));
// count number of children and views
for (int i = 0; i < gf->n_nodes; i++) {
struct ggml_tensor * node = gf->nodes[i];
for (int j = 0; j < GGML_MAX_SRC; j++) {
struct ggml_tensor * parent = node->src[j];
if (parent == NULL) {
break;
}
// todo: ....
node_children_count[parent->id] += 1;
if (ggml_is_view(parent)) {
struct ggml_tensor * ancestor = parent;
do {
node_view_count[ancestor->id] += 1;
ancestor = ancestor->src[0];
} while (ggml_is_view(ancestor));
}
}
}
// allocate tensors
for (int i = 0; i < gf->n_nodes; i++) {
struct ggml_tensor * node = gf->nodes[i];
bool is_view = ggml_is_view(node);
if (is_view) {
// allocate view accordingly to the OP
node->data = node->src[0]->data; // + offset
struct ggml_tensor * ancestor = node->src[0];
while (ggml_is_view(ancestor)) {
ancestor = ancestor->src[0];
}
node_view_count[ancestor->id] -= 1;
} else {
if (node->data == NULL) {
// allocate tensor
// TODO: if last children and size == parent.size, then reuse parent tensor (auto in-place)
// may need a list of ops that can be in-place
ggml_backend_alloc_tensor(buffer, node);
}
}
// update parents
for (int j = 0; j < GGML_MAX_SRC; j++) {
struct ggml_tensor * parent = node->src[j];
if (parent == NULL) {
break;
}
if (is_view) {
node_view_count[parent->id] -= 1;
}
node_children_count[parent->id] -= 1;
if (node_children_count[parent->id] == 0 && node_view_count[parent->id] == 0) {
// free parent
ggml_backend_free_tensor(buffer, parent);
}
}
}
}
#endif

118
ggml-backend.h
View File

@ -5,12 +5,45 @@
#ifdef __cplusplus #ifdef __cplusplus
extern "C" { extern "C" {
#endif #endif
typedef void * ggml_graph_plan_t;
typedef void * ggml_backend_context_t;
typedef void * ggml_backend_buffer_t;
struct ggml_backend; struct ggml_backend;
// backend buffers
typedef void * ggml_buffer_context_t;
struct ggml_backend_buffer;
struct ggml_backend_buffer_interface {
// allocator functions
void (*free_buffer) (struct ggml_backend_buffer * alloc);
void (*alloc_tensor) (struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor);
void (*free_tensor) (struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor);
void (*reset) (struct ggml_backend_buffer * alloc);
// functions overriden by the backend
size_t (*get_alloc_size)(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor); // pre-allocation callback
void (*init_tensor) (struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor); // post-allocation callback
void (*free_data) (struct ggml_backend_buffer * alloc); // free backend-specific data // TODO: better name
};
struct ggml_backend_buffer {
struct ggml_backend_buffer_interface interface;
ggml_buffer_context_t context;
void * backend_data;
};
// backend buffer helper functions
GGML_API void ggml_backend_buffer_free(struct ggml_backend_buffer * alloc);
static inline void ggml_backend_buffer_tensor_alloc(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor) { alloc->interface.alloc_tensor(alloc, tensor); }
static inline void ggml_backend_buffer_free_tensor(struct ggml_backend_buffer * alloc, struct ggml_tensor * tensor) { alloc->interface.free_tensor(alloc, tensor); }
static inline void ggml_backend_buffer_reset(struct ggml_backend_buffer * alloc) { alloc->interface.reset(alloc); }
// default buffer allocators
// simple buffer allocator: cannot free tensors, good for weights and small contexts
// default buffer allocator: can free tensors, good for compute contexts
GGML_API struct ggml_backend_buffer * ggml_allocator_simple_init(void * data, size_t size, size_t alignment);
GGML_API struct ggml_backend_buffer * ggml_allocator_default_init(void * data, size_t size, size_t alignment, int max_free_blocks);
// buffer
// buffers have space for the tensor structs in host memory, and tensor data in backend-specific memory // buffers have space for the tensor structs in host memory, and tensor data in backend-specific memory
struct ggml_buffer { struct ggml_buffer {
// host memory // host memory
@ -19,75 +52,70 @@ extern "C" {
// tensor data // tensor data
struct ggml_backend * backend; struct ggml_backend * backend;
ggml_backend_buffer_t backend_buffer; // backend-specific data struct ggml_backend_buffer * backend_buffer;
}; };
GGML_API struct ggml_buffer * ggml_buffer_alloc(struct ggml_backend * backend, size_t size, size_t max_tensors);
GGML_API void ggml_buffer_free(struct ggml_buffer * buffer);
// backend
typedef void * ggml_backend_context_t;
typedef void * ggml_graph_plan_t;
struct ggml_backend_interface { struct ggml_backend_interface {
const char * (*get_name)(ggml_backend_context_t ctx); const char * (*get_name)(struct ggml_backend * backend);
void (*free_context)(ggml_backend_context_t ctx); void (*free)(struct ggml_backend * backend);
// buffers // buffer allocation
ggml_backend_buffer_t (*alloc_buffer)(ggml_backend_context_t ctx, size_t size); struct ggml_backend_buffer * (*alloc_buffer)(struct ggml_backend * backend, size_t size);
void (*free_buffer) (ggml_backend_context_t ctx, ggml_backend_buffer_t buffer);
void (*reset_buffer)(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer);
void (*alloc_tensor)(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer, struct ggml_tensor * tensor);
// TODO: pinned buffers for faster transfers between host and device
// tensor data access // tensor data access
// these functions can be asynchronous. helper functions are provided for synchronous access that automatically call synchronize // these functions can be asynchronous. helper functions are provided for synchronous access that automatically call synchronize
void (*set_tensor_async)(ggml_backend_context_t ctx, struct ggml_tensor * tensor, const void * data, size_t offset, size_t size); void (*set_tensor_async)(struct ggml_backend * backend, struct ggml_tensor * tensor, const void * data, size_t offset, size_t size);
void (*get_tensor_async)(ggml_backend_context_t ctx, const struct ggml_tensor * tensor, void * data, size_t offset, size_t size); void (*get_tensor_async)(struct ggml_backend * backend, const struct ggml_tensor * tensor, void * data, size_t offset, size_t size);
void (*synchronize)(ggml_backend_context_t ctx); void (*synchronize) (struct ggml_backend * backend);
// (optional) copy tensor between different backends, allow for single-copy tranfers // (optional) copy tensor between different backends, allow for single-copy tranfers
void (*cpy_tensor_from)(ggml_backend_context_t ctx, struct ggml_tensor * src, struct ggml_tensor * dst); void (*cpy_tensor_from)(struct ggml_backend * backend, struct ggml_tensor * src, struct ggml_tensor * dst);
void (*cpy_tensor_to) (ggml_backend_context_t ctx, struct ggml_tensor * src, struct ggml_tensor * dst); void (*cpy_tensor_to) (struct ggml_backend * backend, struct ggml_tensor * src, struct ggml_tensor * dst);
// compute graph with a plan // compute graph with a plan
ggml_graph_plan_t (*graph_plan_create) (ggml_backend_context_t ctx, struct ggml_cgraph * cgraph); ggml_graph_plan_t (*graph_plan_create) (struct ggml_backend * backend, struct ggml_cgraph * cgraph);
void (*graph_plan_free) (ggml_backend_context_t ctx, ggml_graph_plan_t plan); void (*graph_plan_free) (struct ggml_backend * backend, ggml_graph_plan_t plan);
void (*graph_plan_compute)(ggml_backend_context_t ctx, ggml_graph_plan_t plan); void (*graph_plan_compute)(struct ggml_backend * backend, ggml_graph_plan_t plan);
// compute graph without a plan // compute graph without a plan
void (*graph_compute) (ggml_backend_context_t ctx, struct ggml_cgraph * cgraph); void (*graph_compute) (struct ggml_backend * backend, struct ggml_cgraph * cgraph);
// check if a backend supports a given operation // check if a backend supports a given operation
// this could be used to fallback automatically to the CPU backend if a backend doesn't support an operation // this could be used to fallback automatically to the CPU backend if a backend doesn't support an operation
// bool (*supports_op)(ggml_backend_context_t ctx, struct ggml_tensor * op); // bool (*supports_op)(struct ggml_backend * backend, struct ggml_tensor * op);
}; };
struct ggml_backend { struct ggml_backend {
struct ggml_backend_interface * interface; struct ggml_backend_interface interface;
ggml_backend_context_t context; ggml_backend_context_t context;
}; };
// backend helper functions // backend helper functions
