root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-24 10:24:35 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity

backend : offload large batches to GPU (#6083)

Browse Source 
* backend : offload large batches to GPU

* fix hip

* code cleanup

* fix CUDA split buffers

* Update ggml-backend-impl.h

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>

* cuda : fix memset without set_device

* imatrix : remove sched affix from weight names

* sched : add a new split if the current one has too many inputs
reduce max inputs per split
more cleanup

* update backends

ggml-ci

---------

Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
This commit is contained in:
slaren 2024-03-18 11:03:04 +01:00 committed by GitHub
parent 496bc79bc2
commit 2bf8d0f7c4
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
14 changed files with 349 additions and 396 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

32
examples/imatrix/imatrix.cpp
View File

@ -56,13 +56,31 @@ bool IMatrixCollector::collect_imatrix(struct ggml_tensor * t, bool ask, void *
const struct ggml_tensor * src0 = t->src[0]; const struct ggml_tensor * src0 = t->src[0];
const struct ggml_tensor * src1 = t->src[1]; const struct ggml_tensor * src1 = t->src[1];
std::string wname;
{
// remove any prefix and suffixes from the name
// CUDA0#blk.0.attn_k.weight#0 => blk.0.attn_k.weight
const char * p = strchr(src0->name, '#');
if (p != NULL) {
p = p + 1;
const char * q = strchr(p, '#');
if (q != NULL) {
wname = std::string(p, q - p);
} else {
wname = p;
}
} else {
wname = src0->name;
}
}
// when ask is true, the scheduler wants to know if we are interested in data from this tensor // when ask is true, the scheduler wants to know if we are interested in data from this tensor
// if we return true, a follow-up call will be made with ask=false in which we can do the actual collection // if we return true, a follow-up call will be made with ask=false in which we can do the actual collection
if (ask) { if (ask) {
if (t->op == GGML_OP_MUL_MAT_ID) return true; // collect all indirect matrix multiplications if (t->op == GGML_OP_MUL_MAT_ID) return true; // collect all indirect matrix multiplications
if (t->op != GGML_OP_MUL_MAT) return false; if (t->op != GGML_OP_MUL_MAT) return false;
if (src1->ne[1] < 16 || src1->type != GGML_TYPE_F32) return false; if (src1->ne[1] < 16 || src1->type != GGML_TYPE_F32) return false;
if (!(strncmp(src0->name, "blk.", 4) == 0 || (m_params.collect_output_weight && strcmp(src0->name, "output.weight") == 0))) return false; if (!(wname.substr(0, 4) == "blk." || (m_params.collect_output_weight && wname == "output.weight"))) return false;
return true; return true;
} }
@ -94,12 +112,12 @@ bool IMatrixCollector::collect_imatrix(struct ggml_tensor * t, bool ask, void *
// this is necessary to guarantee equal number of "ncall" for each tensor // this is necessary to guarantee equal number of "ncall" for each tensor
for (int ex = 0; ex < n_as; ++ex) { for (int ex = 0; ex < n_as; ++ex) {
src0 = t->src[2 + ex]; src0 = t->src[2 + ex];
auto& e = m_stats[src0->name]; auto& e = m_stats[wname];
if (e.values.empty()) { if (e.values.empty()) {
e.values.resize(src1->ne[0], 0); e.values.resize(src1->ne[0], 0);
} }
else if (e.values.size() != (size_t)src1->ne[0]) { else if (e.values.size() != (size_t)src1->ne[0]) {
fprintf(stderr, "Oops: inconsistent size for %s (%d vs %d)\n", src0->name, (int)e.values.size(), (int)src1->ne[0]); fprintf(stderr, "Oops: inconsistent size for %s (%d vs %d)\n", wname.c_str(), (int)e.values.size(), (int)src1->ne[0]);
exit(1); //GGML_ASSERT(false); exit(1); //GGML_ASSERT(false);
} }
// NOTE: since we select top-k experts, the number of calls for the expert tensors will be k times larger // NOTE: since we select top-k experts, the number of calls for the expert tensors will be k times larger
@ -107,7 +125,7 @@ bool IMatrixCollector::collect_imatrix(struct ggml_tensor * t, bool ask, void *
//if (idx == t->src[0]->ne[0] - 1) ++e.ncall; //if (idx == t->src[0]->ne[0] - 1) ++e.ncall;
++e.ncall; ++e.ncall;
if (m_params.verbosity > 1) { if (m_params.verbosity > 1) {
printf("%s[%d]: %32s, %s, %5d x %5d, %d\n", __func__, m_last_call, src0->name, ggml_op_name(t->op), (int)src1->ne[0], (int)src1->ne[1], (int)src1->type); printf("%s[%d]: %32s, %s, %5d x %5d, %d\n", __func__, m_last_call, wname.c_str(), ggml_op_name(t->op), (int)src1->ne[0], (int)src1->ne[1], (int)src1->type);
} }
for (int row = 0; row < (int)src1->ne[1]; ++row) { for (int row = 0; row < (int)src1->ne[1]; ++row) {
const int excur = m_ids[row*n_as + idx]; const int excur = m_ids[row*n_as + idx];
@ -129,17 +147,17 @@ bool IMatrixCollector::collect_imatrix(struct ggml_tensor * t, bool ask, void *
} }
} }
} else { } else {
auto& e = m_stats[src0->name]; auto& e = m_stats[wname];
if (e.values.empty()) { if (e.values.empty()) {
e.values.resize(src1->ne[0], 0); e.values.resize(src1->ne[0], 0);
} }
else if (e.values.size() != (size_t)src1->ne[0]) { else if (e.values.size() != (size_t)src1->ne[0]) {
fprintf(stderr, "Oops: inconsistent size for %s (%d vs %d)\n", src0->name, (int)e.values.size(), (int)src1->ne[0]); fprintf(stderr, "Oops: inconsistent size for %s (%d vs %d)\n", wname.c_str(), (int)e.values.size(), (int)src1->ne[0]);
exit(1); //GGML_ASSERT(false); exit(1); //GGML_ASSERT(false);
} }
++e.ncall; ++e.ncall;
if (m_params.verbosity > 1) { if (m_params.verbosity > 1) {
printf("%s[%d]: %32s, %s, %5d x %5d, %d\n", __func__, m_last_call, src0->name, ggml_op_name(t->op), (int)src1->ne[0], (int)src1->ne[1], (int)src1->type); printf("%s[%d]: %32s, %s, %5d x %5d, %d\n", __func__, m_last_call, wname.c_str(), ggml_op_name(t->op), (int)src1->ne[0], (int)src1->ne[1], (int)src1->type);
} }
for (int row = 0; row < (int)src1->ne[1]; ++row) { for (int row = 0; row < (int)src1->ne[1]; ++row) {
const float * x = data + row * src1->ne[0]; const float * x = data + row * src1->ne[0];

4
examples/llama-bench/llama-bench.cpp
View File

@ -114,10 +114,10 @@ static std::string get_cpu_info() {
static std::string get_gpu_info() { static std::string get_gpu_info() {
std::string id; std::string id;
#ifdef GGML_USE_CUBLAS #ifdef GGML_USE_CUBLAS
int count = ggml_cuda_get_device_count(); int count = ggml_backend_cuda_get_device_count();
for (int i = 0; i < count; i++) { for (int i = 0; i < count; i++) {
char buf[128]; char buf[128];
ggml_cuda_get_device_description(i, buf, sizeof(buf)); ggml_backend_cuda_get_device_description(i, buf, sizeof(buf));
id += buf; id += buf;
if (i < count - 1) { if (i < count - 1) {
id += "/"; id += "/";

10
ggml-alloc.c
View File

@ -548,7 +548,11 @@ static void ggml_gallocr_alloc_graph_impl(ggml_gallocr_t galloc, struct ggml_cgr
for (int i = 0; i < graph->n_nodes; i++) { for (int i = 0; i < graph->n_nodes; i++) {
struct ggml_tensor * node = graph->nodes[i]; struct ggml_tensor * node = graph->nodes[i];
if (ggml_is_view(node)) { // TODO: better way to add external dependencies
// GGML_OP_NONE does not appear normally in the graph nodes, but is used by ggml-backend to add dependencies to
// control when some tensors are allocated and freed. in this case, the dependencies are in `src`, but the node
// itself is never used and should not be considered a dependency
if (ggml_is_view(node) && node->op != GGML_OP_NONE) {
struct ggml_tensor * view_src = node->view_src; struct ggml_tensor * view_src = node->view_src;
ggml_gallocr_hash_get(galloc, view_src)->n_views += 1; ggml_gallocr_hash_get(galloc, view_src)->n_views += 1;
} }
@ -565,8 +569,8 @@ static void ggml_gallocr_alloc_graph_impl(ggml_gallocr_t galloc, struct ggml_cgr
ggml_gallocr_hash_get(galloc, src)->n_children += 1; ggml_gallocr_hash_get(galloc, src)->n_children += 1;
// allocate explicit inputs and leafs // allocate explicit inputs
if (src->flags & GGML_TENSOR_FLAG_INPUT || src->op == GGML_OP_NONE) { if (src->flags & GGML_TENSOR_FLAG_INPUT) {
ggml_gallocr_allocate_node(galloc, src, get_node_buffer_id(node_buffer_ids, i)); ggml_gallocr_allocate_node(galloc, src, get_node_buffer_id(node_buffer_ids, i));
} }
} }

5
ggml-backend-impl.h
View File

@ -103,6 +103,11 @@ extern "C" {
// check if the backend supports an operation // check if the backend supports an operation
bool (*GGML_CALL supports_op)(ggml_backend_t backend, const struct ggml_tensor * op); bool (*GGML_CALL supports_op)(ggml_backend_t backend, const struct ggml_tensor * op);
// check if the backend wants to run an operation, even if the weights are allocated in a CPU buffer
// these should be expensive operations with large batch sizes that may benefit from running on this backend
// even if the weight has to be copied from the CPU temporarily
bool (*GGML_CALL offload_op)(ggml_backend_t backend, const struct ggml_tensor * op);
// (optional) event synchronization // (optional) event synchronization
ggml_backend_event_t (*GGML_CALL event_new) (ggml_backend_t backend); ggml_backend_event_t (*GGML_CALL event_new) (ggml_backend_t backend);
void (*GGML_CALL event_free) (ggml_backend_event_t event); void (*GGML_CALL event_free) (ggml_backend_event_t event);

