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

remove hack

Browse source
This commit is contained in:
fmz 2024-06-28 07:10:59 -07:00
parent a8d4afb97b
commit f42285f0e5
2 changed files with 0 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

5
include/llama.h
View file

@ -652,11 +652,6 @@ extern "C" {
// State / sessions // State / sessions
// //
// hack
void llama_set_logits_all(
struct llama_context * ctx,
bool logits_all);
// Returns the maximum size in bytes of the state (rng, logits, embedding // Returns the maximum size in bytes of the state (rng, logits, embedding
// and kv_cache) - will often be smaller after compacting tokens // and kv_cache) - will often be smaller after compacting tokens
LLAMA_API size_t llama_state_get_size(const struct llama_context * ctx); LLAMA_API size_t llama_state_get_size(const struct llama_context * ctx);

7
src/llama.cpp
View file

@ -13106,13 +13106,6 @@ static void llama_graph_compute(
// fprintf(stderr, "splits: %d\n", ggml_backend_sched_get_n_splits(lctx.sched)); // fprintf(stderr, "splits: %d\n", ggml_backend_sched_get_n_splits(lctx.sched));
} }
void llama_set_logits_all(
struct llama_context * ctx,
bool logits_all
) {
ctx->logits_all = logits_all;
}
// decode a batch of tokens by evaluating the transformer // decode a batch of tokens by evaluating the transformer
// //
// - lctx: llama context // - lctx: llama context