highperfocused/ollama
1
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2025-11-11 19:57:43 +01:00
Code Issues Packages Projects Releases Wiki Activity

chore: update mllama to use ollama engine (#10637)

Browse Source
This commit is contained in:
Michael Yang
2025-05-13 17:36:02 -07:00
committed by GitHub
parent 0478d440f0
commit 23125648b8
67 changed files with 785 additions and 4354 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
llama/llama.cpp/include/llama.h vendored
View File

@@ -258,7 +258,6 @@ extern "C" {
llama_token * token;
float * embd;
int32_t n_embd;
llama_pos * pos;
int32_t * n_seq_id;
llama_seq_id ** seq_id;
@@ -366,7 +365,6 @@ extern "C" {
bool flash_attn; // whether to use flash attention [EXPERIMENTAL]
bool no_perf; // whether to measure performance timings
bool op_offload; // whether to offload host tensor operations to device
bool cross_attn; // whether to use cross attention
};
// model quantization parameters
@@ -466,10 +464,6 @@ extern "C" {
struct llama_context_params params),
"use llama_init_from_model instead");
// TODO (jmorganca): this should most likely be passed in as part of a batch
// and not set on the context for all batches.
LLAMA_API void llama_set_cross_attention(struct llama_context * ctx, bool cross_attn_state);
// Frees all allocated memory
LLAMA_API void llama_free(struct llama_context * ctx);