278
ggml-backend.c
View File

@ -278,7 +278,7 @@ enum ggml_status ggml_backend_graph_compute(ggml_backend_t backend, struct ggml_
return err; return err;
} }
bool ggml_backend_graph_compute_async(ggml_backend_t backend, struct ggml_cgraph * cgraph) { enum ggml_status ggml_backend_graph_compute_async(ggml_backend_t backend, struct ggml_cgraph * cgraph) {
return backend->iface.graph_compute(backend, cgraph); return backend->iface.graph_compute(backend, cgraph);
} }
@ -286,6 +286,13 @@ bool ggml_backend_supports_op(ggml_backend_t backend, const struct ggml_tensor *
return backend->iface.supports_op(backend, op); return backend->iface.supports_op(backend, op);
} }
bool ggml_backend_offload_op(ggml_backend_t backend, const struct ggml_tensor * op) {
if (backend->iface.offload_op != NULL) {
return backend->iface.offload_op(backend, op);
}
return false;
}
// backend copy // backend copy
static bool ggml_are_same_layout(const struct ggml_tensor * a, const struct ggml_tensor * b) { static bool ggml_are_same_layout(const struct ggml_tensor * a, const struct ggml_tensor * b) {
@ -761,6 +768,10 @@ GGML_CALL static ggml_backend_graph_plan_t ggml_backend_cpu_graph_plan_create(gg
if (cpu_plan->cplan.work_size > 0) { if (cpu_plan->cplan.work_size > 0) {
cpu_plan->cplan.work_data = malloc(cpu_plan->cplan.work_size); cpu_plan->cplan.work_data = malloc(cpu_plan->cplan.work_size);
if (cpu_plan->cplan.work_data == NULL) {
free(cpu_plan);
return NULL;
}
} }
cpu_plan->cplan.abort_callback = cpu_ctx->abort_callback; cpu_plan->cplan.abort_callback = cpu_ctx->abort_callback;
@ -834,6 +845,7 @@ static struct ggml_backend_i cpu_backend_i = {
/* .graph_plan_compute = */ ggml_backend_cpu_graph_plan_compute, /* .graph_plan_compute = */ ggml_backend_cpu_graph_plan_compute,
/* .graph_compute = */ ggml_backend_cpu_graph_compute, /* .graph_compute = */ ggml_backend_cpu_graph_compute,
/* .supports_op = */ ggml_backend_cpu_supports_op, /* .supports_op = */ ggml_backend_cpu_supports_op,
/* .offload_op = */ NULL,
/* .event_new = */ NULL, /* .event_new = */ NULL,
/* .event_free = */ NULL, /* .event_free = */ NULL,
/* .event_record = */ NULL, /* .event_record = */ NULL,
@ -999,11 +1011,11 @@ static bool ggml_is_view_op(enum ggml_op op) {
#endif #endif
#ifndef GGML_SCHED_MAX_SPLITS #ifndef GGML_SCHED_MAX_SPLITS
#define GGML_SCHED_MAX_SPLITS 256 #define GGML_SCHED_MAX_SPLITS 2048
#endif #endif
#ifndef GGML_SCHED_MAX_SPLIT_INPUTS #ifndef GGML_SCHED_MAX_SPLIT_INPUTS
#define GGML_SCHED_MAX_SPLIT_INPUTS 16 #define GGML_SCHED_MAX_SPLIT_INPUTS 4
#endif #endif
#ifndef GGML_SCHED_MAX_COPIES #ifndef GGML_SCHED_MAX_COPIES
@ -1043,8 +1055,9 @@ struct ggml_backend_sched {
struct ggml_cgraph * graph; struct ggml_cgraph * graph;
// graph splits // graph splits
struct ggml_backend_sched_split splits[GGML_SCHED_MAX_SPLITS]; struct ggml_backend_sched_split * splits;
int n_splits; int n_splits;
int splits_capacity;
// pipeline parallelism support // pipeline parallelism support
int n_copies; int n_copies;
@ -1114,40 +1127,48 @@ static int ggml_backend_sched_backend_id_from_cur(ggml_backend_sched_t sched, st
// TODO: use supports_op to check if the backend supports the op // TODO: use supports_op to check if the backend supports the op
// assign pre-allocated nodes to their backend // assign pre-allocated nodes to their backend
// dst int cur_backend_id = ggml_backend_sched_backend_from_buffer(sched, tensor);
int cur_backend = ggml_backend_sched_backend_from_buffer(sched, tensor); if (cur_backend_id != -1) {
if (cur_backend != -1) {
SET_CAUSE(tensor, "1.dst"); SET_CAUSE(tensor, "1.dst");
return cur_backend; return cur_backend_id;
} }
// view_src // view_src
if (tensor->view_src != NULL) { if (tensor->view_src != NULL) {
cur_backend = ggml_backend_sched_backend_from_buffer(sched, tensor->view_src); cur_backend_id = ggml_backend_sched_backend_from_buffer(sched, tensor->view_src);
if (cur_backend != -1) { if (cur_backend_id != -1) {
SET_CAUSE(tensor, "1.vsrc"); SET_CAUSE(tensor, "1.vsrc");
return cur_backend; return cur_backend_id;
} }
} }
// input // graph input
if (tensor->flags & GGML_TENSOR_FLAG_INPUT) { if (tensor->flags & GGML_TENSOR_FLAG_INPUT) {
cur_backend = sched->n_backends - 1; // last backend (assumed CPU) cur_backend_id = sched->n_backends - 1; // last backend (assumed CPU)
SET_CAUSE(tensor, "1.inp"); SET_CAUSE(tensor, "1.inp");
return cur_backend; return cur_backend_id;
} }
// assign nodes that use weights to the backend of the weights // assign nodes that use weights to the backend of the weights
// operations with weights are preferably run on the same backend as the weights
for (int i = 0; i < GGML_MAX_SRC; i++) { for (int i = 0; i < GGML_MAX_SRC; i++) {
const struct ggml_tensor * src = tensor->src[i]; const struct ggml_tensor * src = tensor->src[i];
if (src == NULL) { if (src == NULL) {
continue; continue;
} }
if (src->buffer != NULL && src->buffer->usage == GGML_BACKEND_BUFFER_USAGE_WEIGHTS) { if (src->buffer != NULL && src->buffer->usage == GGML_BACKEND_BUFFER_USAGE_WEIGHTS) {
int src_backend = ggml_backend_sched_backend_from_buffer(sched, src); int src_backend_id = ggml_backend_sched_backend_from_buffer(sched, src);
// operations with weights are always run on the same backend as the weights // check if a backend with higher prio wants to offload the op
if (src_backend_id == sched->n_backends - 1) {
for (int b = 0; b < src_backend_id; b++) {
if (ggml_backend_offload_op(sched->backends[b], tensor)) {
SET_CAUSE(tensor, "1.off");
return b;
}
}
}
SET_CAUSE(tensor, "1.wgt%d", i); SET_CAUSE(tensor, "1.wgt%d", i);
return src_backend; return src_backend_id;
} }
} }
@ -1227,28 +1248,31 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
// pass 1: assign backends to ops with pre-allocated inputs // pass 1: assign backends to ops with pre-allocated inputs
for (int i = 0; i < graph->n_leafs; i++) { for (int i = 0; i < graph->n_leafs; i++) {
struct ggml_tensor * leaf = graph->leafs[i]; struct ggml_tensor * leaf = graph->leafs[i];
if (tensor_backend_id(leaf) != -1) { int * leaf_backend_id = &tensor_backend_id(leaf);
if (*leaf_backend_id != -1) {
// do not overwrite user assignments // do not overwrite user assignments
continue; continue;
} }
tensor_backend_id(leaf) = ggml_backend_sched_backend_id_from_cur(sched, leaf); *leaf_backend_id = ggml_backend_sched_backend_id_from_cur(sched, leaf);
} }
for (int i = 0; i < graph->n_nodes; i++) { for (int i = 0; i < graph->n_nodes; i++) {
struct ggml_tensor * node = graph->nodes[i]; struct ggml_tensor * node = graph->nodes[i];
if (tensor_backend_id(node) != -1) { int * node_backend_id = &tensor_backend_id(node);
if (*node_backend_id != -1) {
// do not overwrite user assignments // do not overwrite user assignments
continue; continue;
} }
tensor_backend_id(node) = ggml_backend_sched_backend_id_from_cur(sched, node); *node_backend_id = ggml_backend_sched_backend_id_from_cur(sched, node);
// src // src
for (int j = 0; j < GGML_MAX_SRC; j++) { for (int j = 0; j < GGML_MAX_SRC; j++) {
struct ggml_tensor * src = node->src[j]; struct ggml_tensor * src = node->src[j];
if (src == NULL) { if (src == NULL) {
continue; continue;
} }
if (tensor_backend_id(src) == -1) { int * src_backend_id = &tensor_backend_id(src);
tensor_backend_id(src) = ggml_backend_sched_backend_id_from_cur(sched, src); if (*src_backend_id == -1) {
*src_backend_id = ggml_backend_sched_backend_id_from_cur(sched, src);
} }
} }
} }
@ -1270,21 +1294,20 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
if (ggml_is_view_op(node->op)) { if (ggml_is_view_op(node->op)) {
continue; continue;
} }
int tensor_backend_id = tensor_backend_id(node); int * node_backend_id = &tensor_backend_id(node);
if (tensor_backend_id != -1) { if (*node_backend_id != -1) {
if (tensor_backend_id == sched->n_backends - 1) { if (*node_backend_id == sched->n_backends - 1) {
// skip cpu (lowest prio backend) // skip cpu (lowest prio backend)
cur_backend_id = -1; cur_backend_id = -1;
} else { } else {
cur_backend_id = tensor_backend_id; cur_backend_id = *node_backend_id;
} }
} else { } else {
tensor_backend_id(node) = cur_backend_id; *node_backend_id = cur_backend_id;
SET_CAUSE(node, "2.2"); SET_CAUSE(node, "2.2");
} }
} }
} }
// pass 2.1 expand gpu up // pass 2.1 expand gpu up
{ {
int cur_backend_id = -1; int cur_backend_id = -1;
@ -1293,22 +1316,20 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
if (ggml_is_view_op(node->op)) { if (ggml_is_view_op(node->op)) {
continue; continue;
} }
int tensor_backend_id = tensor_backend_id(node); int * node_backend_id = &tensor_backend_id(node);
if (tensor_backend_id != -1) { if (*node_backend_id != -1) {
if (tensor_backend_id == sched->n_backends - 1) { if (*node_backend_id == sched->n_backends - 1) {
// skip cpu (lowest prio backend) // skip cpu (lowest prio backend)
cur_backend_id = -1; cur_backend_id = -1;
} else { } else {
cur_backend_id = tensor_backend_id; cur_backend_id = *node_backend_id;
} }
} else { } else {
tensor_backend_id(node) = cur_backend_id; *node_backend_id = cur_backend_id;
SET_CAUSE(node, "2.1"); SET_CAUSE(node, "2.1");
} }
} }
} }
// pass 2.4 expand rest down // pass 2.4 expand rest down
{ {
int cur_backend_id = -1; int cur_backend_id = -1;
@ -1317,16 +1338,16 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
if (ggml_is_view_op(node->op)) { if (ggml_is_view_op(node->op)) {
continue; continue;
} }
int tensor_backend_id = tensor_backend_id(node); int * node_backend_id = &tensor_backend_id(node);
if (tensor_backend_id != -1) { if (*node_backend_id != -1) {
cur_backend_id = tensor_backend_id; cur_backend_id = *node_backend_id;
} else { } else {
tensor_backend_id(node) = cur_backend_id; *node_backend_id = cur_backend_id;
SET_CAUSE(node, "2.4"); SET_CAUSE(node, "2.4");
} }
} }
} }
// pass 2.3 expand rest up // pass 2.3 expand rest up
{ {
int cur_backend_id = -1; int cur_backend_id = -1;
for (int i = graph->n_nodes - 1; i >= 0; i--) { for (int i = graph->n_nodes - 1; i >= 0; i--) {
@ -1334,11 +1355,11 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
if (ggml_is_view_op(node->op)) { if (ggml_is_view_op(node->op)) {
continue; continue;
} }
int tensor_backend_id = tensor_backend_id(node); int * node_backend_id = &tensor_backend_id(node);
if (tensor_backend_id != -1) { if (*node_backend_id != -1) {
cur_backend_id = tensor_backend_id; cur_backend_id = *node_backend_id;
} else { } else {
tensor_backend_id(node) = cur_backend_id; *node_backend_id = cur_backend_id;
SET_CAUSE(node, "2.3"); SET_CAUSE(node, "2.3");
} }
} }
@ -1351,9 +1372,9 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
// pass 3: assign backends to remaining src from dst and view_src // pass 3: assign backends to remaining src from dst and view_src
for (int i = 0; i < graph->n_nodes; i++) { for (int i = 0; i < graph->n_nodes; i++) {
struct ggml_tensor * node = graph->nodes[i]; struct ggml_tensor * node = graph->nodes[i];
int cur_backend_id = tensor_backend_id(node); int * cur_backend_id = &tensor_backend_id(node);
if (node->view_src != NULL && cur_backend_id == -1) { if (node->view_src != NULL && *cur_backend_id == -1) {
cur_backend_id = tensor_backend_id(node) = tensor_backend_id(node->view_src); *cur_backend_id = tensor_backend_id(node->view_src);
SET_CAUSE(node, "3.vsrc"); SET_CAUSE(node, "3.vsrc");
} }
for (int j = 0; j < GGML_MAX_SRC; j++) { for (int j = 0; j < GGML_MAX_SRC; j++) {
@ -1361,14 +1382,14 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
if (src == NULL) { if (src == NULL) {
continue; continue;
} }
int src_backend_id = tensor_backend_id(src); int * src_backend_id = &tensor_backend_id(src);
if (src_backend_id == -1) { if (*src_backend_id == -1) {
if (src->view_src != NULL) { if (src->view_src != NULL) {
// views are always on the same backend as the source // views are always on the same backend as the source
tensor_backend_id(src) = tensor_backend_id(src->view_src); *src_backend_id = tensor_backend_id(src->view_src);
SET_CAUSE(src, "3.vsrc"); SET_CAUSE(src, "3.vsrc");
} else { } else {
tensor_backend_id(src) = cur_backend_id; *src_backend_id = *cur_backend_id;
SET_CAUSE(src, "3.cur"); SET_CAUSE(src, "3.cur");
} }
} }
@ -1380,19 +1401,20 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
// pass 4: split graph, find tensors that need to be copied // pass 4: split graph, find tensors that need to be copied
{ {
int cur_split = 0; int i_split = 0;
struct ggml_backend_sched_split * split = &sched->splits[0];
// find the backend of the first split, skipping view ops // find the backend of the first split, skipping view ops
for (int i = 0; i < graph->n_nodes; i++) { for (int i = 0; i < graph->n_nodes; i++) {
struct ggml_tensor * node = graph->nodes[i]; struct ggml_tensor * node = graph->nodes[i];
if (!ggml_is_view_op(node->op)) { if (!ggml_is_view_op(node->op)) {
sched->splits[0].backend_id = tensor_backend_id(node); split->backend_id = tensor_backend_id(node);
break; break;
} }
} }
sched->splits[0].i_start = 0; split->i_start = 0;
sched->splits[0].n_inputs = 0; split->n_inputs = 0;
memset(sched->splits[0].inputs, 0, sizeof(sched->splits[0].inputs)); //HACK memset(split->inputs, 0, sizeof(split->inputs)); //HACK
int cur_backend_id = sched->splits[0].backend_id; int cur_backend_id = split->backend_id;
for (int i = 0; i < graph->n_nodes; i++) { for (int i = 0; i < graph->n_nodes; i++) {
struct ggml_tensor * node = graph->nodes[i]; struct ggml_tensor * node = graph->nodes[i];
@ -1400,18 +1422,54 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
continue; continue;
} }
int tensor_backend_id = tensor_backend_id(node); const int node_backend_id = tensor_backend_id(node);
GGML_ASSERT(tensor_backend_id != -1); // all nodes should be assigned by now GGML_ASSERT(node_backend_id != -1); // all nodes should be assigned by now
if (tensor_backend_id != cur_backend_id) { // check if we should start a new split based on the sources of the current node
