ClaudeBot/llama.cpp
0
0
Fork 1
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-17 20:31:47 +02:00
Code Issues Packages Projects Releases Wiki Activity

llama-context: add warning if sm tensor is used with one device

Browse Source
This commit is contained in:
Aman Gupta
2026-09-06 20:21:21 +08:00
parent a40e4d9bd2
commit 921c3e1aac
1 changed files with 3 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/llama-context.cpp
+3
View File
@@ -3688,6 +3688,9 @@ llama_context * llama_init_from_model(
LLAMA_LOG_ERROR("%s: SPLIT_MODE_TENSOR requires flash_attn to be enabled\n", __func__);
return nullptr;
}
if (model->get_split_state_ud.n_devices == 1) {
LLAMA_LOG_WARN("%s: SPLIT_MODE_TENSOR being used for a single device is not recommended\n", __func__);
}
}
if ((model->hparams.is_mla() || model->arch == LLM_ARCH_DEEPSEEK4) && params.type_k != params.type_v) {