static inline const char * ggml_backend_name(struct ggml_backend * backend) { return backend->interface->get_name(backend->context); } static inline const char * ggml_backend_name(struct ggml_backend * backend) { return backend->interface.get_name(backend); }
static inline void ggml_backend_free_context(struct ggml_backend * backend) { backend->interface->free_context(backend->context); } static inline void ggml_backend_free(struct ggml_backend * backend) { backend->interface.free(backend); }
static inline void ggml_backend_set_tensor_async(struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) { tensor->backend->interface->set_tensor_async(tensor->backend->context, tensor, data, offset, size); } static inline void ggml_backend_tensor_set_async(struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) { tensor->backend->interface.set_tensor_async(tensor->backend, tensor, data, offset, size); }
static inline void ggml_backend_get_tensor_async(const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) { tensor->backend->interface->get_tensor_async(tensor->backend->context, tensor, data, offset, size); } static inline void ggml_backend_tensor_get_async(const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) { tensor->backend->interface.get_tensor_async(tensor->backend, tensor, data, offset, size); }
static inline void ggml_backend_set_tensor(struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) { tensor->backend->interface->set_tensor_async(tensor->backend->context, tensor, data, offset, size); tensor->backend->interface->synchronize(tensor->backend->context); } static inline void ggml_backend_tensor_set(struct ggml_tensor * tensor, const void * data, size_t offset, size_t size) { tensor->backend->interface.set_tensor_async(tensor->backend, tensor, data, offset, size); tensor->backend->interface.synchronize(tensor->backend); }
static inline void ggml_backend_get_tensor(const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) { tensor->backend->interface->get_tensor_async(tensor->backend->context, tensor, data, offset, size); tensor->backend->interface->synchronize(tensor->backend->context); } static inline void ggml_backend_tensor_get(const struct ggml_tensor * tensor, void * data, size_t offset, size_t size) { tensor->backend->interface.get_tensor_async(tensor->backend, tensor, data, offset, size); tensor->backend->interface.synchronize(tensor->backend); }
static inline void ggml_backend_synchronize(struct ggml_backend * backend) { backend->interface->synchronize(backend->context); } static inline void ggml_backend_synchronize(struct ggml_backend * backend) { backend->interface.synchronize(backend); }
static inline ggml_graph_plan_t ggml_backend_graph_plan_create(struct ggml_backend * backend, struct ggml_cgraph * cgraph) { return backend->interface->graph_plan_create(backend->context, cgraph); } static inline ggml_graph_plan_t ggml_backend_graph_plan_create(struct ggml_backend * backend, struct ggml_cgraph * cgraph) { return backend->interface.graph_plan_create(backend, cgraph); }
static inline void ggml_backend_graph_plan_free(struct ggml_backend * backend, ggml_graph_plan_t plan) { backend->interface->graph_plan_free(backend->context, plan); } static inline void ggml_backend_graph_plan_free(struct ggml_backend * backend, ggml_graph_plan_t plan) { backend->interface.graph_plan_free(backend, plan); }
static inline void ggml_backend_graph_plan_compute(struct ggml_backend * backend, ggml_graph_plan_t plan) { backend->interface->graph_plan_compute(backend->context, plan); } static inline void ggml_backend_graph_plan_compute(struct ggml_backend * backend, ggml_graph_plan_t plan) { backend->interface.graph_plan_compute(backend, plan); }
static inline void ggml_backend_graph_compute(struct ggml_backend * backend, struct ggml_cgraph * cgraph) { backend->interface->graph_compute(backend->context, cgraph); } static inline void ggml_backend_graph_compute(struct ggml_backend * backend, struct ggml_cgraph * cgraph) { backend->interface.graph_compute(backend, cgraph); }
// buffer and tensor allocation
GGML_API struct ggml_buffer ggml_backend_alloc_buffer(struct ggml_backend * backend, size_t size, size_t max_tensors);
GGML_API void ggml_backend_free_buffer(struct ggml_buffer * buffer);
static inline void ggml_backend_reset_buffer(struct ggml_buffer * buffer) { buffer->backend->interface->reset_buffer(buffer->backend->context, buffer->backend_buffer); }
static inline void ggml_backend_alloc_tensor(struct ggml_buffer * buffer, struct ggml_tensor * tensor) { buffer->backend->interface->alloc_tensor(buffer->backend->context, buffer->backend_buffer, tensor); }
// tensor copy between different backends // tensor copy between different backends
GGML_API void ggml_backend_cpy_tensor(struct ggml_tensor * dst, struct ggml_tensor * src); GGML_API void ggml_backend_tensor_copy(struct ggml_tensor * src, struct ggml_tensor * dst);
// CPU backend // CPU backend
GGML_API struct ggml_backend ggml_backend_cpu_init(void); GGML_API struct ggml_backend * ggml_backend_cpu_init(void);
GGML_API void ggml_backend_cpu_set_n_threads(struct ggml_backend * backend_cpu, int n_threads); GGML_API void ggml_backend_cpu_set_n_threads(struct ggml_backend * backend_cpu, int n_threads);
/////////////////////////// ///////////////////////////

118
ggml-cuda.cu
View File

@ -585,6 +585,14 @@ void ggml_cuda_host_free(void * ptr) {
CUDA_CHECK(cudaFreeHost(ptr)); CUDA_CHECK(cudaFreeHost(ptr));
} }
void ggml_cuda_host_register(void * ptr, size_t size) {
CUDA_CHECK(cudaHostRegister(ptr, size, 0));
}
void ggml_cuda_host_unregister(void * ptr) {
CUDA_CHECK(cudaHostUnregister(ptr));
}
template<typename src0_t, typename src1_t, typename dst_t> template<typename src0_t, typename src1_t, typename dst_t>
static void ggml_cuda_op_add( static void ggml_cuda_op_add(
const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst,
@ -792,9 +800,9 @@ static void ggml_cuda_op_rope(
const int64_t ne00 = src0->ne[0]; const int64_t ne00 = src0->ne[0];
const int64_t i01_diff = i01_high - i01_low; const int64_t i01_diff = i01_high - i01_low;
const int n_past = ((int32_t *) dst->params)[0]; const int n_past = ((int32_t *) dst->op_params)[0];
const int n_dims = ((int32_t *) dst->params)[1]; const int n_dims = ((int32_t *) dst->op_params)[1];
const int mode = ((int32_t *) dst->params)[2]; const int mode = ((int32_t *) dst->op_params)[2];
//const int n_ctx = ((int32_t *) dst->params)[3]; //const int n_ctx = ((int32_t *) dst->params)[3];
GGML_ASSERT(mode == 0); GGML_ASSERT(mode == 0);
@ -822,7 +830,7 @@ static void ggml_cuda_op_diag_mask_inf(
const int64_t ne01 = src0->ne[1]; const int64_t ne01 = src0->ne[1];
const int64_t i01_diff = i01_high - i01_low; const int64_t i01_diff = i01_high - i01_low;
const int n_past = ((int32_t *) dst->params)[0]; const int n_past = ((int32_t *) dst->op_params)[0];
// compute // compute
diag_mask_inf_cuda((src0_t *)src0_d, (dst_t *)dst_d, ne00, i01_diff, ne01, n_past, stream); diag_mask_inf_cuda((src0_t *)src0_d, (dst_t *)dst_d, ne00, i01_diff, ne01, n_past, stream);
@ -1689,16 +1697,17 @@ struct ggml_backend_cuda_context {
ggml_cuda_context * cuda_ctx = ggml_cuda_init(); ggml_cuda_context * cuda_ctx = ggml_cuda_init();
}; };
static const char * ggml_backend_cuda_name(ggml_backend_context_t ctx) { static const char * ggml_backend_cuda_name(ggml_backend * backend) {
return "CUDA"; return "CUDA";
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cuda_free_context(ggml_backend_context_t ctx) { static void ggml_backend_cuda_free(ggml_backend * backend) {
ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)ctx; ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
ggml_cuda_free(cuda_ctx->cuda_ctx); ggml_cuda_free(cuda_ctx->cuda_ctx);
delete cuda_ctx; delete cuda_ctx;
delete backend;
} }
struct cuda_backend_buffer { struct cuda_backend_buffer {
@ -1709,116 +1718,82 @@ struct cuda_backend_buffer {
static const size_t TENSOR_ALIGNMENT = 128; static const size_t TENSOR_ALIGNMENT = 128;
static size_t aligned_offset(const void * buffer, size_t offset, size_t alignment) { static void ggml_backend_cuda_free_buffer(struct ggml_backend_buffer * alloc) {
assert(alignment && !(alignment & (alignment - 1))); // power of 2 CUDA_CHECK(cudaFree(alloc->backend_data));
size_t align = (alignment - (((uintptr_t)buffer + offset) % alignment)) % alignment;
return offset + align;
} }