sched->splits[cur_split].i_end = i; bool need_new_split = false;
cur_split++; if (node_backend_id == cur_backend_id && split->n_inputs > 0) {
GGML_ASSERT(cur_split < GGML_SCHED_MAX_SPLITS); for (int j = 0; j < GGML_MAX_SRC; j++) {
sched->splits[cur_split].backend_id = tensor_backend_id; struct ggml_tensor * src = node->src[j];
sched->splits[cur_split].i_start = i; if (src == NULL) {
sched->splits[cur_split].n_inputs = 0; continue;
cur_backend_id = tensor_backend_id; }
// check if a weight is on a different backend
// by starting a new split, the memory of the previously offloaded weights can be reused
if (src->buffer != NULL && src->buffer->usage == GGML_BACKEND_BUFFER_USAGE_WEIGHTS) {
int src_backend_id = tensor_backend_id(src);
if (src_backend_id != -1 && src_backend_id != cur_backend_id) {
need_new_split = true;
break;
}
}
// check if the split has too many inputs
if (split->n_inputs == GGML_SCHED_MAX_SPLIT_INPUTS) {
const size_t id = hash_id(src);
int src_backend_id = sched->tensor_backend_id[id];
if (src_backend_id != cur_backend_id && sched->tensor_copies[hash_id(src)][cur_backend_id][0] == NULL) {
//printf("starting new split because of too many inputs: node %s, input %s\n", node->name, src->name);
need_new_split = true;
break;
}
}
}
}
if (node_backend_id != cur_backend_id || need_new_split) {
split->i_end = i;
i_split++;
if (i_split >= sched->splits_capacity) {
sched->splits_capacity *= 2;
sched->splits = realloc(sched->splits, sched->splits_capacity * sizeof(struct ggml_backend_sched_split));
GGML_ASSERT(sched->splits != NULL);
}
GGML_ASSERT(i_split < GGML_SCHED_MAX_SPLITS);
split = &sched->splits[i_split];
split->backend_id = node_backend_id;
split->i_start = i;
split->n_inputs = 0;
cur_backend_id = node_backend_id;
} }
// find inputs that are not on the same backend // find inputs that are not on the same backend
@ -1421,10 +1479,10 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
continue; continue;
} }
int src_backend_id = tensor_backend_id(src); const int src_backend_id = tensor_backend_id(src);
assert(src_backend_id != -1); // all inputs should be assigned by now assert(src_backend_id != -1); // all inputs should be assigned by now
if (src->flags & GGML_TENSOR_FLAG_INPUT) { if (src->flags & GGML_TENSOR_FLAG_INPUT && sched->n_copies > 1) {
size_t id = hash_id(src); size_t id = hash_id(src);
if (sched->tensor_copies[id][src_backend_id][0] == NULL) { if (sched->tensor_copies[id][src_backend_id][0] == NULL) {
ggml_backend_t backend = sched->backends[src_backend_id]; ggml_backend_t backend = sched->backends[src_backend_id];
@ -1441,7 +1499,6 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
ggml_set_output(tensor_copy); // prevent ggml-alloc from overwriting the tensor ggml_set_output(tensor_copy); // prevent ggml-alloc from overwriting the tensor
} }
sched->tensor_copies[id][src_backend_id][c] = tensor_copy; sched->tensor_copies[id][src_backend_id][c] = tensor_copy;
tensor_backend_id(tensor_copy) = src_backend_id;
SET_CAUSE(tensor_copy, "4.cpy"); SET_CAUSE(tensor_copy, "4.cpy");
} }
int n_graph_inputs = sched->n_graph_inputs++; int n_graph_inputs = sched->n_graph_inputs++;
@ -1450,9 +1507,9 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
} }
} }
if (src_backend_id != tensor_backend_id) { if (src_backend_id != node_backend_id) {
// create a copy of the input in the split's backend // create a copy of the input in the split's backend
size_t id = hash_id(src); const size_t id = hash_id(src);
if (sched->tensor_copies[id][cur_backend_id][0] == NULL) { if (sched->tensor_copies[id][cur_backend_id][0] == NULL) {
ggml_backend_t backend = sched->backends[cur_backend_id]; ggml_backend_t backend = sched->backends[cur_backend_id];
for (int c = 0; c < sched->n_copies; c++) { for (int c = 0; c < sched->n_copies; c++) {
@ -1463,76 +1520,42 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
ggml_set_output(tensor_copy); // prevent ggml-alloc from overwriting the tensor ggml_set_output(tensor_copy); // prevent ggml-alloc from overwriting the tensor
} }
sched->tensor_copies[id][cur_backend_id][c] = tensor_copy; sched->tensor_copies[id][cur_backend_id][c] = tensor_copy;
tensor_backend_id(tensor_copy) = cur_backend_id;
SET_CAUSE(tensor_copy, "4.cpy"); SET_CAUSE(tensor_copy, "4.cpy");
} }
int n_inputs = sched->splits[cur_split].n_inputs++; int n_inputs = split->n_inputs++;
GGML_ASSERT(n_inputs < GGML_SCHED_MAX_SPLIT_INPUTS); GGML_ASSERT(n_inputs < GGML_SCHED_MAX_SPLIT_INPUTS);
sched->splits[cur_split].inputs[n_inputs] = src; split->inputs[n_inputs] = src;
} }
node->src[j] = sched->tensor_copies[id][cur_backend_id][sched->cur_copy]; node->src[j] = sched->tensor_copies[id][cur_backend_id][sched->cur_copy];
} }
} }
} }
sched->splits[cur_split].i_end = graph->n_nodes; split->i_end = graph->n_nodes;
sched->n_splits = cur_split + 1; sched->n_splits = i_split + 1;
} }
#ifdef DEBUG_PASS4 #ifdef DEBUG_PASS4
fprintf(stderr, "PASS 4 ASSIGNMENTS\n"); ggml_backend_sched_print_assignments(sched, graph); fprintf(stderr, "PASS 4 ASSIGNMENTS\n"); ggml_backend_sched_print_assignments(sched, graph);
#endif #endif
#ifndef NDEBUG
// sanity check: all sources should have the same backend as the node
for (int i = 0; i < graph->n_nodes; i++) {
struct ggml_tensor * node = graph->nodes[i];
ggml_backend_t tensor_backend = ggml_backend_sched_get_tensor_backend(sched, node);
if (tensor_backend == NULL) {
fprintf(stderr, "!!!!!!! %s has no backend\n", node->name);
}
if (node->view_src != NULL && tensor_backend != ggml_backend_sched_get_tensor_backend(sched, node->view_src)) {
fprintf(stderr, "!!!!!!! %s has backend %s, view_src %s has backend %s\n",
node->name, tensor_backend ? ggml_backend_name(tensor_backend) : "NULL",
node->view_src->name, ggml_backend_sched_get_tensor_backend(sched, node->view_src) ?
ggml_backend_name(ggml_backend_sched_get_tensor_backend(sched, node->view_src)) : "NULL");
}
for (int j = 0; j < GGML_MAX_SRC; j++) {
struct ggml_tensor * src = node->src[j];
if (src == NULL) {
continue;
}
ggml_backend_t src_backend = ggml_backend_sched_get_tensor_backend(sched, src);
if (src_backend != tensor_backend /* && src_backend != NULL */) {
fprintf(stderr, "!!!! %s has backend %s, src %d (%s) has backend %s\n",
node->name, tensor_backend ? ggml_backend_name(tensor_backend) : "NULL",
j, src->name, src_backend ? ggml_backend_name(src_backend) : "NULL");
}
if (src->view_src != NULL && src_backend != ggml_backend_sched_get_tensor_backend(sched, src->view_src)) {
fprintf(stderr, "!!!!!!! [src] %s has backend %s, view_src %s has backend %s\n",
src->name, src_backend ? ggml_backend_name(src_backend) : "NULL",
src->view_src->name, ggml_backend_sched_get_tensor_backend(sched, src->view_src) ?
ggml_backend_name(ggml_backend_sched_get_tensor_backend(sched, src->view_src)) : "NULL");
}
}
}
fflush(stderr);
#endif
// create copies of the graph for each split // create copies of the graph for each split
// TODO: avoid this copy // TODO: avoid this copy
struct ggml_cgraph * graph_copy = ggml_new_graph_custom(sched->ctx, graph->n_nodes + sched->n_splits*GGML_SCHED_MAX_SPLIT_INPUTS, false); struct ggml_cgraph * graph_copy = ggml_new_graph_custom(sched->ctx, graph->n_nodes + sched->n_splits*GGML_SCHED_MAX_SPLIT_INPUTS*2, false);
for (int i = 0; i < sched->n_splits; i++) { for (int i = 0; i < sched->n_splits; i++) {
struct ggml_backend_sched_split * split = &sched->splits[i]; struct ggml_backend_sched_split * split = &sched->splits[i];
split->graph = ggml_graph_view(graph, split->i_start, split->i_end); split->graph = ggml_graph_view(graph, split->i_start, split->i_end);
// add inputs to the graph copy so that they are allocated by ggml-alloc at the start of the split // add inputs to the graph copy so that they are allocated by ggml-alloc at the start of the split
for (int j = 0; j < split->n_inputs; j++) { for (int j = 0; j < split->n_inputs; j++) {
assert(graph_copy->size > (graph_copy->n_nodes + 1));
struct ggml_tensor * input = split->inputs[j]; struct ggml_tensor * input = split->inputs[j];
struct ggml_tensor * input_cpy = sched->tensor_copies[hash_id(input)][split->backend_id][sched->cur_copy]; const size_t input_id = hash_id(input);
struct ggml_tensor * input_cpy = sched->tensor_copies[input_id][split->backend_id][sched->cur_copy];
// add a dependency to the input source so that it is not freed before the copy is done // add a dependency to the input source so that it is not freed before the copy is done
struct ggml_tensor * input_dep = ggml_view_tensor(sched->ctx, input); struct ggml_tensor * input_dep = ggml_view_tensor(sched->ctx, input);
input_dep->src[0] = input; input_dep->src[0] = input;
sched->node_backend_ids[graph_copy->n_nodes] = tensor_backend_id(input); sched->node_backend_ids[graph_copy->n_nodes] = sched->tensor_backend_id[input_id];
graph_copy->nodes[graph_copy->n_nodes++] = input_dep; graph_copy->nodes[graph_copy->n_nodes++] = input_dep;
// add a dependency to the input copy so that it is allocated at the start of the split // add a dependency to the input copy so that it is allocated at the start of the split
@ -1541,6 +1564,7 @@ static void ggml_backend_sched_split_graph(ggml_backend_sched_t sched, struct gg
} }
for (int j = split->i_start; j < split->i_end; j++) { for (int j = split->i_start; j < split->i_end; j++) {
assert(graph_copy->size > graph_copy->n_nodes);
sched->node_backend_ids[graph_copy->n_nodes] = tensor_backend_id(graph->nodes[j]); sched->node_backend_ids[graph_copy->n_nodes] = tensor_backend_id(graph->nodes[j]);
graph_copy->nodes[graph_copy->n_nodes++] = graph->nodes[j]; graph_copy->nodes[graph_copy->n_nodes++] = graph->nodes[j];
} }
@ -1625,13 +1649,12 @@ static enum ggml_status ggml_backend_sched_compute_splits(ggml_backend_sched_t s
} }
ggml_backend_tensor_copy(input, input_cpy); ggml_backend_tensor_copy(input, input_cpy);
} else { } else {
// wait for the split backend to finish using the input before overwriting it
if (sched->events[split_backend_id][sched->cur_copy] != NULL) { if (sched->events[split_backend_id][sched->cur_copy] != NULL) {
ggml_backend_event_wait(split_backend, sched->events[split_backend_id][sched->cur_copy]); ggml_backend_event_wait(split_backend, sched->events[split_backend_id][sched->cur_copy]);
} else { } else {
ggml_backend_synchronize(split_backend); ggml_backend_synchronize(split_backend);
ggml_backend_synchronize(input_backend);
} }
ggml_backend_tensor_copy_async(input_backend, split_backend, input, input_cpy); ggml_backend_tensor_copy_async(input_backend, split_backend, input, input_cpy);
} }
} }
@ -1701,17 +1724,21 @@ ggml_backend_sched_t ggml_backend_sched_new(
struct ggml_backend_sched * sched = calloc(sizeof(struct ggml_backend_sched), 1); struct ggml_backend_sched * sched = calloc(sizeof(struct ggml_backend_sched), 1);
// initialize hash table // initialize hash table
sched->hash_set = ggml_hash_set_new(graph_size + GGML_SCHED_MAX_SPLITS*GGML_SCHED_MAX_SPLIT_INPUTS); sched->hash_set = ggml_hash_set_new(graph_size);
sched->tensor_backend_id = calloc(sizeof(sched->tensor_backend_id[0]), sched->hash_set.size); sched->tensor_backend_id = calloc(sizeof(sched->tensor_backend_id[0]), sched->hash_set.size);
sched->tensor_copies = calloc(sizeof(sched->tensor_copies[0]), sched->hash_set.size); sched->tensor_copies = calloc(sizeof(sched->tensor_copies[0]), sched->hash_set.size);
sched->node_backend_ids = calloc(sizeof(sched->node_backend_ids[0]), graph_size);
sched->leaf_backend_ids = calloc(sizeof(sched->leaf_backend_ids[0]), graph_size); const size_t nodes_size = graph_size + GGML_SCHED_MAX_SPLITS*GGML_SCHED_MAX_SPLIT_INPUTS*2;
sched->node_backend_ids = calloc(sizeof(sched->node_backend_ids[0]), nodes_size);
sched->leaf_backend_ids = calloc(sizeof(sched->leaf_backend_ids[0]), nodes_size);
sched->n_backends = n_backends; sched->n_backends = n_backends;
sched->n_copies = parallel ? GGML_SCHED_MAX_COPIES : 1; sched->n_copies = parallel ? GGML_SCHED_MAX_COPIES : 1;
GGML_ASSERT(sched->n_copies <= GGML_SCHED_MAX_COPIES); const int initial_splits_capacity = 16;
sched->splits = calloc(sizeof(sched->splits[0]), initial_splits_capacity);
sched->splits_capacity = initial_splits_capacity;
for (int b = 0; b < n_backends; b++) { for (int b = 0; b < n_backends; b++) {
sched->backends[b] = backends[b]; sched->backends[b] = backends[b];
@ -1742,6 +1769,7 @@ void ggml_backend_sched_free(ggml_backend_sched_t sched) {
} }
ggml_gallocr_free(sched->galloc); ggml_gallocr_free(sched->galloc);
ggml_free(sched->ctx); ggml_free(sched->ctx);
free(sched->splits);
free(sched->hash_set.keys); free(sched->hash_set.keys);
free(sched->tensor_backend_id); free(sched->tensor_backend_id);
free(sched->tensor_copies); free(sched->tensor_copies);
@ -1762,6 +1790,8 @@ void ggml_backend_sched_reset(ggml_backend_sched_t sched) {
} }
bool ggml_backend_sched_reserve(ggml_backend_sched_t sched, struct ggml_cgraph * measure_graph) { bool ggml_backend_sched_reserve(ggml_backend_sched_t sched, struct ggml_cgraph * measure_graph) {
GGML_ASSERT((int)sched->hash_set.size >= measure_graph->n_nodes);
ggml_backend_sched_split_graph(sched, measure_graph); ggml_backend_sched_split_graph(sched, measure_graph);
// TODO: extract this to a separate function // TODO: extract this to a separate function
@ -1776,7 +1806,7 @@ bool ggml_backend_sched_reserve(ggml_backend_sched_t sched, struct ggml_cgraph *
} }
bool ggml_backend_sched_alloc_graph(ggml_backend_sched_t sched, struct ggml_cgraph * graph) { bool ggml_backend_sched_alloc_graph(ggml_backend_sched_t sched, struct ggml_cgraph * graph) {
GGML_ASSERT((int)sched->hash_set.size >= graph->n_nodes + GGML_SCHED_MAX_SPLITS*GGML_SCHED_MAX_SPLIT_INPUTS); GGML_ASSERT((int)sched->hash_set.size >= graph->n_nodes);
ggml_backend_sched_split_graph(sched, graph); ggml_backend_sched_split_graph(sched, graph);

