EngineX-Ascend/enginex-ascend-910-llama.cpp
10
0
Fork 0
Code Issues Pull Requests Actions 4 Projects Releases Wiki Activity

llama : reuse compute graphs (#14482)

Browse Source 
* llama : reuse compute graphs

ggml-ci

* llama-bench : add graph reuse parameter

ggml-ci

* cont : remove the parameter and the sched resets

ggml-ci

* graph : rename update() to can_reuse()

ggml-ci

* params : remove is_same()

ggml-ci

* graph : set res->params in llm_graph_context constructor

ggml-ci

* graph : avoid set_max_nodes in llm_graph_result

ggml-ci

* kv-cache : reuse llama_context's graph result instance

ggml-ci

* context : reset the previous graph result upon memory updates

ggml-ci

* batch : llama_ubatch now carries its data instead of pointing to balloc

ggml-ci

* merge : fix build

ggml-ci

* graph : fix can_reuse() checks when flash-attention is disabled

* graph : move llm_graph_result impl in source file + debug env

ggml-ci
This commit is contained in:
Georgi Gerganov
2025-07-17 19:08:33 +03:00
committed by GitHub
parent 086cf81e88
commit 01612b7409
12 changed files with 548 additions and 289 deletions
Download Patch File Download Diff File Expand all files Collapse all files

38
src/llama-batch.h
View File

@@ -8,12 +8,17 @@
#include <vector>
#include <set>
#include <bitset>
#include <memory>
#include <unordered_map>
// keep this struct lightweight
// it points to data in `llama_batch_allocr`
struct llama_ubatch {
bool equal_seqs;
bool equal_seqs() const {
return b_equal_seqs != 0;
}
uint32_t b_equal_seqs; // note: this is a boolean, but we use an int32_t for alignment
// otherwise address sanitizer complains
// TODO: whole_seqs for embeddings?
uint32_t n_tokens; // total tokens (n_seq_tokens * n_seqs)
@@ -34,6 +39,20 @@ struct llama_ubatch {
llama_seq_id * seq_id_unq; // [n_seqs_unq] | s | seq_id
int32_t * seq_idx; // [LLAMA_MAX_SEQ] | - | seq_idx
int8_t * output; // [n_tokens] | i | -
struct data_t {
std::vector<llama_token> token;
std::vector<float> embd;
std::vector<llama_pos> pos;
std::vector<int32_t> n_seq_id;
std::vector<llama_seq_id *> seq_id;
std::vector<llama_seq_id> seq_id_unq;
std::vector<int32_t> seq_idx;
std::vector<int8_t> output;
};
// the llama_ubatch pointers above point to this data if set. otherwise - points to non-owning data
std::shared_ptr<data_t> data;
};
// a helper for sanitizing, fulfilling and splitting a batch
@@ -137,20 +156,5 @@ private:
// used[i] indicates if token i has already been used in a previous ubatch
std::vector<bool> used;
// llama_ubatch points to this data:
struct ubatch {
std::vector<llama_token> token;
std::vector<float> embd;
std::vector<llama_pos> pos;
std::vector<int32_t> n_seq_id;
std::vector<llama_seq_id *> seq_id;
std::vector<llama_seq_id> seq_id_unq;
std::vector<int32_t> seq_idx;
std::vector<int8_t> output;
};
// current splitting state:
std::vector<ubatch> ubatches;
int debug;
};