static ggml_backend_buffer_t ggml_backend_cuda_alloc_buffer(ggml_backend_context_t ctx, size_t size) { static ggml_backend_buffer * ggml_backend_cuda_alloc_buffer(ggml_backend * backend, size_t size) {
cuda_backend_buffer * buffer = new cuda_backend_buffer; void * data;
CUDA_CHECK(cudaMalloc(&data, size));
CUDA_CHECK(cudaMalloc(&buffer->data, size)); ggml_backend_buffer * buffer = ggml_allocator_simple_init(data, size, TENSOR_ALIGNMENT);
buffer->offset = 0; // cudaMalloc returns aligned pointers buffer->interface.free_data = ggml_backend_cuda_free_buffer;
buffer->size = size; buffer->backend_data = data;
return buffer; return buffer;
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cuda_free_buffer(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer) { static void ggml_backend_cuda_set_tensor_async(ggml_backend * backend, ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
cuda_backend_buffer * cuda_buffer = (cuda_backend_buffer *)buffer;
CUDA_CHECK(cudaFree(cuda_buffer->data));
delete cuda_buffer;
UNUSED(ctx);
}
static void ggml_backend_cuda_reset_buffer(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer) {
cuda_backend_buffer * cuda_buffer = (cuda_backend_buffer *)buffer;
cuda_buffer->offset = 0;
UNUSED(ctx);
}
static void ggml_backend_cuda_alloc_tensor(ggml_backend_context_t ctx, ggml_backend_buffer_t buffer, ggml_tensor * tensor) {
cuda_backend_buffer * cuda_buffer = (cuda_backend_buffer *)buffer;
if (cuda_buffer->offset + ggml_nbytes(tensor) > cuda_buffer->size) {
fprintf(stderr, "%s: not enough space in the CUDA buffer (needed %zu, available %zu)\n",
__func__, ggml_nbytes(tensor), cuda_buffer->size - cuda_buffer->offset);
GGML_ASSERT(false);
}
tensor->data = (char*)cuda_buffer->data + cuda_buffer->offset;
cuda_buffer->offset = aligned_offset(cuda_buffer->data, cuda_buffer->offset + ggml_nbytes(tensor), TENSOR_ALIGNMENT);
UNUSED(ctx);
}
static void ggml_backend_cuda_set_tensor_async(ggml_backend_context_t ctx, ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor write out of bounds"); GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor write out of bounds");
GGML_ASSERT(tensor->data != NULL && "tensor not allocated"); GGML_ASSERT(tensor->data != NULL && "tensor not allocated");
//ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)ctx; //ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
CUDA_CHECK(cudaMemcpyAsync((char*)tensor->data + offset, data, size, cudaMemcpyHostToDevice, g_cudaStream_main)); CUDA_CHECK(cudaMemcpyAsync((char*)tensor->data + offset, data, size, cudaMemcpyHostToDevice, g_cudaStream_main));
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cuda_get_tensor_async(ggml_backend_context_t ctx, const ggml_tensor * tensor, void * data, size_t offset, size_t size) { static void ggml_backend_cuda_get_tensor_async(ggml_backend * backend, const ggml_tensor * tensor, void * data, size_t offset, size_t size) {
GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor read out of bounds"); GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && "tensor read out of bounds");
GGML_ASSERT(tensor->data != NULL && "tensor not allocated"); GGML_ASSERT(tensor->data != NULL && "tensor not allocated");
//ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)ctx; //ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
CUDA_CHECK(cudaMemcpyAsync(data, (const char*)tensor->data + offset, size, cudaMemcpyDeviceToHost, g_cudaStream_main)); CUDA_CHECK(cudaMemcpyAsync(data, (const char*)tensor->data + offset, size, cudaMemcpyDeviceToHost, g_cudaStream_main));
UNUSED(ctx); UNUSED(backend);
} }
static void ggml_backend_cuda_synchronize(ggml_backend_context_t ctx) { static void ggml_backend_cuda_synchronize(ggml_backend * backend) {
ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)ctx; ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
ggml_cuda_synchronize(cuda_ctx->cuda_ctx); ggml_cuda_synchronize(cuda_ctx->cuda_ctx);
} }
static ggml_graph_plan_t ggml_backend_cuda_graph_plan_create(ggml_backend_context_t ctx, ggml_cgraph * cgraph) { static ggml_graph_plan_t ggml_backend_cuda_graph_plan_create(ggml_backend * backend, ggml_cgraph * cgraph) {
GGML_ASSERT(false); GGML_ASSERT(false);
return nullptr; return nullptr;
UNUSED(ctx); UNUSED(backend);
UNUSED(cgraph); UNUSED(cgraph);
} }
static void ggml_backend_cuda_graph_plan_free(ggml_backend_context_t ctx, ggml_graph_plan_t plan) { static void ggml_backend_cuda_graph_plan_free(ggml_backend * backend, ggml_graph_plan_t plan) {
GGML_ASSERT(false); GGML_ASSERT(false);
UNUSED(ctx); UNUSED(backend);
UNUSED(plan); UNUSED(plan);
} }
static void ggml_backend_cuda_graph_plan_compute(ggml_backend_context_t ctx, ggml_graph_plan_t plan) { static void ggml_backend_cuda_graph_plan_compute(ggml_backend * backend, ggml_graph_plan_t plan) {
GGML_ASSERT(false); GGML_ASSERT(false);
UNUSED(ctx); UNUSED(backend);
UNUSED(plan); UNUSED(plan);
} }
static void ggml_backend_cuda_graph_compute(ggml_backend_context_t ctx, ggml_cgraph * cgraph) { static void ggml_backend_cuda_graph_compute(ggml_backend * backend, ggml_cgraph * cgraph) {
ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)ctx; ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
ggml_cuda_cgraph_compute(cuda_ctx->cuda_ctx, cgraph); ggml_cuda_cgraph_compute(cuda_ctx->cuda_ctx, cgraph);
} }
static ggml_backend_interface cuda_backend_interface = { static ggml_backend_interface cuda_backend_interface = {
/* .get_name = */ ggml_backend_cuda_name, /* .get_name = */ ggml_backend_cuda_name,
/* .free_context = */ ggml_backend_cuda_free_context, /* .free = */ ggml_backend_cuda_free,
/* .alloc_buffer = */ ggml_backend_cuda_alloc_buffer, /* .alloc_buffer = */ ggml_backend_cuda_alloc_buffer,
/* .free_buffer = */ ggml_backend_cuda_free_buffer,
/* .reset_buffer = */ ggml_backend_cuda_reset_buffer,
/* .alloc_tensor = */ ggml_backend_cuda_alloc_tensor,
/* .set_tensor_async = */ ggml_backend_cuda_set_tensor_async, /* .set_tensor_async = */ ggml_backend_cuda_set_tensor_async,
/* .get_tensor_async = */ ggml_backend_cuda_get_tensor_async, /* .get_tensor_async = */ ggml_backend_cuda_get_tensor_async,
/* .synchronize = */ ggml_backend_cuda_synchronize, /* .synchronize = */ ggml_backend_cuda_synchronize,
@ -1830,11 +1805,12 @@ static ggml_backend_interface cuda_backend_interface = {
/* .graph_compute = */ ggml_backend_cuda_graph_compute /* .graph_compute = */ ggml_backend_cuda_graph_compute
}; };
ggml_backend ggml_backend_cuda_init(void) { ggml_backend * ggml_backend_cuda_init(void) {
ggml_backend_cuda_context * ctx = new ggml_backend_cuda_context; ggml_backend_cuda_context * ctx = new ggml_backend_cuda_context;
ggml_backend cuda_backend = { ggml_backend * cuda_backend = new ggml_backend;
/* .interface = */ &cuda_backend_interface, *cuda_backend = (ggml_backend){
/* .interface = */ cuda_backend_interface,
/* .context = */ ctx /* .context = */ ctx
}; };
return cuda_backend; return cuda_backend;

8
ggml-cuda.h
View File

@ -6,12 +6,14 @@
extern "C" { extern "C" {
#endif #endif
void * ggml_cuda_host_malloc(size_t size); GGML_API void * ggml_cuda_host_malloc(size_t size);
void ggml_cuda_host_free(void * ptr); GGML_API void ggml_cuda_host_free(void * ptr);
GGML_API void ggml_cuda_host_register(void * ptr, size_t size);
GGML_API void ggml_cuda_host_unregister(void * ptr);
// backend API // backend API
struct ggml_backend ggml_backend_cuda_init(); GGML_API struct ggml_backend * ggml_backend_cuda_init();
#ifdef __cplusplus #ifdef __cplusplus

78
ggml.c
View File

@ -4393,7 +4393,7 @@ struct ggml_context * ggml_init(struct ggml_init_params params) {
/*.compute_type =*/ params.compute_type, /*.compute_type =*/ params.compute_type,
}; };
ggml_backend_reset_buffer(params.buffer); ggml_backend_buffer_reset(params.buffer->backend_buffer);
GGML_ASSERT(ctx->mem_buffer != NULL); GGML_ASSERT(ctx->mem_buffer != NULL);
@ -4526,17 +4526,17 @@ struct ggml_tensor * ggml_new_tensor_impl(
/*.ne =*/ { 1, 1, 1, 1 }, /*.ne =*/ { 1, 1, 1, 1 },
/*.nb =*/ { 0, 0, 0, 0 }, /*.nb =*/ { 0, 0, 0, 0 },
/*.op =*/ GGML_OP_NONE, /*.op =*/ GGML_OP_NONE,
/*.op_params =*/ { 0 },
/*.is_param =*/ false, /*.is_param =*/ false,
/*.grad =*/ NULL, /*.grad =*/ NULL,
/*.src =*/ { NULL }, /*.src =*/ { NULL },
/*.node_id =*/ -1,
/*.perf_runs =*/ 0, /*.perf_runs =*/ 0,
/*.perf_cycles =*/ 0, /*.perf_cycles =*/ 0,
/*.perf_time_us =*/ 0, /*.perf_time_us =*/ 0,
/*.params =*/ { 0 },
/*.data =*/ data, /*.data =*/ data,
/*.name =*/ { 0 }, /*.name =*/ { 0 },
