ver4a/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

llama : add support for control vectors (#5970)

Browse source 
* control vector api and implementation

* control-vectors : minor code style updates

* disable control vector when data == nullptr

use -1 for disabled range (also on init) in case we ever support controlling layer 0 (embeddings)

---------

Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Theia Vogel 2024-03-15 13:43:02 -07:00 • committed by GitHub
parent 12247f4c69
commit 877b4d0c62
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
4 changed files with 392 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

23
llama.h
View file

@ -388,6 +388,7 @@ extern "C" {
LLAMA_API int32_t llama_n_vocab (const struct llama_model * model);
LLAMA_API int32_t llama_n_ctx_train(const struct llama_model * model);
LLAMA_API int32_t llama_n_embd (const struct llama_model * model);
LLAMA_API int32_t llama_n_layer (const struct llama_model * model);
// Get the model's RoPE frequency scaling factor
LLAMA_API float llama_rope_freq_scale_train(const struct llama_model * model);
@ -435,10 +436,24 @@ extern "C" {
// Returns 0 on success
LLAMA_API int32_t llama_model_apply_lora_from_file(
const struct llama_model * model,
const char * path_lora,
float scale,
const char * path_base_model,
int32_t n_threads);
const char * path_lora,
float scale,
const char * path_base_model,
int32_t n_threads);
// Apply a loaded control vector to a llama_context, or if data is NULL, clear
// the currently loaded vector.
// n_embd should be the size of a single layer's control, and data should point
// to an n_embd x n_layers buffer starting from layer 1.
// il_start and il_end are the layer range the vector should apply to (both inclusive)
// See llama_control_vector_load in common to load a control vector.
LLAMA_API int32_t llama_control_vector_apply(
struct llama_context * lctx,
const float * data,
size_t len,
int32_t n_embd,
int32_t il_start,
int32_t il_end);
//
// KV cache