8
ggml-backend.h
View File

@ -70,11 +70,11 @@ extern "C" {
GGML_API ggml_backend_graph_plan_t ggml_backend_graph_plan_create(ggml_backend_t backend, struct ggml_cgraph * cgraph); GGML_API ggml_backend_graph_plan_t ggml_backend_graph_plan_create(ggml_backend_t backend, struct ggml_cgraph * cgraph);
GGML_API void ggml_backend_graph_plan_free (ggml_backend_t backend, ggml_backend_graph_plan_t plan); GGML_API void ggml_backend_graph_plan_free (ggml_backend_t backend, ggml_backend_graph_plan_t plan);
GGML_API enum ggml_status ggml_backend_graph_plan_compute(ggml_backend_t backend, ggml_backend_graph_plan_t plan); GGML_API enum ggml_status ggml_backend_graph_plan_compute (ggml_backend_t backend, ggml_backend_graph_plan_t plan);
GGML_API enum ggml_status ggml_backend_graph_compute (ggml_backend_t backend, struct ggml_cgraph * cgraph); GGML_API enum ggml_status ggml_backend_graph_compute (ggml_backend_t backend, struct ggml_cgraph * cgraph);
GGML_API enum ggml_status ggml_backend_graph_compute_async(ggml_backend_t backend, struct ggml_cgraph * cgraph);
GGML_API bool ggml_backend_graph_compute_async(ggml_backend_t backend, struct ggml_cgraph * cgraph);
GGML_API bool ggml_backend_supports_op(ggml_backend_t backend, const struct ggml_tensor * op); GGML_API bool ggml_backend_supports_op(ggml_backend_t backend, const struct ggml_tensor * op);
GGML_API bool ggml_backend_offload_op(ggml_backend_t backend, const struct ggml_tensor * op);
// tensor copy between different backends // tensor copy between different backends
GGML_API void ggml_backend_tensor_copy(struct ggml_tensor * src, struct ggml_tensor * dst); GGML_API void ggml_backend_tensor_copy(struct ggml_tensor * src, struct ggml_tensor * dst);