/*.extra =*/ NULL, /*.extra =*/ NULL,
/*.visited =*/ false,
/*.pad =*/ { 0 }, /*.pad =*/ { 0 },
}; };
@ -4551,7 +4551,7 @@ struct ggml_tensor * ggml_new_tensor_impl(
} }
if (data == NULL && !ctx->no_alloc) { if (data == NULL && !ctx->no_alloc) {
ggml_backend_alloc_tensor(ctx->buffer, result); ggml_backend_buffer_tensor_alloc(ctx->buffer->backend_buffer, result);
} }
// TODO: this should not be needed as long as we don't rely on aligned SIMD loads // TODO: this should not be needed as long as we don't rely on aligned SIMD loads
@ -4730,7 +4730,7 @@ struct ggml_tensor * ggml_set_f32(struct ggml_tensor * tensor, float value) {
} }
*/ */
for (int i = 0; i < ggml_nelements(tensor); i++) { for (int i = 0; i < ggml_nelements(tensor); i++) {
ggml_backend_set_tensor(tensor, &value, sizeof(float)*i, sizeof(float)); ggml_backend_tensor_set(tensor, &value, sizeof(float)*i, sizeof(float));
} }
} break; } break;
default: default:
@ -4839,7 +4839,7 @@ float ggml_get_f32_1d(const struct ggml_tensor * tensor, int i) {
GGML_ASSERT(tensor->nb[0] == sizeof(float)); GGML_ASSERT(tensor->nb[0] == sizeof(float));
//return ((float *)(tensor->data))[i]; //return ((float *)(tensor->data))[i];
float value; float value;
ggml_backend_get_tensor(tensor, &value, sizeof(float)*i, sizeof(float)); ggml_backend_tensor_get(tensor, &value, sizeof(float)*i, sizeof(float));
return value; return value;
} break; } break;
default: default:
@ -4912,6 +4912,11 @@ struct ggml_tensor * ggml_format_name(struct ggml_tensor * tensor, const char *
return tensor; return tensor;
} }
static void ggml_set_op_params(struct ggml_tensor * tensor, void * params, size_t params_size) {
GGML_ASSERT(params_size <= GGML_MAX_OP_PARAMS);
memcpy(tensor->op_params, params, params_size);
}
struct ggml_tensor * ggml_view_tensor( struct ggml_tensor * ggml_view_tensor(
struct ggml_context * ctx, struct ggml_context * ctx,
const struct ggml_tensor * src) { const struct ggml_tensor * src) {
@ -6385,8 +6390,7 @@ struct ggml_tensor * ggml_view_1d(
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 1, &ne0, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 1, &ne0, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name); ggml_format_name(result, "%s (view)", a->name);
assert(GGML_MAX_OP_PARAMS >= sizeof(offset)); ggml_set_op_params(result, &offset, sizeof(offset));
memcpy(result->params, &offset, sizeof(offset));
result->op = GGML_OP_VIEW; result->op = GGML_OP_VIEW;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -6417,8 +6421,7 @@ struct ggml_tensor * ggml_view_2d(
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 2, ne, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 2, ne, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name); ggml_format_name(result, "%s (view)", a->name);
assert(GGML_MAX_OP_PARAMS >= sizeof(offset)); ggml_set_op_params(result, &offset, sizeof(offset));
memcpy(result->params, &offset, sizeof(offset));
result->nb[1] = nb1; result->nb[1] = nb1;
result->nb[2] = result->nb[1]*ne1; result->nb[2] = result->nb[1]*ne1;
@ -6455,8 +6458,7 @@ struct ggml_tensor * ggml_view_3d(
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 3, ne, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 3, ne, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name); ggml_format_name(result, "%s (view)", a->name);
assert(GGML_MAX_OP_PARAMS >= sizeof(offset)); ggml_set_op_params(result, &offset, sizeof(offset));
memcpy(result->params, &offset, sizeof(offset));
result->nb[1] = nb1; result->nb[1] = nb1;
result->nb[2] = nb2; result->nb[2] = nb2;
@ -6495,8 +6497,7 @@ struct ggml_tensor * ggml_view_4d(
struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 4, ne, (char *) a->data + offset); struct ggml_tensor * result = ggml_new_tensor_impl(ctx, a->type, 4, ne, (char *) a->data + offset);
ggml_format_name(result, "%s (view)", a->name); ggml_format_name(result, "%s (view)", a->name);
assert(GGML_MAX_OP_PARAMS >= sizeof(offset)); ggml_set_op_params(result, &offset, sizeof(offset));
memcpy(result->params, &offset, sizeof(offset));
result->nb[1] = nb1; result->nb[1] = nb1;
result->nb[2] = nb2; result->nb[2] = nb2;
@ -6569,8 +6570,7 @@ struct ggml_tensor * ggml_permute(
result->src[1] = NULL; result->src[1] = NULL;
int32_t params[] = { axis0, axis1, axis2, axis3 }; int32_t params[] = { axis0, axis1, axis2, axis3 };
assert(GGML_MAX_OP_PARAMS >= sizeof(params)); ggml_set_op_params(result, &params, sizeof(params));
memcpy(result->params, params, sizeof(params));
return result; return result;
} }
@ -6694,8 +6694,7 @@ struct ggml_tensor * ggml_diag_mask_inf_impl(
struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a); struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a);
int32_t params[] = { n_past, inplace ? 1 : 0 }; int32_t params[] = { n_past, inplace ? 1 : 0 };
assert(GGML_MAX_OP_PARAMS >= sizeof(params)); ggml_set_op_params(result, &params, sizeof(params));
memcpy(result->params, params, sizeof(params));
result->op = GGML_OP_DIAG_MASK_INF; result->op = GGML_OP_DIAG_MASK_INF;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -6854,12 +6853,10 @@ struct ggml_tensor * ggml_rope_impl(
struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a); struct ggml_tensor * result = inplace ? ggml_view_tensor(ctx, a) : ggml_dup_tensor(ctx, a);
// TODO: just use a struct
int32_t params[6] = { n_past, n_dims, mode, n_ctx }; int32_t params[6] = { n_past, n_dims, mode, n_ctx };
memcpy(params + 4, &freq_base, sizeof(float)); memcpy(params + 4, &freq_base, sizeof(float));
memcpy(params + 5, &freq_scale, sizeof(float)); memcpy(params + 5, &freq_scale, sizeof(float));
assert(GGML_MAX_OP_PARAMS >= sizeof(params)); ggml_set_op_params(result, &params, sizeof(params));
memcpy(result->params, &params, sizeof(params));
result->op = GGML_OP_ROPE; result->op = GGML_OP_ROPE;
result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL; result->grad = is_node ? ggml_dup_tensor(ctx, result) : NULL;
@ -11392,8 +11389,8 @@ static void ggml_compute_forward_diag_mask_f32(
const int ith = params->ith; const int ith = params->ith;
const int nth = params->nth; const int nth = params->nth;
const int n_past = ((int32_t *) dst->params)[0]; const int n_past = ((int32_t *) dst->op_params)[0];
const bool inplace = (bool)((int32_t *) dst->params)[1]; const bool inplace = (bool)((int32_t *) dst->op_params)[1];
GGML_ASSERT(n_past >= 0); GGML_ASSERT(n_past >= 0);
@ -11910,12 +11907,12 @@ static void ggml_compute_forward_rope_f32(
float freq_base; float freq_base;
float freq_scale; float freq_scale;
const int n_past = ((int32_t *) dst->params)[0]; const int n_past = ((int32_t *) dst->op_params)[0];
const int n_dims = ((int32_t *) dst->params)[1]; const int n_dims = ((int32_t *) dst->op_params)[1];
const int mode = ((int32_t *) dst->params)[2]; const int mode = ((int32_t *) dst->op_params)[2];
const int n_ctx = ((int32_t *) dst->params)[3]; const int n_ctx = ((int32_t *) dst->op_params)[3];
memcpy(&freq_base, (int32_t *) dst->params + 4, sizeof(float)); memcpy(&freq_base, (int32_t *) dst->op_params + 4, sizeof(float));
memcpy(&freq_scale, (int32_t *) dst->params + 5, sizeof(float)); memcpy(&freq_scale, (int32_t *) dst->op_params + 5, sizeof(float));
assert(n_past >= 0); assert(n_past >= 0);
@ -12039,12 +12036,12 @@ static void ggml_compute_forward_rope_f16(
float freq_base; float freq_base;
float freq_scale; float freq_scale;
const int n_past = ((int32_t *) dst->params)[0]; const int n_past = ((int32_t *) dst->op_params)[0];
const int n_dims = ((int32_t *) dst->params)[1]; const int n_dims = ((int32_t *) dst->op_params)[1];
const int mode = ((int32_t *) dst->params)[2]; const int mode = ((int32_t *) dst->op_params)[2];
const int n_ctx = ((int32_t *) dst->params)[3]; const int n_ctx = ((int32_t *) dst->op_params)[3];
memcpy(&freq_base, (int32_t *) dst->params + 4, sizeof(float)); memcpy(&freq_base, (int32_t *) dst->op_params + 4, sizeof(float));
memcpy(&freq_scale, (int32_t *) dst->params + 5, sizeof(float)); memcpy(&freq_scale, (int32_t *) dst->op_params + 5, sizeof(float));
assert(n_past >= 0); assert(n_past >= 0);
@ -15810,10 +15807,9 @@ static void ggml_visit_parents(struct ggml_cgraph * cgraph, struct ggml_tensor *
} }
// check if already visited // check if already visited
if (node->visited) { if (node->node_id != -1) {
return; return;
} }
node->visited = true;
for (int i = 0; i < GGML_MAX_SRC; ++i) { for (int i = 0; i < GGML_MAX_SRC; ++i) {
