vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

context : minor

Browse source 
ggml-ci
This commit is contained in:
Georgi Gerganov 2025-01-15 10:54:21 +02:00
parent 17b363afd3
commit a19f671fe0
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
5 changed files with 37 additions and 47 deletions
Download patch file Download diff file Expand all files Collapse all files

8
src/llama-context.h
View file

@ -18,7 +18,7 @@ struct llama_context {
llama_context(const llama_model & model)
: model(model)
, t_start_us(model.t_start_us)
, t_load_us(model.t_load_us) {}
, t_load_us (model.t_load_us) {}
const struct llama_model & model;
@ -107,13 +107,11 @@ struct llama_context {
struct ggml_tensor * inp_pos_bucket; // I32 [n_batch|n_kv, n_batch]
struct ggml_tensor * inp_embd_enc; // F32 [n_embd, n_outputs_enc]
struct ggml_tensor * inp_KQ_mask_cross; // F32 [n_outputs_enc, n_batch]
void set_k_shift(llama_kv_cache & kv);
};
// TODO: make these methods of llama_context
void llama_set_k_shift(struct llama_context & lctx);
void llama_set_s_copy(struct llama_context & lctx);
void llama_set_inputs(llama_context & lctx, const llama_ubatch & ubatch);
// Make sure enough space is available for outputs.