297
ggml-cuda.cu
View File

@ -82,6 +82,10 @@
#define cudaGetDeviceProperties hipGetDeviceProperties #define cudaGetDeviceProperties hipGetDeviceProperties
#define cudaGetErrorString hipGetErrorString #define cudaGetErrorString hipGetErrorString
#define cudaGetLastError hipGetLastError #define cudaGetLastError hipGetLastError
#define cudaHostRegister hipHostRegister
#define cudaHostRegisterPortable hipHostRegisterPortable
#define cudaHostRegisterReadOnly hipHostRegisterReadOnly
#define cudaHostUnregister hipHostUnregister
#define cudaLaunchHostFunc hipLaunchHostFunc #define cudaLaunchHostFunc hipLaunchHostFunc
#ifdef GGML_HIP_UMA #ifdef GGML_HIP_UMA
#define cudaMalloc hipMallocManaged #define cudaMalloc hipMallocManaged
@ -7787,11 +7791,7 @@ struct cuda_pool_alloc {
static bool g_cublas_loaded = false; static bool g_cublas_loaded = false;
GGML_CALL bool ggml_cublas_loaded(void) { static void ggml_init_cublas() {
return g_cublas_loaded;
}
GGML_CALL void ggml_init_cublas() {
static bool initialized = false; static bool initialized = false;
if (!initialized) { if (!initialized) {
@ -7880,7 +7880,7 @@ GGML_CALL void ggml_init_cublas() {
} }
} }
GGML_CALL void * ggml_cuda_host_malloc(size_t size) { static void * ggml_cuda_host_malloc(size_t size) {
if (getenv("GGML_CUDA_NO_PINNED") != nullptr) { if (getenv("GGML_CUDA_NO_PINNED") != nullptr) {
return nullptr; return nullptr;
} }
@ -7890,7 +7890,7 @@ GGML_CALL void * ggml_cuda_host_malloc(size_t size) {
if (err != cudaSuccess) { if (err != cudaSuccess) {
// clear the error // clear the error
cudaGetLastError(); cudaGetLastError();
fprintf(stderr, "WARNING: failed to allocate %.2f MB of pinned memory: %s\n", fprintf(stderr, "%s: warning: failed to allocate %.2f MiB of pinned memory: %s\n", __func__,
size/1024.0/1024.0, cudaGetErrorString(err)); size/1024.0/1024.0, cudaGetErrorString(err));
return nullptr; return nullptr;
} }
@ -7898,7 +7898,7 @@ GGML_CALL void * ggml_cuda_host_malloc(size_t size) {
return ptr; return ptr;
} }
GGML_CALL void ggml_cuda_host_free(void * ptr) { static void ggml_cuda_host_free(void * ptr) {
CUDA_CHECK(cudaFreeHost(ptr)); CUDA_CHECK(cudaFreeHost(ptr));
} }
@ -9036,21 +9036,13 @@ static void ggml_cuda_op_soft_max(
// positions tensor // positions tensor
float * src2_dd = nullptr; float * src2_dd = nullptr;
cuda_pool_alloc<float> src2_f;
ggml_tensor * src2 = dst->src[2]; ggml_tensor * src2 = dst->src[2];
const bool use_src2 = src2 != nullptr; const bool use_src2 = src2 != nullptr;
if (use_src2) { if (use_src2) {
const bool src2_on_device = src2->backend == GGML_BACKEND_TYPE_GPU; ggml_tensor_extra_gpu * src2_extra = (ggml_tensor_extra_gpu *) src2->extra;
src2_dd = (float *) src2_extra->data_device[g_main_device];
if (src2_on_device) {
ggml_tensor_extra_gpu * src2_extra = (ggml_tensor_extra_gpu *) src2->extra;
src2_dd = (float *) src2_extra->data_device[g_main_device];
} else {
src2_dd = src2_f.alloc(ggml_nelements(src2));
CUDA_CHECK(ggml_cuda_cpy_tensor_2d(src2_dd, src2, 0, 0, 0, 1, main_stream));
}
} }
soft_max_f32_cuda(src0_dd, src1 ? src1_dd : nullptr, src2_dd, dst_dd, ne00, nrows_x, nrows_y, scale, max_bias, main_stream); soft_max_f32_cuda(src0_dd, src1 ? src1_dd : nullptr, src2_dd, dst_dd, ne00, nrows_x, nrows_y, scale, max_bias, main_stream);
@ -9107,55 +9099,24 @@ static void ggml_cuda_op_flatten(const ggml_tensor * src0, const ggml_tensor * s
ggml_tensor_extra_gpu * src1_extra = use_src1 ? (ggml_tensor_extra_gpu *) src1->extra : nullptr; ggml_tensor_extra_gpu * src1_extra = use_src1 ? (ggml_tensor_extra_gpu *) src1->extra : nullptr;
ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra; ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra;
const bool src0_on_device = src0->backend == GGML_BACKEND_TYPE_GPU || src0->backend == GGML_BACKEND_TYPE_GPU_SPLIT;
const bool src1_on_device = use_src1 && src1->backend == GGML_BACKEND_TYPE_GPU;
const bool dst_on_device = dst->backend == GGML_BACKEND_TYPE_GPU;
// dd = data device // dd = data device
float * src0_ddf = nullptr; float * src0_ddf = nullptr;
float * src1_ddf = nullptr; float * src1_ddf = nullptr;
float * dst_ddf = nullptr; float * dst_ddf = nullptr;
cuda_pool_alloc<float> src0_f;
cuda_pool_alloc<float> src1_f;
cuda_pool_alloc<float> dst_f;
ggml_cuda_set_device(g_main_device); ggml_cuda_set_device(g_main_device);
cudaStream_t main_stream = g_cudaStreams[g_main_device][0]; cudaStream_t main_stream = g_cudaStreams[g_main_device][0];
if (src0_on_device) { src0_ddf = (float *) src0_extra->data_device[g_main_device];
src0_ddf = (float *) src0_extra->data_device[g_main_device];
} else {
src0_ddf = src0_f.alloc(ggml_nelements(src0));
CUDA_CHECK(ggml_cuda_cpy_tensor_2d(src0_ddf, src0, 0, 0, 0, nrows0, main_stream));
}
if (use_src1) { if (use_src1) {
if (src1_on_device) { src1_ddf = (float *) src1_extra->data_device[g_main_device];
src1_ddf = (float *) src1_extra->data_device[g_main_device];
} else {
src1_ddf = src1_f.alloc(ggml_nelements(src1));
CUDA_CHECK(ggml_cuda_cpy_tensor_2d(src1_ddf, src1, 0, 0, 0, nrows1, main_stream));
}
}
if (dst_on_device) {
dst_ddf = (float *) dst_extra->data_device[g_main_device];
} else {
dst_ddf = dst_f.alloc(ggml_nelements(dst));
} }
dst_ddf = (float *) dst_extra->data_device[g_main_device];
// do the computation // do the computation
op(src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream); op(src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream);
CUDA_CHECK(cudaGetLastError()); CUDA_CHECK(cudaGetLastError());
// copy dst to host if necessary
if (!dst_on_device) {
CUDA_CHECK(cudaMemcpyAsync(dst->data, dst_ddf, ggml_nbytes(dst), cudaMemcpyDeviceToHost, main_stream));
}
if (dst->backend == GGML_BACKEND_TYPE_CPU) {
CUDA_CHECK(cudaDeviceSynchronize());
}
} }
static void ggml_cuda_set_peer_access(const int n_tokens) { static void ggml_cuda_set_peer_access(const int n_tokens) {
@ -9251,7 +9212,6 @@ static void ggml_cuda_op_mul_mat(
ggml_tensor_extra_gpu * src1_extra = (ggml_tensor_extra_gpu *) src1->extra; ggml_tensor_extra_gpu * src1_extra = (ggml_tensor_extra_gpu *) src1->extra;
ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra; ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra;
const bool src0_on_device = src0->backend == GGML_BACKEND_TYPE_GPU || src0->backend == GGML_BACKEND_TYPE_GPU_SPLIT;
const bool src0_is_contiguous = ggml_is_contiguous(src0); const bool src0_is_contiguous = ggml_is_contiguous(src0);
const bool src1_is_contiguous = ggml_is_contiguous(src1); const bool src1_is_contiguous = ggml_is_contiguous(src1);
@ -9322,13 +9282,13 @@ static void ggml_cuda_op_mul_mat(
used_devices++; used_devices++;
const bool src1_on_device = src1->backend == GGML_BACKEND_TYPE_GPU && id == g_main_device; const bool src1_on_device = id == g_main_device; // TODO: check from buffer
const bool dst_on_device = dst->backend == GGML_BACKEND_TYPE_GPU && id == g_main_device; const bool dst_on_device = id == g_main_device;
ggml_cuda_set_device(id); ggml_cuda_set_device(id);
cudaStream_t stream = g_cudaStreams[id][0]; cudaStream_t stream = g_cudaStreams[id][0];
if (src0_on_device && src0_is_contiguous) { if (src0_is_contiguous) {
dev[id].src0_dd = (char *) src0_extra->data_device[id]; dev[id].src0_dd = (char *) src0_extra->data_device[id];
} else { } else {
dev[id].src0_dd = dev[id].src0_dd_alloc.alloc(ggml_nbytes(src0)); dev[id].src0_dd = dev[id].src0_dd_alloc.alloc(ggml_nbytes(src0));
@ -9374,8 +9334,8 @@ static void ggml_cuda_op_mul_mat(
continue; continue;
} }
const bool src1_on_device = src1->backend == GGML_BACKEND_TYPE_GPU && id == g_main_device; const bool src1_on_device = id == g_main_device; // TODO: check from buffer
const bool dst_on_device = dst->backend == GGML_BACKEND_TYPE_GPU && id == g_main_device; const bool dst_on_device = id == g_main_device;
const int64_t row_diff = dev[id].row_high - dev[id].row_low; const int64_t row_diff = dev[id].row_high - dev[id].row_low;
ggml_cuda_set_device(id); ggml_cuda_set_device(id);
@ -9400,12 +9360,12 @@ static void ggml_cuda_op_mul_mat(
// the main device memory buffer can be on VRAM scratch, with space for all partial results // the main device memory buffer can be on VRAM scratch, with space for all partial results
// in that case an offset on dst_ddf_i is needed // in that case an offset on dst_ddf_i is needed
if (dst->backend == GGML_BACKEND_TYPE_GPU && id == g_main_device) { if (id == g_main_device) {
dst_dd_i += dev[id].row_low; // offset is 0 if no tensor split dst_dd_i += dev[id].row_low; // offset is 0 if no tensor split
} }
// copy src0, src1 to device if necessary // copy src0, src1 to device if necessary
if (src1->backend == GGML_BACKEND_TYPE_GPU && src1_is_contiguous) { if (src1_is_contiguous) {
if (id != g_main_device) { if (id != g_main_device) {
if (convert_src1_to_q8_1) { if (convert_src1_to_q8_1) {
char * src1_ddq_i_source = dev[g_main_device].src1_ddq + src1_ddq_i_offset; char * src1_ddq_i_source = dev[g_main_device].src1_ddq + src1_ddq_i_offset;
@ -9418,19 +9378,19 @@ static void ggml_cuda_op_mul_mat(
src1_ncols*ne10*sizeof(float), stream)); src1_ncols*ne10*sizeof(float), stream));
} }
} }
} else if (src1->backend == GGML_BACKEND_TYPE_CPU || (src1_on_device && !src1_is_contiguous)) { } else if (src1_on_device && !src1_is_contiguous) {
CUDA_CHECK(ggml_cuda_cpy_tensor_2d( CUDA_CHECK(ggml_cuda_cpy_tensor_2d(
src1_ddf_i, src1, i03, i02, src1_col_0, src1_col_0+src1_ncols, stream)); src1_ddf_i, src1, i03, i02, src1_col_0, src1_col_0+src1_ncols, stream));
} else { } else {
GGML_ASSERT(false); GGML_ASSERT(false);
} }
if (convert_src1_to_q8_1 && (src1->backend == GGML_BACKEND_TYPE_CPU || !src1_is_contiguous)) { if (convert_src1_to_q8_1 && !src1_is_contiguous) {
quantize_row_q8_1_cuda(src1_ddf_i, src1_ddq_i, ne10, src1_ncols, src1_padded_col_size, stream); quantize_row_q8_1_cuda(src1_ddf_i, src1_ddq_i, ne10, src1_ncols, src1_padded_col_size, stream);
CUDA_CHECK(cudaGetLastError()); CUDA_CHECK(cudaGetLastError());
} }
if (src1_col_0 == 0 && (!src0_on_device || !src0_is_contiguous) && i02 % i02_divisor == 0) { if (src1_col_0 == 0 && !src0_is_contiguous && i02 % i02_divisor == 0) {
CUDA_CHECK(ggml_cuda_cpy_tensor_2d(src0_dd_i, src0, i03, i02/i02_divisor, dev[id].row_low, dev[id].row_high, stream)); CUDA_CHECK(ggml_cuda_cpy_tensor_2d(src0_dd_i, src0, i03, i02/i02_divisor, dev[id].row_low, dev[id].row_high, stream));
} }
@ -9441,17 +9401,7 @@ static void ggml_cuda_op_mul_mat(
// copy dst to host or other device if necessary // copy dst to host or other device if necessary
if (!dst_on_device) { if (!dst_on_device) {
void * dst_off_device; void * dst_off_device = dst_extra->data_device[g_main_device];
cudaMemcpyKind kind;
if (dst->backend == GGML_BACKEND_TYPE_CPU) {
dst_off_device = dst->data;
kind = cudaMemcpyDeviceToHost;
} else if (dst->backend == GGML_BACKEND_TYPE_GPU) {
dst_off_device = dst_extra->data_device[g_main_device];
kind = cudaMemcpyDeviceToDevice;
} else {
GGML_ASSERT(false);
}
if (split) { if (split) {
// src0 = weight matrix is saved as a transposed matrix for better memory layout. // src0 = weight matrix is saved as a transposed matrix for better memory layout.
// dst is NOT transposed. // dst is NOT transposed.
@ -9462,28 +9412,26 @@ static void ggml_cuda_op_mul_mat(
GGML_ASSERT(dst->nb[1] == ne0*sizeof(float)); GGML_ASSERT(dst->nb[1] == ne0*sizeof(float));
dhf_dst_i += src1_col_0*ne0 + dev[id].row_low; dhf_dst_i += src1_col_0*ne0 + dev[id].row_low;
#if !defined(GGML_USE_HIPBLAS) #if !defined(GGML_USE_HIPBLAS)
if (kind == cudaMemcpyDeviceToDevice) { // cudaMemcpy2DAsync may fail with copies between vmm pools of different devices
// cudaMemcpy2DAsync may fail with copies between vmm pools of different devices cudaMemcpy3DPeerParms p = {};
cudaMemcpy3DPeerParms p = {}; p.dstDevice = g_main_device;
p.dstDevice = g_main_device; p.dstPtr = make_cudaPitchedPtr(dhf_dst_i, ne0*sizeof(float), row_diff, src1_ncols);
p.dstPtr = make_cudaPitchedPtr(dhf_dst_i, ne0*sizeof(float), row_diff, src1_ncols); p.srcDevice = id;
p.srcDevice = id; p.srcPtr = make_cudaPitchedPtr(dst_dd_i, row_diff*sizeof(float), row_diff, src1_ncols);
p.srcPtr = make_cudaPitchedPtr(dst_dd_i, row_diff*sizeof(float), row_diff, src1_ncols); p.extent = make_cudaExtent(row_diff*sizeof(float), src1_ncols, 1);
p.extent = make_cudaExtent(row_diff*sizeof(float), src1_ncols, 1); CUDA_CHECK(cudaMemcpy3DPeerAsync(&p, stream));
CUDA_CHECK(cudaMemcpy3DPeerAsync(&p, stream)); #else
} else // HIP does not support cudaMemcpy3DPeerAsync or vmm pools
CUDA_CHECK(cudaMemcpy2DAsync(dhf_dst_i, ne0*sizeof(float),
dst_dd_i, row_diff*sizeof(float),
row_diff*sizeof(float), src1_ncols,
cudaMemcpyDeviceToDevice, stream));
#endif #endif
{
CUDA_CHECK(cudaMemcpy2DAsync(dhf_dst_i, ne0*sizeof(float),
dst_dd_i, row_diff*sizeof(float),
row_diff*sizeof(float), src1_ncols,
kind, stream));
}
} else { } else {
float * dhf_dst_i = (float *) ((char *) dst_off_device + i02*nb2 + i03*nb3); float * dhf_dst_i = (float *) ((char *) dst_off_device + i02*nb2 + i03*nb3);
GGML_ASSERT(dst->nb[1] == ne0*sizeof(float)); GGML_ASSERT(dst->nb[1] == ne0*sizeof(float));
dhf_dst_i += src1_col_0*ne0; dhf_dst_i += src1_col_0*ne0;
CUDA_CHECK(cudaMemcpyAsync(dhf_dst_i, dst_dd_i, src1_ncols*ne0*sizeof(float), kind, stream)); CUDA_CHECK(cudaMemcpyAsync(dhf_dst_i, dst_dd_i, src1_ncols*ne0*sizeof(float), cudaMemcpyDeviceToDevice, stream));
} }
} }
@ -9510,11 +9458,6 @@ static void ggml_cuda_op_mul_mat(
} }
} }
} }
if (dst->backend == GGML_BACKEND_TYPE_CPU) {
ggml_cuda_set_device(g_main_device);
CUDA_CHECK(cudaDeviceSynchronize());
}
} }
static void ggml_cuda_repeat(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) { static void ggml_cuda_repeat(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
@ -9599,36 +9542,19 @@ static void ggml_cuda_pad(const ggml_tensor * src0, const ggml_tensor * src1, gg
static void ggml_cuda_arange(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) { static void ggml_cuda_arange(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra; ggml_tensor_extra_gpu * dst_extra = (ggml_tensor_extra_gpu *) dst->extra;
const bool dst_on_device = dst->backend == GGML_BACKEND_TYPE_GPU;
// dd = data device // dd = data device
float * src0_ddf = nullptr; float * src0_ddf = nullptr;
float * src1_ddf = nullptr; float * src1_ddf = nullptr;
float * dst_ddf = nullptr; float * dst_ddf = nullptr;
cuda_pool_alloc<float> dst_f;
ggml_cuda_set_device(g_main_device); ggml_cuda_set_device(g_main_device);
cudaStream_t main_stream = g_cudaStreams[g_main_device][0]; cudaStream_t main_stream = g_cudaStreams[g_main_device][0];
if (dst_on_device) { dst_ddf = (float *) dst_extra->data_device[g_main_device];
dst_ddf = (float *) dst_extra->data_device[g_main_device];
} else {
dst_ddf = dst_f.alloc(ggml_nelements(dst));
}
// do the computation // do the computation
ggml_cuda_op_arange(src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream); ggml_cuda_op_arange(src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream);
CUDA_CHECK(cudaGetLastError()); CUDA_CHECK(cudaGetLastError());
// copy dst to host if necessary
if (!dst_on_device) {
CUDA_CHECK(cudaMemcpyAsync(dst->data, dst_ddf, ggml_nbytes(dst), cudaMemcpyDeviceToHost, main_stream));
}
if (dst->backend == GGML_BACKEND_TYPE_CPU) {
CUDA_CHECK(cudaDeviceSynchronize());
}
} }
static void ggml_cuda_timestep_embedding(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) { static void ggml_cuda_timestep_embedding(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
@ -9639,21 +9565,6 @@ static void ggml_cuda_rms_norm(const ggml_tensor * src0, const ggml_tensor * src
ggml_cuda_op_flatten(src0, src1, dst, ggml_cuda_op_rms_norm); ggml_cuda_op_flatten(src0, src1, dst, ggml_cuda_op_rms_norm);
} }
GGML_CALL bool ggml_cuda_can_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst) {
if (!g_cublas_loaded) return false;
const int64_t ne10 = src1->ne[0];
const int64_t ne0 = dst->ne[0];
const int64_t ne1 = dst->ne[1];
// TODO: find the optimal values for these
return (src0->type == GGML_TYPE_F32 || src0->type == GGML_TYPE_F16 || ggml_is_quantized(src0->type)) &&
src1->type == GGML_TYPE_F32 &&
dst->type == GGML_TYPE_F32 &&
(ne0 >= 32 && ne1 >= 32 && ne10 >= 32);
}
static void ggml_cuda_mul_mat_vec_p021(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst){ static void ggml_cuda_mul_mat_vec_p021(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst){
GGML_ASSERT(ggml_is_permuted(src0) && ggml_is_permuted(src1)); GGML_ASSERT(ggml_is_permuted(src0) && ggml_is_permuted(src1));
GGML_ASSERT(src0->backend != GGML_BACKEND_TYPE_GPU_SPLIT); GGML_ASSERT(src0->backend != GGML_BACKEND_TYPE_GPU_SPLIT);
@ -9891,11 +9802,6 @@ static void ggml_cuda_mul_mat_batched_cublas(const ggml_tensor * src0, const ggm
} }
static void ggml_cuda_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) { static void ggml_cuda_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
const bool all_on_device =
(src0->backend == GGML_BACKEND_TYPE_GPU || src0->backend == GGML_BACKEND_TYPE_GPU_SPLIT) &&
(src1->backend == GGML_BACKEND_TYPE_GPU) &&
( dst->backend == GGML_BACKEND_TYPE_GPU);
const bool split = src0->backend == GGML_BACKEND_TYPE_GPU_SPLIT; const bool split = src0->backend == GGML_BACKEND_TYPE_GPU_SPLIT;