if (node->src[i]) { if (node->src[i]) {
@ -15821,6 +15817,8 @@ static void ggml_visit_parents(struct ggml_cgraph * cgraph, struct ggml_tensor *
} }
} }
//node->id = cgraph->n_nodes + cgraph->n_leafs;
// TODO: add ggml_dependency instead of checking for NULL // TODO: add ggml_dependency instead of checking for NULL
if (node->op == GGML_OP_NONE && node->src[0] == NULL && node->src[1] == NULL && node->grad == NULL) { if (node->op == GGML_OP_NONE && node->src[0] == NULL && node->src[1] == NULL && node->grad == NULL) {
// reached a leaf node, not part of the gradient graph (e.g. a constant) // reached a leaf node, not part of the gradient graph (e.g. a constant)
@ -15830,6 +15828,7 @@ static void ggml_visit_parents(struct ggml_cgraph * cgraph, struct ggml_tensor *
ggml_format_name(node, "leaf_%d", cgraph->n_leafs); ggml_format_name(node, "leaf_%d", cgraph->n_leafs);
} }
node->node_id = cgraph->n_leafs;
cgraph->leafs[cgraph->n_leafs] = node; cgraph->leafs[cgraph->n_leafs] = node;
cgraph->n_leafs++; cgraph->n_leafs++;
} else { } else {
@ -15839,6 +15838,7 @@ static void ggml_visit_parents(struct ggml_cgraph * cgraph, struct ggml_tensor *
ggml_format_name(node, "node_%d", cgraph->n_nodes); ggml_format_name(node, "node_%d", cgraph->n_nodes);
} }
node->node_id = cgraph->n_nodes;
cgraph->nodes[cgraph->n_nodes] = node; cgraph->nodes[cgraph->n_nodes] = node;
cgraph->grads[cgraph->n_nodes] = node->grad; cgraph->grads[cgraph->n_nodes] = node->grad;
cgraph->n_nodes++; cgraph->n_nodes++;
@ -15872,10 +15872,10 @@ void ggml_build_forward_expand(struct ggml_cgraph * cgraph, struct ggml_tensor *
// TODO: this can be removed when ggml_build_forward_expand is removed // TODO: this can be removed when ggml_build_forward_expand is removed
void ggml_graph_close(struct ggml_cgraph * cgraph) { void ggml_graph_close(struct ggml_cgraph * cgraph) {
for (int i = 0; i < cgraph->n_nodes; ++i) { for (int i = 0; i < cgraph->n_nodes; ++i) {
cgraph->nodes[i]->visited = false; cgraph->nodes[i]->node_id = -1;
} }
for (int i = 0; i < cgraph->n_leafs; ++i) { for (int i = 0; i < cgraph->n_leafs; ++i) {
cgraph->leafs[i]->visited = false; cgraph->leafs[i]->node_id = -1;
} }
cgraph->closed = true; cgraph->closed = true;
} }

9
ggml.h
View File

@ -414,19 +414,21 @@ extern "C" {
// compute data // compute data
enum ggml_op op; enum ggml_op op;
// op params - allocated as int32_t for alignment
int32_t op_params[GGML_MAX_OP_PARAMS / sizeof(uint32_t)];
bool is_param; bool is_param;
struct ggml_tensor * grad; struct ggml_tensor * grad;
struct ggml_tensor * src[GGML_MAX_SRC]; struct ggml_tensor * src[GGML_MAX_SRC];
int node_id; // used to build graphs
// performance // performance
int perf_runs; int perf_runs;
int64_t perf_cycles; int64_t perf_cycles;
int64_t perf_time_us; int64_t perf_time_us;
// op params
// allocated as int32_t to avoid alignment issues
int32_t params[GGML_MAX_OP_PARAMS / sizeof(uint32_t)];
void * data; void * data;
@ -434,7 +436,6 @@ extern "C" {
void * extra; // extra things e.g. for ggml-cuda.cu void * extra; // extra things e.g. for ggml-cuda.cu
bool visited; // used to build graphs
char padding[4]; char padding[4];
}; };

183
llama.cpp
View File

@ -172,7 +172,7 @@ struct llama_kv_cache {
struct ggml_context * ctx = NULL; struct ggml_context * ctx = NULL;
ggml_buffer buf; ggml_buffer * buf;
int n; // number of tokens currently in the cache int n; // number of tokens currently in the cache
@ -225,29 +225,29 @@ struct llama_model {
llama_vocab vocab; llama_vocab vocab;
// backends // backends
ggml_backend backend_cpu; ggml_backend * backend_cpu = NULL;
ggml_buffer buf_cpu; ggml_buffer * buf_cpu = NULL;
ggml_context * ctx_cpu = NULL; ggml_context * ctx_cpu = NULL;
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
ggml_backend backend_cuda; ggml_backend * backend_cuda = NULL;
ggml_buffer buf_cuda; ggml_buffer * buf_cuda = NULL;
ggml_context * ctx_cuda = NULL; ggml_context * ctx_cuda = NULL;
#endif #endif
// backend assigned to each layer // backend assigned to each layer
ggml_backend * backend_input = NULL; ggml_backend * backend_inp = NULL;
ggml_backend * backend_output = NULL; ggml_backend * backend_out = NULL;
std::vector<ggml_backend *> backend_layers; std::vector<ggml_backend *> backend_layers;
~llama_model() { ~llama_model() {
if (ctx_cpu) { if (ctx_cpu) {
ggml_free(ctx_cpu); ggml_free(ctx_cpu);
ggml_backend_free_buffer(&buf_cpu); ggml_buffer_free(buf_cpu);
} }
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if (ctx_cuda) { if (ctx_cuda) {
ggml_free(ctx_cuda); ggml_free(ctx_cuda);
ggml_backend_free_buffer(&buf_cuda); ggml_buffer_free(buf_cuda);
} }
#endif #endif
} }
@ -286,9 +286,9 @@ struct llama_context {
std::vector<float> embedding; std::vector<float> embedding;
// memory buffers used to evaluate the model // memory buffers used to evaluate the model
ggml_buffer buf_compute_cpu = {}; ggml_buffer * buf_compute_cpu;
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
ggml_buffer buf_compute_cuda = {}; ggml_buffer * buf_compute_cuda;
#endif #endif
// input tensors // input tensors
@ -300,8 +300,19 @@ struct llama_context {
struct ggml_tensor * graph_embeddings_out = nullptr; struct ggml_tensor * graph_embeddings_out = nullptr;
// buffers to store the inputs and outputs of the graphs // buffers to store the inputs and outputs of the graphs
ggml_buffer buf_input = {}; ggml_buffer * buf_input;
ggml_buffer buf_output = {}; ggml_buffer * buf_output;
/*
~llama_context() {
if (model_owner) {
delete &model;
}
if (buf_compute_cpu) {
ggml_buffer_free(buf_compute_cpu);
}
}
*/
}; };
template <typename T> template <typename T>
@ -601,9 +612,6 @@ struct llama_model_loader {
void load_all_data(llama_progress_callback progress_callback, void * progress_callback_user_data, llama_mlock * lmlock) { void load_all_data(llama_progress_callback progress_callback, void * progress_callback_user_data, llama_mlock * lmlock) {
size_t data_size = 0; size_t data_size = 0;
size_t lock_size = 0; size_t lock_size = 0;
for (const llama_load_tensor & lt : tensors_map.tensors) {
data_size += lt.size;
}
if (use_mmap) { if (use_mmap) {
mapping.reset(new llama_mmap(&file_loader->file, false, ggml_is_numa())); mapping.reset(new llama_mmap(&file_loader->file, false, ggml_is_numa()));
@ -613,14 +621,28 @@ struct llama_model_loader {
} }
size_t done_size = 0; size_t done_size = 0;
std::vector<uint8_t> tmp_buf; std::vector<uint8_t> load_buf;
size_t load_buf_size = 0;
for (llama_load_tensor & lt : tensors_map.tensors) {
bool is_cpu = lt.ggml_tensor->backend == model->backend_cpu;
if (!use_mmap && !is_cpu) {
load_buf_size = std::max(load_buf_size, lt.size);
}
data_size += lt.size;
}
if (load_buf_size > 0) {
load_buf.resize(load_buf_size);
// may improve CUDA loading speed without mmap
//ggml_cuda_host_register(load_buf.data(), load_buf.size());
}
for (llama_load_tensor & lt : tensors_map.tensors) { for (llama_load_tensor & lt : tensors_map.tensors) {
if (progress_callback) { if (progress_callback) {
progress_callback((float) done_size / data_size, progress_callback_user_data); progress_callback((float) done_size / data_size, progress_callback_user_data);
} }
LLAMA_ASSERT(lt.ggml_tensor); // unused tensors should have been caught by load_data already LLAMA_ASSERT(lt.ggml_tensor); // unused tensors should have been caught by load_data already
bool is_cpu = lt.ggml_tensor->backend == &model->backend_cpu; bool is_cpu = lt.ggml_tensor->backend == model->backend_cpu;
// select buffer to load data into // select buffer to load data into
if (!use_mmap) { if (!use_mmap) {
@ -628,8 +650,7 @@ struct llama_model_loader {
lt.data = (uint8_t *) lt.ggml_tensor->data; lt.data = (uint8_t *) lt.ggml_tensor->data;
} else { } else {
// read to temporary buffer // read to temporary buffer
tmp_buf.resize(lt.size); lt.data = (uint8_t *) load_buf.data();
lt.data = (uint8_t *) tmp_buf.data();
} }
} }
@ -645,7 +666,7 @@ struct llama_model_loader {
} }
} }
} else { } else {
ggml_backend_set_tensor(lt.ggml_tensor, lt.data, 0, lt.size); ggml_backend_tensor_set(lt.ggml_tensor, lt.data, 0, lt.size);
if (use_mmap) { if (use_mmap) {
// hint the OS that we don't need the data anymore // hint the OS that we don't need the data anymore
// TODO: this may be a bad idea with devices that use the system memory (Metal?) // TODO: this may be a bad idea with devices that use the system memory (Metal?)
