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

llama : second attempt to refactor vision API

Browse source
This commit is contained in:
Xuan Son Nguyen 2025-01-18 20:56:35 +01:00
parent 2a458d1a9d
commit 0a81051ae2
20 changed files with 695 additions and 145 deletions
Download patch file Download diff file Expand all files Collapse all files

2
src/llama-context.cpp
View file

@ -73,7 +73,7 @@ void llama_set_inputs(llama_context & lctx, const llama_ubatch & ubatch) {
ggml_backend_tensor_set(lctx.inp_tokens, ubatch.token, 0, n_tokens*ggml_element_size(lctx.inp_tokens));
}
if (ubatch.embd) {
if (ubatch.embd && !ubatch.embd_tensor) {
const int64_t n_embd = hparams.n_embd;
const int64_t n_tokens = ubatch.n_tokens;