int64_t min_compute_capability = INT_MAX; int64_t min_compute_capability = INT_MAX;
@ -9972,13 +9878,13 @@ static void ggml_cuda_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1
//printf("src0 is contiguous %d, transposed %d, type = %s, name = %s\n", ggml_is_contiguous(src0), ggml_is_transposed(src0), ggml_type_name(src0->type), src0->name); //printf("src0 is contiguous %d, transposed %d, type = %s, name = %s\n", ggml_is_contiguous(src0), ggml_is_transposed(src0), ggml_type_name(src0->type), src0->name);
//printf("src1 is contiguous %d, transposed %d, type = %s, name = %s\n", ggml_is_contiguous(src1), ggml_is_transposed(src1), ggml_type_name(src1->type), src1->name); //printf("src1 is contiguous %d, transposed %d, type = %s, name = %s\n", ggml_is_contiguous(src1), ggml_is_transposed(src1), ggml_type_name(src1->type), src1->name);
if (!split && all_on_device && !fp16_performance_good && src0->type == GGML_TYPE_F16 && ggml_is_permuted(src0) && ggml_is_permuted(src1) && src1->ne[1] == 1) { if (!split && !fp16_performance_good && src0->type == GGML_TYPE_F16 && ggml_is_permuted(src0) && ggml_is_permuted(src1) && src1->ne[1] == 1) {
// KQ single-batch // KQ single-batch
ggml_cuda_mul_mat_vec_p021(src0, src1, dst); ggml_cuda_mul_mat_vec_p021(src0, src1, dst);
} else if (!split && all_on_device && !fp16_performance_good && src0->type == GGML_TYPE_F16 && !ggml_is_contiguous(src0) && !ggml_is_transposed(src1) && src1->ne[1] == 1) { } else if (!split && !fp16_performance_good && src0->type == GGML_TYPE_F16 && !ggml_is_contiguous(src0) && !ggml_is_transposed(src1) && src1->ne[1] == 1) {
// KQV single-batch // KQV single-batch
ggml_cuda_mul_mat_vec_nc(src0, src1, dst); ggml_cuda_mul_mat_vec_nc(src0, src1, dst);
} else if (!split && all_on_device && fp16_performance_good && src0->type == GGML_TYPE_F16 && !ggml_is_transposed(src0) && !ggml_is_transposed(src1) && src1->ne[2]*src1->ne[3] > 1) { } else if (!split && fp16_performance_good && src0->type == GGML_TYPE_F16 && !ggml_is_transposed(src0) && !ggml_is_transposed(src1) && src1->ne[2]*src1->ne[3] > 1) {
// KQ + KQV multi-batch // KQ + KQV multi-batch
ggml_cuda_mul_mat_batched_cublas(src0, src1, dst); ggml_cuda_mul_mat_batched_cublas(src0, src1, dst);
} else if (use_dequantize_mul_mat_vec) { } else if (use_dequantize_mul_mat_vec) {
@ -10178,6 +10084,7 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
ggml_cuda_mul_mat_id_cublas(dst); ggml_cuda_mul_mat_id_cublas(dst);
// TODO: mmq/mmv support // TODO: mmq/mmv support
#endif #endif
cudaStream_t stream = g_cudaStreams[g_main_device][0];
const size_t nb11 = src1->nb[1]; const size_t nb11 = src1->nb[1];
const size_t nb1 = dst->nb[1]; const size_t nb1 = dst->nb[1];
@ -10187,16 +10094,9 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
const int32_t n_as = ((int32_t *) dst->op_params)[1]; const int32_t n_as = ((int32_t *) dst->op_params)[1];
std::vector<char> ids_host(ggml_nbytes(ids)); std::vector<char> ids_host(ggml_nbytes(ids));
const char * ids_dev = (const char *)((const ggml_tensor_extra_gpu *)ids->extra)->data_device[g_main_device];
cudaStream_t stream = g_cudaStreams[g_main_device][0]; CUDA_CHECK(cudaMemcpyAsync(ids_host.data(), ids_dev, ggml_nbytes(ids), cudaMemcpyDeviceToHost, stream));
CUDA_CHECK(cudaStreamSynchronize(stream));
if (ids->backend == GGML_BACKEND_TYPE_GPU) {
const char * ids_dev = (const char *)((const ggml_tensor_extra_gpu *)ids->extra)->data_device[g_main_device];
CUDA_CHECK(cudaMemcpyAsync(ids_host.data(), ids_dev, ggml_nbytes(ids), cudaMemcpyDeviceToHost, stream));
CUDA_CHECK(cudaStreamSynchronize(stream));
} else {
memcpy(ids_host.data(), ids->data, ggml_nbytes(ids));
}
const ggml_tensor_extra_gpu * src1_extra = (const ggml_tensor_extra_gpu *) src1->extra; const ggml_tensor_extra_gpu * src1_extra = (const ggml_tensor_extra_gpu *) src1->extra;
const ggml_tensor_extra_gpu * dst_extra = (const ggml_tensor_extra_gpu *) dst->extra; const ggml_tensor_extra_gpu * dst_extra = (const ggml_tensor_extra_gpu *) dst->extra;
@ -10213,20 +10113,11 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
src1_row.extra = &src1_row_extra; src1_row.extra = &src1_row_extra;
dst_row.extra = &dst_row_extra; dst_row.extra = &dst_row_extra;
char * src1_original = src1->backend == GGML_BACKEND_TYPE_CPU ? char * src1_original = (char *) src1_extra->data_device[g_main_device];
(char *) src1->data : (char *) src1_extra->data_device[g_main_device]; char * dst_original = (char *) dst_extra->data_device[g_main_device];
char * dst_original = dst->backend == GGML_BACKEND_TYPE_CPU ?
(char *) dst->data : (char *) dst_extra->data_device[g_main_device];
if (src1->ne[1] == 1) { if (src1->ne[1] == 1) {
GGML_ASSERT(src1->backend == GGML_BACKEND_TYPE_GPU);
GGML_ASSERT(dst->backend == GGML_BACKEND_TYPE_GPU);
for (int64_t i01 = 0; i01 < ids->ne[1]; i01++) { for (int64_t i01 = 0; i01 < ids->ne[1]; i01++) {
//int32_t row_id;
//CUDA_CHECK(cudaMemcpyAsync(&row_id, ids_dev + i01*ids->nb[1] + id*ids->nb[0], sizeof(int32_t), cudaMemcpyDeviceToHost, g_cudaStreams[g_main_device][0]));
//CUDA_CHECK(cudaStreamSynchronize(g_cudaStreams[g_main_device][0]));
const int32_t row_id = *(const int32_t *) (ids_host.data() + i01*ids->nb[1] + id*ids->nb[0]); const int32_t row_id = *(const int32_t *) (ids_host.data() + i01*ids->nb[1] + id*ids->nb[0]);
GGML_ASSERT(row_id >= 0 && row_id < n_as); GGML_ASSERT(row_id >= 0 && row_id < n_as);
@ -10248,11 +10139,6 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
src1_row_extra.data_device[g_main_device] = src1_contiguous.get(); src1_row_extra.data_device[g_main_device] = src1_contiguous.get();
dst_row_extra.data_device[g_main_device] = dst_contiguous.get(); dst_row_extra.data_device[g_main_device] = dst_contiguous.get();
const cudaMemcpyKind src1_kind = src1->backend == GGML_BACKEND_TYPE_CPU ?
cudaMemcpyHostToDevice : cudaMemcpyDeviceToDevice;
const cudaMemcpyKind dst_kind = dst->backend == GGML_BACKEND_TYPE_CPU ?
cudaMemcpyDeviceToHost : cudaMemcpyDeviceToDevice;
for (int32_t row_id = 0; row_id < n_as; ++row_id) { for (int32_t row_id = 0; row_id < n_as; ++row_id) {
const struct ggml_tensor * src0_row = dst->src[row_id + 2]; const struct ggml_tensor * src0_row = dst->src[row_id + 2];
@ -10267,7 +10153,7 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
GGML_ASSERT(row_id >= 0 && row_id < n_as); GGML_ASSERT(row_id >= 0 && row_id < n_as);
CUDA_CHECK(cudaMemcpyAsync(src1_contiguous.get() + num_src1_rows*nb11, src1_original + i01*nb11, CUDA_CHECK(cudaMemcpyAsync(src1_contiguous.get() + num_src1_rows*nb11, src1_original + i01*nb11,
nb11, src1_kind, stream)); nb11, cudaMemcpyDeviceToDevice, stream));
num_src1_rows++; num_src1_rows++;
} }
@ -10299,15 +10185,11 @@ static void ggml_cuda_mul_mat_id(const ggml_tensor * src0, const ggml_tensor * s
GGML_ASSERT(row_id >= 0 && row_id < n_as); GGML_ASSERT(row_id >= 0 && row_id < n_as);
CUDA_CHECK(cudaMemcpyAsync(dst_original + i01*nb1, dst_contiguous.get() + num_src1_rows*nb1, CUDA_CHECK(cudaMemcpyAsync(dst_original + i01*nb1, dst_contiguous.get() + num_src1_rows*nb1,
nb1, dst_kind, stream)); nb1, cudaMemcpyDeviceToDevice, stream));
num_src1_rows++; num_src1_rows++;
} }
} }
} }
if (dst->backend == GGML_BACKEND_TYPE_CPU) {
CUDA_CHECK(cudaStreamSynchronize(stream));
}
} }
static void ggml_cuda_scale(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) { static void ggml_cuda_scale(const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst) {
@ -10435,7 +10317,7 @@ static size_t ggml_nbytes_split(const struct ggml_tensor * tensor, int nrows_spl
return nrows_split*ggml_row_size(tensor->type, tensor->ne[0]); return nrows_split*ggml_row_size(tensor->type, tensor->ne[0]);
} }
GGML_CALL static void ggml_cuda_set_main_device(const int main_device) { static void ggml_cuda_set_main_device(const int main_device) {
if (main_device >= g_device_count) { if (main_device >= g_device_count) {
fprintf(stderr, "warning: cannot set main_device=%d because there are only %d devices. Using device %d instead.\n", fprintf(stderr, "warning: cannot set main_device=%d because there are only %d devices. Using device %d instead.\n",
main_device, g_device_count, g_main_device); main_device, g_device_count, g_main_device);
@ -10450,18 +10332,9 @@ GGML_CALL static void ggml_cuda_set_main_device(const int main_device) {
} }
} }
GGML_CALL bool ggml_cuda_compute_forward(struct ggml_compute_params * params, struct ggml_tensor * tensor) { static bool ggml_cuda_compute_forward(struct ggml_tensor * tensor) {
if (!g_cublas_loaded) return false; if (!g_cublas_loaded) return false;
ggml_cuda_func_t func;
const bool any_on_device = tensor->backend == GGML_BACKEND_TYPE_GPU
|| (tensor->src[0] != nullptr && (tensor->src[0]->backend == GGML_BACKEND_TYPE_GPU || tensor->src[0]->backend == GGML_BACKEND_TYPE_GPU_SPLIT))
|| (tensor->src[1] != nullptr && tensor->src[1]->backend == GGML_BACKEND_TYPE_GPU);
if (!any_on_device && tensor->op != GGML_OP_MUL_MAT && tensor->op != GGML_OP_MUL_MAT_ID) {
return false;
}
if (tensor->op == GGML_OP_MUL_MAT) { if (tensor->op == GGML_OP_MUL_MAT) {
if (tensor->src[0]->ne[3] != tensor->src[1]->ne[3]) { if (tensor->src[0]->ne[3] != tensor->src[1]->ne[3]) {
#ifndef NDEBUG #ifndef NDEBUG
@ -10471,6 +10344,8 @@ GGML_CALL bool ggml_cuda_compute_forward(struct ggml_compute_params * params, st
} }
} }
ggml_cuda_func_t func;
switch (tensor->op) { switch (tensor->op) {
case GGML_OP_REPEAT: case GGML_OP_REPEAT:
func = ggml_cuda_repeat; func = ggml_cuda_repeat;
@ -10548,15 +10423,9 @@ GGML_CALL bool ggml_cuda_compute_forward(struct ggml_compute_params * params, st
func = ggml_cuda_rms_norm; func = ggml_cuda_rms_norm;
break; break;
case GGML_OP_MUL_MAT: case GGML_OP_MUL_MAT:
if (!any_on_device && !ggml_cuda_can_mul_mat(tensor->src[0], tensor->src[1], tensor)) {
return false;
}
func = ggml_cuda_mul_mat; func = ggml_cuda_mul_mat;
break; break;
case GGML_OP_MUL_MAT_ID: case GGML_OP_MUL_MAT_ID:
if (!any_on_device && !ggml_cuda_can_mul_mat(tensor->src[2], tensor->src[1], tensor)) {
return false;
}
func = ggml_cuda_mul_mat_id; func = ggml_cuda_mul_mat_id;
break; break;
case GGML_OP_SCALE: case GGML_OP_SCALE:
@ -10613,17 +10482,11 @@ GGML_CALL bool ggml_cuda_compute_forward(struct ggml_compute_params * params, st
ggml_cuda_set_peer_access(tensor->src[1]->ne[1]); ggml_cuda_set_peer_access(tensor->src[1]->ne[1]);
} }
if (params->ith != 0) {
return true;
}
if (params->type == GGML_TASK_TYPE_INIT || params->type == GGML_TASK_TYPE_FINALIZE) {
return true;
}
func(tensor->src[0], tensor->src[1], tensor); func(tensor->src[0], tensor->src[1], tensor);
return true; return true;
} }
GGML_CALL int ggml_cuda_get_device_count() { static int ggml_cuda_get_device_count() {
int device_count; int device_count;
if (cudaGetDeviceCount(&device_count) != cudaSuccess) { if (cudaGetDeviceCount(&device_count) != cudaSuccess) {
return 0; return 0;
@ -10631,7 +10494,7 @@ GGML_CALL int ggml_cuda_get_device_count() {
return device_count; return device_count;
} }
GGML_CALL void ggml_cuda_get_device_description(int device, char * description, size_t description_size) { static void ggml_cuda_get_device_description(int device, char * description, size_t description_size) {
cudaDeviceProp prop; cudaDeviceProp prop;
CUDA_CHECK(cudaGetDeviceProperties(&prop, device)); CUDA_CHECK(cudaGetDeviceProperties(&prop, device));
snprintf(description, description_size, "%s", prop.name); snprintf(description, description_size, "%s", prop.name);
@ -10736,6 +10599,7 @@ GGML_CALL static void ggml_backend_cuda_buffer_init_tensor(ggml_backend_buffer_t
size_t padded_size = ggml_backend_buft_get_alloc_size(buffer->buft, tensor); size_t padded_size = ggml_backend_buft_get_alloc_size(buffer->buft, tensor);
if (padded_size > original_size && tensor->view_src == nullptr) { if (padded_size > original_size && tensor->view_src == nullptr) {
ggml_cuda_set_device(ctx->device);
CUDA_CHECK(cudaMemset((char *)tensor->data + original_size, 0, padded_size - original_size)); CUDA_CHECK(cudaMemset((char *)tensor->data + original_size, 0, padded_size - original_size));
} }
} }
@ -10873,6 +10737,8 @@ static ggml_backend_buffer_type_i ggml_backend_cuda_buffer_type_interface = {
}; };
GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_buffer_type(int device) { GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_buffer_type(int device) {
ggml_init_cublas();
// FIXME: this is not thread safe // FIXME: this is not thread safe
if (device >= ggml_backend_cuda_get_device_count()) { if (device >= ggml_backend_cuda_get_device_count()) {
return nullptr; return nullptr;
@ -11157,6 +11023,8 @@ static ggml_backend_buffer_type_i ggml_backend_cuda_split_buffer_type_interface
}; };
GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_split_buffer_type(const float * tensor_split) { GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_split_buffer_type(const float * tensor_split) {
ggml_init_cublas();
// FIXME: this is not thread safe // FIXME: this is not thread safe
static std::map<std::array<float, GGML_CUDA_MAX_DEVICES>, struct ggml_backend_buffer_type> buft_map; static std::map<std::array<float, GGML_CUDA_MAX_DEVICES>, struct ggml_backend_buffer_type> buft_map;
@ -11348,9 +11216,6 @@ GGML_CALL static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t
ggml_cuda_set_main_device(cuda_ctx->device); ggml_cuda_set_main_device(cuda_ctx->device);
ggml_compute_params params = {};
params.type = GGML_TASK_TYPE_COMPUTE;
params.ith = 0;
for (int i = 0; i < cgraph->n_nodes; i++) { for (int i = 0; i < cgraph->n_nodes; i++) {
ggml_tensor * node = cgraph->nodes[i]; ggml_tensor * node = cgraph->nodes[i];
@ -11372,7 +11237,7 @@ GGML_CALL static enum ggml_status ggml_backend_cuda_graph_compute(ggml_backend_t
} }
#endif #endif
bool ok = ggml_cuda_compute_forward(&params, node); bool ok = ggml_cuda_compute_forward(node);
if (!ok) { if (!ok) {
fprintf(stderr, "%s: error: op not supported %s (%s)\n", __func__, node->name, ggml_op_name(node->op)); fprintf(stderr, "%s: error: op not supported %s (%s)\n", __func__, node->name, ggml_op_name(node->op));
} }
@ -11509,6 +11374,14 @@ GGML_CALL static bool ggml_backend_cuda_supports_op(ggml_backend_t backend, cons
UNUSED(backend); UNUSED(backend);
} }
GGML_CALL static bool ggml_backend_cuda_offload_op(ggml_backend_t backend, const ggml_tensor * op) {
const int min_batch_size = 32;
return op->ne[1] >= min_batch_size && op->op != GGML_OP_GET_ROWS;
UNUSED(backend);
}
static ggml_backend_event_t ggml_backend_cuda_event_new(ggml_backend_t backend) { static ggml_backend_event_t ggml_backend_cuda_event_new(ggml_backend_t backend) {
ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context; ggml_backend_cuda_context * cuda_ctx = (ggml_backend_cuda_context *)backend->context;
@ -11571,6 +11444,7 @@ static ggml_backend_i ggml_backend_cuda_interface = {
/* .graph_plan_compute = */ NULL, /* .graph_plan_compute = */ NULL,
/* .graph_compute = */ ggml_backend_cuda_graph_compute, /* .graph_compute = */ ggml_backend_cuda_graph_compute,
/* .supports_op = */ ggml_backend_cuda_supports_op, /* .supports_op = */ ggml_backend_cuda_supports_op,
/* .offload_op = */ ggml_backend_cuda_offload_op,
/* .event_new = */ ggml_backend_cuda_event_new, /* .event_new = */ ggml_backend_cuda_event_new,
/* .event_free = */ ggml_backend_cuda_event_free, /* .event_free = */ ggml_backend_cuda_event_free,
/* .event_record = */ ggml_backend_cuda_event_record, /* .event_record = */ ggml_backend_cuda_event_record,
@ -11584,7 +11458,7 @@ static ggml_guid_t ggml_backend_cuda_guid() {
} }
GGML_CALL ggml_backend_t ggml_backend_cuda_init(int device) { GGML_CALL ggml_backend_t ggml_backend_cuda_init(int device) {
ggml_init_cublas(); // TODO: remove from ggml.c ggml_init_cublas();
if (device < 0 || device >= ggml_cuda_get_device_count()) { if (device < 0 || device >= ggml_cuda_get_device_count()) {
fprintf(stderr, "%s: error: invalid device %d\n", __func__, device); fprintf(stderr, "%s: error: invalid device %d\n", __func__, device);
@ -11627,6 +11501,31 @@ GGML_CALL void ggml_backend_cuda_get_device_memory(int device, size_t * free, si
CUDA_CHECK(cudaMemGetInfo(free, total)); CUDA_CHECK(cudaMemGetInfo(free, total));
} }
GGML_CALL bool ggml_backend_cuda_register_host_buffer(void * buffer, size_t size) {
if (getenv("GGML_CUDA_NO_PINNED") != nullptr) {
return false;
}
cudaError_t err = cudaHostRegister(buffer, size, cudaHostRegisterPortable | cudaHostRegisterReadOnly);
if (err != cudaSuccess) {
// clear the error
cudaGetLastError();
fprintf(stderr, "%s: warning: failed to register %.2f MiB of pinned memory: %s\n", __func__,
size/1024.0/1024.0, cudaGetErrorString(err));
return false;
}
return true;
}
GGML_CALL void ggml_backend_cuda_unregister_host_buffer(void * buffer) {
cudaError_t err = cudaHostUnregister(buffer);
if (err != cudaSuccess) {
// clear the error
cudaGetLastError();
}
}
// backend registry // backend registry
GGML_CALL static ggml_backend_t ggml_backend_reg_cuda_init(const char * params, void * user_data) { GGML_CALL static ggml_backend_t ggml_backend_reg_cuda_init(const char * params, void * user_data) {
ggml_backend_t cuda_backend = ggml_backend_cuda_init((int) (intptr_t) user_data); ggml_backend_t cuda_backend = ggml_backend_cuda_init((int) (intptr_t) user_data);