@ -655,6 +676,9 @@ struct llama_model_loader {
done_size += lt.size; done_size += lt.size;
} }
//if (load_buf_size > 0) {
// ggml_cuda_host_unregister(load_buf.data());
//}
} }
void load_data_for(llama_load_tensor & lt) { void load_data_for(llama_load_tensor & lt) {
@ -701,11 +725,11 @@ static bool kv_cache_init(
size_t size = 2u*n_elements*ggml_type_size(wtype) + 2u*MB; size_t size = 2u*n_elements*ggml_type_size(wtype) + 2u*MB;
cache.buf = ggml_backend_alloc_buffer(backend, size, 2); cache.buf = ggml_buffer_alloc(backend, size, 2);
cache.n = 0; cache.n = 0;
struct ggml_init_params params = ggml_init_params_default(); struct ggml_init_params params = ggml_init_params_default();
params.buffer = &cache.buf; params.buffer = cache.buf;
cache.ctx = ggml_init(params); cache.ctx = ggml_init(params);
@ -771,7 +795,7 @@ void llama_backend_init(bool numa) {
// needed to initialize f16 tables // needed to initialize f16 tables
{ {
struct ggml_init_params params = ggml_init_params_default(); struct ggml_init_params params = ggml_init_params_default();
params.buffer = {0}; params.buffer = NULL;
struct ggml_context * ctx = ggml_init(params); struct ggml_context * ctx = ggml_init(params);
ggml_free(ctx); ggml_free(ctx);
} }
@ -940,30 +964,30 @@ static void llama_model_load_internal(
const uint32_t n_layer = hparams.n_layer; const uint32_t n_layer = hparams.n_layer;
model.backend_cpu = ggml_backend_cpu_init(); model.backend_cpu = ggml_backend_cpu_init();
ggml_backend * backend_gpu = &model.backend_cpu; // hack until we have a proper backend selection ggml_backend * backend_gpu = model.backend_cpu; // hack until we have a proper backend selection
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if (n_gpu_layers > 0) { if (n_gpu_layers > 0) {
model.backend_cuda = ggml_backend_cuda_init(); model.backend_cuda = ggml_backend_cuda_init();
backend_gpu = &model.backend_cuda; backend_gpu = model.backend_cuda;
} }
#endif #endif
// assign splits to the backends // assign splits to the backends
const int i_gpu_start = std::max(0, (int)n_layer - n_gpu_layers); const int i_gpu_start = std::max(0, (int)n_layer - n_gpu_layers);
model.backend_input = n_gpu_layers > (int)n_layer ? backend_gpu : &model.backend_cpu; model.backend_inp = n_gpu_layers > (int)n_layer ? backend_gpu : model.backend_cpu;
model.backend_output = n_gpu_layers > 0 ? backend_gpu : &model.backend_cpu; model.backend_out = n_gpu_layers > 0 ? backend_gpu : model.backend_cpu;
model.backend_layers.resize(n_layer); model.backend_layers.resize(n_layer);
std::fill(model.backend_layers.begin(), model.backend_layers.begin() + i_gpu_start, &model.backend_cpu); std::fill(model.backend_layers.begin(), model.backend_layers.begin() + i_gpu_start, model.backend_cpu);
std::fill(model.backend_layers.begin() + i_gpu_start, model.backend_layers.end(), backend_gpu); std::fill(model.backend_layers.begin() + i_gpu_start, model.backend_layers.end(), backend_gpu);
// calculate the size of each context // calculate the size of each context
std::unordered_map<struct ggml_backend *, size_t> ctx_sizes; std::unordered_map<struct ggml_backend *, size_t> ctx_sizes;
for (const llama_load_tensor & lt : ml->tensors_map.tensors) { for (const llama_load_tensor & lt : ml->tensors_map.tensors) {
if (lt.name == "tok_embeddings.weight") { if (lt.name == "tok_embeddings.weight") {
ctx_sizes[model.backend_input] += lt.size; ctx_sizes[model.backend_inp] += lt.size;
} }
else if (lt.name == "norm.weight" || lt.name == "output.weight") { else if (lt.name == "norm.weight" || lt.name == "output.weight") {
ctx_sizes[model.backend_output] += lt.size; ctx_sizes[model.backend_out] += lt.size;
} }
else { else {
// parse layer number from name // parse layer number from name
@ -980,14 +1004,14 @@ static void llama_model_load_internal(
// TODO: generalize support for mmap // TODO: generalize support for mmap
size_t mmap_size = 0; size_t mmap_size = 0;
if (ml->use_mmap) { if (ml->use_mmap) {
mmap_size = ctx_sizes[&model.backend_cpu]; mmap_size = ctx_sizes[model.backend_cpu];
ctx_sizes[&model.backend_cpu] = 0; ctx_sizes[model.backend_cpu] = 0;
} }
fprintf(stderr, "%s: ggml ctx sizes:\n", __func__); fprintf(stderr, "%s: ggml ctx sizes:\n", __func__);
for (const auto & it : ctx_sizes) { for (const auto & it : ctx_sizes) {
fprintf(stderr, "%8s = %7.2f MB", ggml_backend_name(it.first), it.second / 1024.0 / 1024.0); fprintf(stderr, "%8s = %7.2f MB", ggml_backend_name(it.first), it.second / 1024.0 / 1024.0);
if (it.first == &model.backend_cpu && ml->use_mmap) { if (it.first == model.backend_cpu && ml->use_mmap) {
fprintf(stderr, " + %7.2f MB (mmap)", mmap_size / 1024.0 / 1024.0); fprintf(stderr, " + %7.2f MB (mmap)", mmap_size / 1024.0 / 1024.0);
} }
fprintf(stderr, "\n"); fprintf(stderr, "\n");
@ -996,10 +1020,10 @@ static void llama_model_load_internal(
// create the buffers and contexts // create the buffers and contexts
{ {
size_t cpu_num_tensors = ml->tensors_map.tensors.size(); size_t cpu_num_tensors = ml->tensors_map.tensors.size();
size_t ctx_size = ctx_sizes[&model.backend_cpu]; size_t ctx_size = ctx_sizes[model.backend_cpu];
model.buf_cpu = ggml_backend_alloc_buffer(&model.backend_cpu, ctx_size, cpu_num_tensors); model.buf_cpu = ggml_buffer_alloc(model.backend_cpu, ctx_size, cpu_num_tensors);
struct ggml_init_params params = ggml_init_params_default(); struct ggml_init_params params = ggml_init_params_default();
params.buffer = &model.buf_cpu; params.buffer = model.buf_cpu;
params.no_alloc = ml->use_mmap; params.no_alloc = ml->use_mmap;
model.ctx_cpu = ggml_init(params); model.ctx_cpu = ggml_init(params);
if (!model.ctx_cpu) { if (!model.ctx_cpu) {
@ -1011,10 +1035,10 @@ static void llama_model_load_internal(
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if (n_gpu_layers > 0) { if (n_gpu_layers > 0) {
size_t gpu_num_tensors = ml->tensors_map.tensors.size(); size_t gpu_num_tensors = ml->tensors_map.tensors.size();
size_t ctx_size = ctx_sizes[&model.backend_cuda]; size_t ctx_size = ctx_sizes[model.backend_cuda];
model.buf_cuda = ggml_backend_alloc_buffer(&model.backend_cuda, ctx_size, gpu_num_tensors); model.buf_cuda = ggml_buffer_alloc(model.backend_cuda, ctx_size, gpu_num_tensors);
struct ggml_init_params params = ggml_init_params_default(); struct ggml_init_params params = ggml_init_params_default();
params.buffer = &model.buf_cuda; params.buffer = model.buf_cuda;
model.ctx_cuda = ggml_init(params); model.ctx_cuda = ggml_init(params);
if (!model.ctx_cuda) { if (!model.ctx_cuda) {
throw std::runtime_error(format("ggml_init() failed for CUDA backend")); throw std::runtime_error(format("ggml_init() failed for CUDA backend"));
@ -1025,9 +1049,9 @@ static void llama_model_load_internal(
// TODO: clean this // TODO: clean this
