Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-20 01:31:42 +02:00
Code Issues Packages Projects Releases Wiki Activity

some tweaks and cleanup

Browse Source
This commit is contained in:
Concedo
2025-01-13 23:50:54 +08:00
parent 636beac6d2
commit e77d566268
3 changed files with 8 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-context.cpp
+1 -1
View File
@@ -87,7 +87,7 @@ void llama_set_inputs(llama_context & lctx, const llama_ubatch & ubatch) {
//GGML_ASSERT(lctx.inp_out_ids && "every model that can must skip unused outputs");
if (!lctx.inp_out_ids) {
LLAMA_LOG_WARN("%s: 'lctx.inp_out_ids' is not created\n", __func__);
//LLAMA_LOG_WARN("%s: 'lctx.inp_out_ids' is not created\n", __func__);
} else {
const int64_t n_tokens = ubatch.n_tokens;
src/llama.cpp
+1 -1
View File
@@ -11124,7 +11124,7 @@ static int llama_decode_impl(
GGML_ASSERT_CONTINUE(n_tokens_all <= cparams.n_batch);
GGML_ASSERT_CONTINUE((cparams.causal_attn || cparams.n_ubatch >= n_tokens_all) && "non-causal attention requires n_ubatch >= n_tokens");
//GGML_ASSERT_CONTINUE((cparams.causal_attn || cparams.n_ubatch >= n_tokens_all) && "non-causal attention requires n_ubatch >= n_tokens");
if (lctx.t_compute_start_us == 0) {
lctx.t_compute_start_us = ggml_time_us();