21
ggml-cuda.h
View File

@ -17,29 +17,17 @@ extern "C" {
#define GGML_CUDA_MAX_DEVICES 16 #define GGML_CUDA_MAX_DEVICES 16
// Always success. To check if CUDA is actually loaded, use `ggml_cublas_loaded`.
GGML_API GGML_CALL void ggml_init_cublas(void);
// Returns `true` if there are available CUDA devices and cublas loads successfully; otherwise, it returns `false`.
GGML_API GGML_CALL bool ggml_cublas_loaded(void);
GGML_API GGML_CALL void * ggml_cuda_host_malloc(size_t size);
GGML_API GGML_CALL void ggml_cuda_host_free(void * ptr);
GGML_API GGML_CALL bool ggml_cuda_can_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
GGML_API GGML_CALL bool ggml_cuda_compute_forward(struct ggml_compute_params * params, struct ggml_tensor * tensor);
GGML_API GGML_CALL int ggml_cuda_get_device_count(void);
GGML_API GGML_CALL void ggml_cuda_get_device_description(int device, char * description, size_t description_size);
// backend API // backend API
GGML_API GGML_CALL ggml_backend_t ggml_backend_cuda_init(int device); GGML_API GGML_CALL ggml_backend_t ggml_backend_cuda_init(int device);
GGML_API GGML_CALL bool ggml_backend_is_cuda(ggml_backend_t backend); GGML_API GGML_CALL bool ggml_backend_is_cuda(ggml_backend_t backend);
// device buffer
GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_buffer_type(int device); GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_buffer_type(int device);
// split tensor buffer that splits matrices by rows across multiple devices // split tensor buffer that splits matrices by rows across multiple devices
GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_split_buffer_type(const float * tensor_split); GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_split_buffer_type(const float * tensor_split);
// pinned host buffer for use with the CPU backend for faster copies between CPU and GPU // pinned host buffer for use with the CPU backend for faster copies between CPU and GPU
GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_host_buffer_type(void); GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_cuda_host_buffer_type(void);
@ -47,6 +35,9 @@ GGML_API GGML_CALL int ggml_backend_cuda_get_device_count(void);
GGML_API GGML_CALL void ggml_backend_cuda_get_device_description(int device, char * description, size_t description_size); GGML_API GGML_CALL void ggml_backend_cuda_get_device_description(int device, char * description, size_t description_size);
GGML_API GGML_CALL void ggml_backend_cuda_get_device_memory(int device, size_t * free, size_t * total); GGML_API GGML_CALL void ggml_backend_cuda_get_device_memory(int device, size_t * free, size_t * total);
GGML_API GGML_CALL bool ggml_backend_cuda_register_host_buffer(void * buffer, size_t size);
GGML_API GGML_CALL void ggml_backend_cuda_unregister_host_buffer(void * buffer);
#ifdef __cplusplus #ifdef __cplusplus
} }
#endif #endif

1
ggml-kompute.cpp
View File

@ -1951,6 +1951,7 @@ static struct ggml_backend_i kompute_backend_i = {
/* .graph_plan_compute = */ NULL, /* .graph_plan_compute = */ NULL,
/* .graph_compute = */ ggml_backend_kompute_graph_compute, /* .graph_compute = */ ggml_backend_kompute_graph_compute,
/* .supports_op = */ ggml_backend_kompute_supports_op, /* .supports_op = */ ggml_backend_kompute_supports_op,
/* .offload_op = */ NULL,
/* .event_new = */ NULL, /* .event_new = */ NULL,
/* .event_free = */ NULL, /* .event_free = */ NULL,
/* .event_record = */ NULL, /* .event_record = */ NULL,

1
ggml-metal.m
View File

@ -2837,6 +2837,7 @@ static struct ggml_backend_i ggml_backend_metal_i = {
/* .graph_plan_compute = */ NULL, /* .graph_plan_compute = */ NULL,
/* .graph_compute = */ ggml_backend_metal_graph_compute, /* .graph_compute = */ ggml_backend_metal_graph_compute,
/* .supports_op = */ ggml_backend_metal_supports_op, /* .supports_op = */ ggml_backend_metal_supports_op,
/* .offload_op = */ NULL,
/* .event_new = */ NULL, /* .event_new = */ NULL,
/* .event_free = */ NULL, /* .event_free = */ NULL,
/* .event_record = */ NULL, /* .event_record = */ NULL,

1
ggml-sycl.cpp
View File

@ -17390,6 +17390,7 @@ static ggml_backend_i ggml_backend_sycl_interface = {
/* .graph_plan_compute = */ NULL, /* .graph_plan_compute = */ NULL,
/* .graph_compute = */ ggml_backend_sycl_graph_compute, /* .graph_compute = */ ggml_backend_sycl_graph_compute,
/* .supports_op = */ ggml_backend_sycl_supports_op, /* .supports_op = */ ggml_backend_sycl_supports_op,
/* .offload_op = */ NULL,
/* .event_new = */ NULL, /* .event_new = */ NULL,
/* .event_free = */ NULL, /* .event_free = */ NULL,
/* .event_record = */ NULL, /* .event_record = */ NULL,