ggml_context * ctx_input = model.ctx_cpu; ggml_context * ctx_input = model.ctx_cpu;
if (model.backend_input == backend_gpu) ctx_input = ctx_gpu; if (model.backend_inp == backend_gpu) ctx_input = ctx_gpu;
ggml_context * ctx_output = model.ctx_cpu; ggml_context * ctx_output = model.ctx_cpu;
if (model.backend_output == backend_gpu) ctx_output = ctx_gpu; if (model.backend_out == backend_gpu) ctx_output = ctx_gpu;
std::vector<ggml_context *> ctx_layers(n_layer, model.ctx_cpu); std::vector<ggml_context *> ctx_layers(n_layer, model.ctx_cpu);
for (uint32_t i = 0; i < n_layer; ++i) { for (uint32_t i = 0; i < n_layer; ++i) {
if (model.backend_layers[i] == backend_gpu) { if (model.backend_layers[i] == backend_gpu) {
@ -1181,18 +1205,18 @@ static ggml_graph_splits llama_build_graph(
// initialize contexts for every backend // initialize contexts for every backend
struct ggml_context * ctx_cpu = nullptr; struct ggml_context * ctx_cpu = nullptr;
if (lctx.buf_compute_cpu.mem_size > 0) { if (lctx.buf_compute_cpu != nullptr) {
struct ggml_init_params params = ggml_init_params_default(); struct ggml_init_params params = ggml_init_params_default();
params.buffer = &lctx.buf_compute_cpu; params.buffer = lctx.buf_compute_cpu;
params.compute_type = compute_type; params.compute_type = compute_type;
ctx_cpu = ggml_init(params); ctx_cpu = ggml_init(params);
} }
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
struct ggml_context * ctx_cuda = nullptr; struct ggml_context * ctx_cuda = nullptr;
if (lctx.buf_compute_cuda.mem_size > 0) { if (lctx.buf_compute_cuda != nullptr) {
struct ggml_init_params params = ggml_init_params_default(); struct ggml_init_params params = ggml_init_params_default();
params.buffer = &lctx.buf_compute_cuda; params.buffer = lctx.buf_compute_cuda;
params.compute_type = compute_type; params.compute_type = compute_type;
ctx_cuda = ggml_init(params); ctx_cuda = ggml_init(params);
} }
@ -1204,26 +1228,30 @@ static ggml_graph_splits llama_build_graph(
struct ggml_context * ctx_o = nullptr; struct ggml_context * ctx_o = nullptr;
struct ggml_context * ctx_kv = nullptr; struct ggml_context * ctx_kv = nullptr;
if (lctx.model.backend_input == &lctx.model.backend_cpu) ctx_i = ctx_cpu; if (lctx.model.backend_inp == lctx.model.backend_cpu) ctx_i = ctx_cpu;
if (lctx.model.backend_output == &lctx.model.backend_cpu) ctx_o = ctx_cpu; if (lctx.model.backend_out == lctx.model.backend_cpu) ctx_o = ctx_cpu;
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if (lctx.model.backend_input == &lctx.model.backend_cuda) ctx_i = ctx_cuda; if (lctx.model.backend_inp == lctx.model.backend_cuda) ctx_i = ctx_cuda;
if (lctx.model.backend_output == &lctx.model.backend_cuda) ctx_o = ctx_cuda; if (lctx.model.backend_out == lctx.model.backend_cuda) ctx_o = ctx_cuda;
#endif #endif
for (int il = 0; il < n_layer; il++) { for (int il = 0; il < n_layer; il++) {
if (lctx.model.backend_layers[il] == &lctx.model.backend_cpu) ctx_ls[il] = ctx_cpu; if (lctx.model.backend_layers[il] == lctx.model.backend_cpu) ctx_ls[il] = ctx_cpu;
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if (lctx.model.backend_layers[il] == &lctx.model.backend_cuda) ctx_ls[il] = ctx_cuda; if (lctx.model.backend_layers[il] == lctx.model.backend_cuda) ctx_ls[il] = ctx_cuda;
#endif #endif
} }
if (lctx.backend_kv == &lctx.model.backend_cpu) ctx_kv = ctx_cpu; if (lctx.backend_kv == lctx.model.backend_cpu) ctx_kv = ctx_cpu;
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if (lctx.backend_kv == &lctx.model.backend_cuda) ctx_kv = ctx_cuda; if (lctx.backend_kv == lctx.model.backend_cuda) ctx_kv = ctx_cuda;
#endif #endif
struct ggml_tensor * inpL; struct ggml_tensor * inpL;
// reuse the scale tensor for all layers since it requires a memory transfer
struct ggml_tensor * KQ_scale = ggml_new_f32(ctx_kv, 1.0f/sqrtf(float(n_embd)/n_head));
ggml_set_name(KQ_scale, "1/sqrt(n_embd/n_head)");
if (embeddings_input) { if (embeddings_input) {
// use embeddings as input // use embeddings as input
struct ggml_tensor * embd_in = lctx.graph_embeddings_in; struct ggml_tensor * embd_in = lctx.graph_embeddings_in;
@ -1236,10 +1264,6 @@ static ggml_graph_splits llama_build_graph(
inpL = ggml_get_rows(ctx_i, model.tok_embeddings, token_in); inpL = ggml_get_rows(ctx_i, model.tok_embeddings, token_in);
} }
// reuse the scale tensor for all layers since it requires a memory transfer
struct ggml_tensor * KQ_scale = ggml_new_f32(ctx_kv, 1.0f/sqrtf(float(n_embd)/n_head));
ggml_set_name(KQ_scale, "1/sqrt(n_embd/n_head)");
struct ggml_tensor * cur = nullptr; struct ggml_tensor * cur = nullptr;
for (int il = 0; il < n_layer; ++il) { for (int il = 0; il < n_layer; ++il) {
struct ggml_context * ctx_l = ctx_ls[il]; struct ggml_context * ctx_l = ctx_ls[il];
@ -1540,16 +1564,16 @@ static bool llama_eval_internal(
// for big prompts, if BLAS is enabled, it is better to use only one thread // for big prompts, if BLAS is enabled, it is better to use only one thread
// otherwise, the threads are spin-lock waiting for the BLAS calls and are degrading the performance // otherwise, the threads are spin-lock waiting for the BLAS calls and are degrading the performance
n_threads = N >= 32 && ggml_cpu_has_blas() ? 1 : n_threads; n_threads = N >= 32 && ggml_cpu_has_blas() ? 1 : n_threads;
ggml_backend_cpu_set_n_threads(const_cast<ggml_backend*>(&model.backend_cpu), n_threads); ggml_backend_cpu_set_n_threads(const_cast<ggml_backend*>(model.backend_cpu), n_threads);
struct ggml_graph_splits splits = llama_build_graph(lctx, N, n_past, embd_input); struct ggml_graph_splits splits = llama_build_graph(lctx, N, n_past, embd_input);
if (tokens != nullptr) { if (tokens != nullptr) {
// copy the tokens to the input tensor // copy the tokens to the input tensor
ggml_backend_set_tensor_async(lctx.graph_tokens_in, tokens, 0, N*ggml_element_size(lctx.graph_tokens_in)); ggml_backend_tensor_set_async(lctx.graph_tokens_in, tokens, 0, N*ggml_element_size(lctx.graph_tokens_in));
} else { } else {
// copy the embeddings to the input tensor // copy the embeddings to the input tensor
ggml_backend_set_tensor_async(lctx.graph_embeddings_in, embd, 0, N*n_embd*ggml_element_size(lctx.graph_embeddings_in)); ggml_backend_tensor_set_async(lctx.graph_embeddings_in, embd, 0, N*n_embd*ggml_element_size(lctx.graph_embeddings_in));
} }
// run the computation // run the computation
@ -1577,11 +1601,11 @@ static bool llama_eval_internal(
if (lctx.logits_all) { if (lctx.logits_all) {
logits_out.resize(n_vocab * N); logits_out.resize(n_vocab * N);
ggml_backend_get_tensor_async(lctx.graph_logits, logits_out.data(), 0, N*n_vocab*sizeof(float)); ggml_backend_tensor_get_async(lctx.graph_logits, logits_out.data(), 0, N*n_vocab*sizeof(float));
} else { } else {
// return result for just the last token // return result for just the last token
