public_git/whisper.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/whisper.cpp.git synced 2026-10-04 21:41:24 +02:00
Code Issues Packages Projects Releases Wiki Activity

talk-llama : sync llama.cpp

Browse Source
This commit is contained in:
Georgi Gerganov
2026-03-16 13:10:15 +02:00
parent ab1252c19e
commit 2bc630f197
131 changed files with 2374 additions and 1519 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/talk-llama/models/smallthinker.cpp
+4 -4
View File
@@ -2,10 +2,10 @@
template <bool iswa>
llm_build_smallthinker<iswa>::llm_build_smallthinker(const llama_model & model, const llm_graph_params & params) : llm_graph_context(params){
const int64_t n_embd_head = hparams.n_embd_head_v;
const int64_t n_embd_head = hparams.n_embd_head_v();
GGML_ASSERT(n_embd_head == hparams.n_embd_head_k);
GGML_ASSERT(n_embd_head == hparams.n_rot);
GGML_ASSERT(n_embd_head == hparams.n_embd_head_k());
GGML_ASSERT(n_embd_head == n_rot);
ggml_tensor * cur;
ggml_tensor * inpL;
@@ -93,7 +93,7 @@ llm_build_smallthinker<iswa>::llm_build_smallthinker(const llama_model & model,
nullptr,
n_expert, n_expert_used,
LLM_FFN_RELU, true,
false, 0.0,
hparams.expert_weights_scale,
static_cast<llama_expert_gating_func_type>(hparams.expert_gating_func),
il, probs);