1
ggml-vulkan.cpp
View File

@ -5699,6 +5699,7 @@ static ggml_backend_i ggml_backend_vk_interface = {
/* .graph_plan_compute = */ NULL, /* .graph_plan_compute = */ NULL,
/* .graph_compute = */ ggml_backend_vk_graph_compute, /* .graph_compute = */ ggml_backend_vk_graph_compute,
/* .supports_op = */ ggml_backend_vk_supports_op, /* .supports_op = */ ggml_backend_vk_supports_op,
/* .offload_op = */ NULL,
/* .event_new = */ NULL, /* .event_new = */ NULL,
/* .event_free = */ NULL, /* .event_free = */ NULL,
/* .event_record = */ NULL, /* .event_record = */ NULL,

19
ggml.c
View File

@ -282,8 +282,6 @@ inline static void * ggml_calloc(size_t num, size_t size) {
#else #else
#include <cblas.h> #include <cblas.h>
#endif #endif
#elif defined(GGML_USE_CUBLAS)
#include "ggml-cuda.h"
#elif defined(GGML_USE_CLBLAST) #elif defined(GGML_USE_CLBLAST)
#include "ggml-opencl.h" #include "ggml-opencl.h"
#elif defined(GGML_USE_VULKAN) #elif defined(GGML_USE_VULKAN)
@ -2640,9 +2638,7 @@ struct ggml_context * ggml_init(struct ggml_init_params params) {
GGML_PRINT_DEBUG("%s: g_state initialized in %f ms\n", __func__, (t_end - t_start)/1000.0f); GGML_PRINT_DEBUG("%s: g_state initialized in %f ms\n", __func__, (t_end - t_start)/1000.0f);
} }
#if defined(GGML_USE_CUBLAS) #if defined(GGML_USE_CLBLAST)
ggml_init_cublas();
#elif defined(GGML_USE_CLBLAST)
ggml_cl_init(); ggml_cl_init();
#elif defined(GGML_USE_VULKAN) #elif defined(GGML_USE_VULKAN)
ggml_vk_init_cpu_assist(); ggml_vk_init_cpu_assist();
@ -11105,7 +11101,6 @@ static void ggml_compute_forward_out_prod_f32(
// nb01 >= nb00 - src0 is not transposed // nb01 >= nb00 - src0 is not transposed
// compute by src0 rows // compute by src0 rows
// TODO: #if defined(GGML_USE_CUBLAS) ggml_cuda_out_prod
// TODO: #if defined(GGML_USE_CLBLAST) // TODO: #if defined(GGML_USE_CLBLAST)
#if defined(GGML_USE_ACCELERATE) || defined(GGML_USE_OPENBLAS) #if defined(GGML_USE_ACCELERATE) || defined(GGML_USE_OPENBLAS)
@ -11305,7 +11300,6 @@ static void ggml_compute_forward_out_prod_q_f32(
// nb01 >= nb00 - src0 is not transposed // nb01 >= nb00 - src0 is not transposed
// compute by src0 rows // compute by src0 rows
// TODO: #if defined(GGML_USE_CUBLAS) ggml_cuda_out_prod
// TODO: #if defined(GGML_USE_ACCELERATE) || defined(GGML_USE_OPENBLAS) || defined(GGML_USE_CLBLAST) // TODO: #if defined(GGML_USE_ACCELERATE) || defined(GGML_USE_OPENBLAS) || defined(GGML_USE_CLBLAST)
if (params->type == GGML_TASK_TYPE_INIT) { if (params->type == GGML_TASK_TYPE_INIT) {
@ -16051,14 +16045,7 @@ static void ggml_compute_forward(struct ggml_compute_params * params, struct ggm
return; return;
} }
#ifdef GGML_USE_CUBLAS #if defined(GGML_USE_VULKAN)
bool skip_cpu = ggml_cuda_compute_forward(params, tensor);
if (skip_cpu) {
return;
}
GGML_ASSERT(tensor->src[0] == NULL || tensor->src[0]->backend == GGML_BACKEND_TYPE_CPU);
GGML_ASSERT(tensor->src[1] == NULL || tensor->src[1]->backend == GGML_BACKEND_TYPE_CPU);
#elif defined(GGML_USE_VULKAN)
const bool skip_cpu = ggml_vk_compute_forward_cpu_assist(params, tensor); const bool skip_cpu = ggml_vk_compute_forward_cpu_assist(params, tensor);
#ifdef GGML_VULKAN_CHECK_RESULTS #ifdef GGML_VULKAN_CHECK_RESULTS
if (skip_cpu) { if (skip_cpu) {
@ -16070,7 +16057,7 @@ static void ggml_compute_forward(struct ggml_compute_params * params, struct ggm
} }
GGML_ASSERT(tensor->src[0] == NULL || tensor->src[0]->backend == GGML_BACKEND_TYPE_CPU); GGML_ASSERT(tensor->src[0] == NULL || tensor->src[0]->backend == GGML_BACKEND_TYPE_CPU);
GGML_ASSERT(tensor->src[1] == NULL || tensor->src[1]->backend == GGML_BACKEND_TYPE_CPU); GGML_ASSERT(tensor->src[1] == NULL || tensor->src[1]->backend == GGML_BACKEND_TYPE_CPU);
#endif // GGML_USE_CUBLAS #endif // GGML_USE_VULKAN
#ifdef GGML_USE_SYCL #ifdef GGML_USE_SYCL
bool skip_cpu = ggml_sycl_compute_forward(params, tensor); bool skip_cpu = ggml_sycl_compute_forward(params, tensor);

67
llama.cpp
View File

@ -2040,6 +2040,11 @@ struct llama_model {
ggml_free(ctx); ggml_free(ctx);
} }
for (ggml_backend_buffer_t buf : bufs) { for (ggml_backend_buffer_t buf : bufs) {
#ifdef GGML_USE_CUBLAS
if (ggml_backend_buffer_get_type(buf) == ggml_backend_cpu_buffer_type()) {
ggml_backend_cuda_unregister_host_buffer(ggml_backend_buffer_get_base(buf));
}
#endif
ggml_backend_buffer_free(buf); ggml_backend_buffer_free(buf);
} }
} }
@ -5033,6 +5038,13 @@ static bool llm_load_tensors(
size_t first, last; size_t first, last;
ml.get_mapping_range(&first, &last, ctx); ml.get_mapping_range(&first, &last, ctx);
buf = ggml_backend_cpu_buffer_from_ptr((char *) ml.mapping->addr + first, last - first); buf = ggml_backend_cpu_buffer_from_ptr((char *) ml.mapping->addr + first, last - first);
#ifdef GGML_USE_CUBLAS
if (n_layer >= n_gpu_layers) {
ggml_backend_cuda_register_host_buffer(
ggml_backend_buffer_get_base(buf),
ggml_backend_buffer_get_size(buf));
}
#endif
} }
#ifdef GGML_USE_METAL #ifdef GGML_USE_METAL
else if (ml.use_mmap && buft == ggml_backend_metal_buffer_type()) { else if (ml.use_mmap && buft == ggml_backend_metal_buffer_type()) {
@ -8231,7 +8243,6 @@ struct llm_build_context {
cur = llm_build_kv(ctx0, model, hparams, kv_self, gf, cur = llm_build_kv(ctx0, model, hparams, kv_self, gf,
model.layers[il].wo, model.layers[il].bo, model.layers[il].wo, model.layers[il].bo,
Kcur, Vcur, Qcur, KQ_mask, nullptr, n_ctx, n_tokens, kv_head, n_kv, 1.0f/sqrtf(float(n_embd_head)), cb, il); Kcur, Vcur, Qcur, KQ_mask, nullptr, n_ctx, n_tokens, kv_head, n_kv, 1.0f/sqrtf(float(n_embd_head)), cb, il);
cb(cur, "kqv_out", il);
} }
struct ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA); struct ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);
@ -8601,12 +8612,15 @@ static struct ggml_cgraph * llama_build_graph(
} }
// norm may be automatically assigned to the backend of the previous layer, increasing data transfer between backends // norm may be automatically assigned to the backend of the previous layer, increasing data transfer between backends
// to fix this, we assign the norm layer manually to the backend of its layer // FIXME: fix in ggml_backend_sched
if (il != -1 && strcmp(name, "norm") == 0) { const bool full_offload = lctx.model.n_gpu_layers > (int)lctx.model.hparams.n_layer;
for (auto * backend : lctx.backends) { if (batch.n_tokens < 32 || full_offload) {
if (ggml_backend_buft_supports_backend(lctx.model.buft_layer[il].buft, backend)) { if (il != -1 && strcmp(name, "norm") == 0) {
ggml_backend_sched_set_tensor_backend(lctx.sched, cur, backend); for (auto * backend : lctx.backends) {
break; if (ggml_backend_buft_supports_backend(lctx.model.buft_layer[il].buft, backend)) {
ggml_backend_sched_set_tensor_backend(lctx.sched, cur, backend);
break;
}
} }
} }
} }
@ -13107,27 +13121,25 @@ struct llama_context * llama_new_context_with_model(
ctx->backends.push_back(ctx->backend_metal); ctx->backends.push_back(ctx->backend_metal);
} }
#elif defined(GGML_USE_CUBLAS) #elif defined(GGML_USE_CUBLAS)
if (model->n_gpu_layers > 0) { if (model->split_mode == LLAMA_SPLIT_MODE_NONE || model->split_mode == LLAMA_SPLIT_MODE_ROW) {
// with split_mode LLAMA_SPLIT_MODE_NONE or LLAMA_SPLIT_MODE_ROW, only the main GPU backend is used // with split_mode LLAMA_SPLIT_MODE_NONE or LLAMA_SPLIT_MODE_ROW, only the main GPU backend is used
if (model->split_mode == LLAMA_SPLIT_MODE_NONE || model->split_mode == LLAMA_SPLIT_MODE_ROW) { ggml_backend_t backend = ggml_backend_cuda_init(model->main_gpu);
ggml_backend_t backend = ggml_backend_cuda_init(model->main_gpu); if (backend == nullptr) {
LLAMA_LOG_ERROR("%s: failed to initialize CUDA%d backend\n", __func__, model->main_gpu);
llama_free(ctx);
return nullptr;
}
ctx->backends.push_back(backend);
} else {
// LLAMA_SPLIT_MODE_LAYER requires a backend for each GPU
for (int device = 0; device < ggml_backend_cuda_get_device_count(); ++device) {
ggml_backend_t backend = ggml_backend_cuda_init(device);
if (backend == nullptr) { if (backend == nullptr) {
LLAMA_LOG_ERROR("%s: failed to initialize CUDA%d backend\n", __func__, model->main_gpu); LLAMA_LOG_ERROR("%s: failed to initialize CUDA%d backend\n", __func__, device);
llama_free(ctx); llama_free(ctx);
return nullptr; return nullptr;
} }
ctx->backends.push_back(backend); ctx->backends.push_back(backend);
} else {
// LLAMA_SPLIT_MODE_LAYER requires a backend for each GPU
for (int device = 0; device < ggml_backend_cuda_get_device_count(); ++device) {
ggml_backend_t backend = ggml_backend_cuda_init(device);
if (backend == nullptr) {
LLAMA_LOG_ERROR("%s: failed to initialize CUDA%d backend\n", __func__, device);
llama_free(ctx);
return nullptr;
}
ctx->backends.push_back(backend);
}
} }
} }
#elif defined(GGML_USE_VULKAN) #elif defined(GGML_USE_VULKAN)
@ -13285,14 +13297,17 @@ struct llama_context * llama_new_context_with_model(
ggml_backend_t backend = ctx->backends[i]; ggml_backend_t backend = ctx->backends[i];
ggml_backend_buffer_type_t buft = backend_buft[i]; ggml_backend_buffer_type_t buft = backend_buft[i];
size_t size = ggml_backend_sched_get_buffer_size(ctx->sched, backend); size_t size = ggml_backend_sched_get_buffer_size(ctx->sched, backend);
LLAMA_LOG_INFO("%s: %10s compute buffer size = %8.2f MiB\n", __func__, if (size > 1) {
ggml_backend_buft_name(buft), LLAMA_LOG_INFO("%s: %10s compute buffer size = %8.2f MiB\n", __func__,
size / 1024.0 / 1024.0); ggml_backend_buft_name(buft),
size / 1024.0 / 1024.0);
}
} }
// note: the number of splits during measure is higher than during inference due to the kv shift // note: the number of splits during measure is higher than during inference due to the kv shift
int n_splits = ggml_backend_sched_get_n_splits(ctx->sched); int n_splits = ggml_backend_sched_get_n_splits(ctx->sched);
LLAMA_LOG_INFO("%s: graph splits: %d\n", __func__, n_splits); LLAMA_LOG_INFO("%s: graph nodes = %d\n", __func__, gf->n_nodes);
LLAMA_LOG_INFO("%s: graph splits = %d\n", __func__, n_splits);
} }
} }