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

Apply suggestions from code review

Browse source 
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
This commit is contained in:
Xuan Son Nguyen 2024-10-22 15:23:42 +02:00 • committed by GitHub
parent 11b1564efb
commit f732003622
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
src/llama.cpp
View file

@ -17170,7 +17170,7 @@ static int llama_decode_internal(
// temporary allocate memory for the input batch if needed // temporary allocate memory for the input batch if needed
llama_batch_allocr batch_allocr(lctx, inp_batch); llama_batch_allocr batch_allocr(lctx, inp_batch);
llama_batch batch = batch_allocr.batch; const llama_batch & batch = batch_allocr.batch;
const uint32_t n_tokens_all = batch.n_tokens; const uint32_t n_tokens_all = batch.n_tokens;
const auto & model = lctx.model; const auto & model = lctx.model;
@ -17488,7 +17488,7 @@ static int llama_encode_internal(
// temporary allocate memory for the input batch if needed // temporary allocate memory for the input batch if needed
llama_batch_allocr batch_allocr(lctx, inp_batch); llama_batch_allocr batch_allocr(lctx, inp_batch);
llama_batch batch = batch_allocr.batch; const llama_batch & batch = batch_allocr.batch;
const uint32_t n_tokens = batch.n_tokens; const uint32_t n_tokens = batch.n_tokens;
const auto & model = lctx.model; const auto & model = lctx.model;