logits_out.resize(n_vocab); logits_out.resize(n_vocab);
ggml_backend_get_tensor_async(lctx.graph_logits, logits_out.data(), 0, n_vocab*sizeof(float)); ggml_backend_tensor_get_async(lctx.graph_logits, logits_out.data(), 0, n_vocab*sizeof(float));
} }
} }
@ -1589,13 +1613,13 @@ static bool llama_eval_internal(
if (!lctx.embedding.empty()) { if (!lctx.embedding.empty()) {
auto & embedding_out = lctx.embedding; auto & embedding_out = lctx.embedding;
embedding_out.resize(n_embd); embedding_out.resize(n_embd);
ggml_backend_get_tensor_async(lctx.graph_embeddings_out, embedding_out.data(), 0, n_embd*sizeof(float)); ggml_backend_tensor_get_async(lctx.graph_embeddings_out, embedding_out.data(), 0, n_embd*sizeof(float));
} }
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
// wait for the async copy to finish // wait for the async copy to finish
if (lctx.model.n_gpu_layers > 0) { if (lctx.model.n_gpu_layers > 0) {
ggml_backend_synchronize(const_cast<ggml_backend*>(&lctx.model.backend_cuda)); ggml_backend_synchronize(const_cast<ggml_backend*>(lctx.model.backend_cuda));
} }
#endif #endif
@ -2063,7 +2087,7 @@ void llama_sample_classifier_free_guidance(
struct llama_context * guidance_ctx, struct llama_context * guidance_ctx,
float scale, float scale,
float smooth_factor) { float smooth_factor) {
int64_t t_start_sample_us = t_start_sample_us = ggml_time_us(); int64_t t_start_sample_us = ggml_time_us();
assert(ctx); assert(ctx);
auto n_vocab = llama_n_vocab(ctx); auto n_vocab = llama_n_vocab(ctx);
@ -2608,13 +2632,13 @@ struct llama_context * llama_new_context_with_model(
// TODO: choose backend depending on n_layers/low_vram // TODO: choose backend depending on n_layers/low_vram
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if ((uint32_t)params.n_gpu_layers >= model->hparams.n_layer/2) { if ((uint32_t)params.n_gpu_layers >= model->hparams.n_layer/2 && !params.low_vram) {
ctx->backend_kv = &model->backend_cuda; ctx->backend_kv = model->backend_cuda;
} else { } else {
ctx->backend_kv = &model->backend_cpu; ctx->backend_kv = model->backend_cpu;
} }
#else #else
ctx->backend_kv = &model->backend_cpu; ctx->backend_kv = model->backend_cpu;
#endif #endif
ggml_type memory_type = params.f16_kv ? GGML_TYPE_F16 : GGML_TYPE_F32; ggml_type memory_type = params.f16_kv ? GGML_TYPE_F16 : GGML_TYPE_F32;
@ -2639,10 +2663,12 @@ struct llama_context * llama_new_context_with_model(
} }
// TODO: size the buffers more accurately - depends on improved memory management // TODO: size the buffers more accurately - depends on improved memory management
ctx->buf_compute_cpu = ggml_backend_alloc_buffer(&model->backend_cpu, MEM_REQ_EVAL().at(ctx->model.type), 2048); ctx->buf_compute_cpu = ggml_buffer_alloc(model->backend_cpu, MEM_REQ_EVAL().at(ctx->model.type), 2048);
// TODO: pinned memory for faster host-device transfers
//ggml_cuda_host_register(*(void**)ctx->buf_compute_cpu.backend_buffer, MEM_REQ_EVAL().at(ctx->model.type) + 128*2048);
#ifdef GGML_USE_CUDA #ifdef GGML_USE_CUDA
if (params.n_gpu_layers > 0) { if (params.n_gpu_layers > 0) {
ctx->buf_compute_cuda = ggml_backend_alloc_buffer(&model->backend_cuda, MEM_REQ_EVAL().at(ctx->model.type), 2048); ctx->buf_compute_cuda = ggml_buffer_alloc(model->backend_cuda, MEM_REQ_EVAL().at(ctx->model.type), 2048);
} }
#endif #endif
@ -2653,10 +2679,10 @@ struct llama_context * llama_new_context_with_model(
buf_input_size += hparams.n_ctx * ggml_type_size(GGML_TYPE_F32); // input tokens buf_input_size += hparams.n_ctx * ggml_type_size(GGML_TYPE_F32); // input tokens
// TODO: input embeddings should be optional to save memory // TODO: input embeddings should be optional to save memory
buf_input_size += hparams.n_embd * hparams.n_ctx * ggml_type_size(GGML_TYPE_F32); // input embeddings buf_input_size += hparams.n_embd * hparams.n_ctx * ggml_type_size(GGML_TYPE_F32); // input embeddings
ctx->buf_input = ggml_backend_alloc_buffer(model->backend_input, buf_input_size, 2); ctx->buf_input = ggml_buffer_alloc(model->backend_inp, buf_input_size, 2);
struct ggml_init_params ggml_params = ggml_init_params_default(); struct ggml_init_params ggml_params = ggml_init_params_default();
ggml_params.buffer = &ctx->buf_input; ggml_params.buffer = ctx->buf_input;
ggml_context * ctx0 = ggml_init(ggml_params); ggml_context * ctx0 = ggml_init(ggml_params);
ctx->graph_tokens_in = ggml_new_tensor_1d(ctx0, GGML_TYPE_I32, hparams.n_ctx); ctx->graph_tokens_in = ggml_new_tensor_1d(ctx0, GGML_TYPE_I32, hparams.n_ctx);
@ -2677,10 +2703,10 @@ struct llama_context * llama_new_context_with_model(
if (params.embedding) { if (params.embedding) {
buf_output_size += hparams.n_embd * ggml_type_size(GGML_TYPE_F32); buf_output_size += hparams.n_embd * ggml_type_size(GGML_TYPE_F32);
} }
ctx->buf_output = ggml_backend_alloc_buffer(model->backend_output, buf_output_size, 2); ctx->buf_output = ggml_buffer_alloc(model->backend_out, buf_output_size, 2);
struct ggml_init_params ggml_params = ggml_init_params_default(); struct ggml_init_params ggml_params = ggml_init_params_default();
ggml_params.buffer = &ctx->buf_output; ggml_params.buffer = ctx->buf_output;
ggml_context * ctx0 = ggml_init(ggml_params); ggml_context * ctx0 = ggml_init(ggml_params);
ctx->graph_logits = ggml_new_tensor_2d(ctx0, GGML_TYPE_F32, hparams.n_vocab, params.logits_all ? hparams.n_ctx : 1); ctx->graph_logits = ggml_new_tensor_2d(ctx0, GGML_TYPE_F32, hparams.n_vocab, params.logits_all ? hparams.n_ctx : 1);
@ -2706,7 +2732,7 @@ struct llama_context * llama_new_context_with_model(
} }
fprintf(stderr, "%s: layer backends: ", __func__); fprintf(stderr, "%s: layer backends: ", __func__);
fprintf(stderr, "input: %s, ", ggml_backend_name(ctx->model.backend_input)); fprintf(stderr, "input: %s, ", ggml_backend_name(ctx->model.backend_inp));
int start = 0; int start = 0;
struct ggml_backend * prev_backend = ctx->model.backend_layers[0]; struct ggml_backend * prev_backend = ctx->model.backend_layers[0];
@ -2721,7 +2747,7 @@ struct llama_context * llama_new_context_with_model(
prev_backend = ctx->model.backend_layers[i]; prev_backend = ctx->model.backend_layers[i];
} }
} }
fprintf(stderr, "output: %s, ", ggml_backend_name(ctx->model.backend_output)); fprintf(stderr, "output: %s, ", ggml_backend_name(ctx->model.backend_out));
fprintf(stderr, "kv: %s\n", ggml_backend_name(ctx->backend_kv)); fprintf(stderr, "kv: %s\n", ggml_backend_name(ctx->backend_kv));
#ifdef GGML_USE_MPI #ifdef GGML_USE_MPI
@ -2753,6 +2779,7 @@ struct llama_context * llama_init_from_file(
} }
void llama_free(struct llama_context * ctx) { void llama_free(struct llama_context * ctx) {
// TODO: free buffers - move this to destructor like llama_model
if (ctx->model_owner) { if (ctx->model_owner) {
delete &ctx->model; delete &ctx->model;
} }