Public/koboldcpp
1
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2026-09-01 18:47:36 +02:00
Code Issues Packages Projects Releases Wiki Activity

Merge commit '12280ae9058600b0b14335fe59848e601e10eebc' into concedo_experimental

Browse Source 
# Conflicts:
#	.github/workflows/build.yml
#	common/CMakeLists.txt
#	docs/docker.md
#	examples/model-conversion/scripts/causal/compare-logits.py
#	ggml/src/ggml-hexagon/htp/rope-ops.c
#	tests/test-backend-ops.cpp
#	tests/test-barrier.cpp
#	tools/server/CMakeLists.txt
#	tools/server/README.md
This commit is contained in:
Concedo
2025-12-16 16:29:01 +08:00
parent 78bbe89956 12280ae905
commit e88bf41fdc
49 changed files with 1380 additions and 583 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-context.cpp
-8
View File
@@ -96,14 +96,6 @@ llama_context::llama_context(
// with causal attention, the batch size is limited by the context size
cparams.n_batch = cparams.causal_attn ? std::min(cparams.n_ctx, params.n_batch) : params.n_batch;
// the batch has to be at least GGML_KQ_MASK_PAD because we will be padding the KQ_mask
// this is required by GPU kernels in order to avoid out-of-bounds accesses (e.g. ggml_flash_attn_ext)
// ref: https://github.com/ggerganov/llama.cpp/pull/5021
// TODO: this padding is not needed for the cache-less context so we should probably move it to llama_memory
if (cparams.n_batch < GGML_KQ_MASK_PAD) {
LLAMA_LOG_WARN("%s: n_batch is less than GGML_KQ_MASK_PAD - increasing to %d\n", __func__, GGML_KQ_MASK_PAD);
cparams.n_batch = GGML_KQ_MASK_PAD;
}
cparams.n_ubatch = std::min(cparams.n_batch, params.n_ubatch == 0 ? params.n_batch : params.n_ubatch);
cparams.op_offload = params.op_offload;