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

Fix preci failures

Browse source
This commit is contained in:
z5269887 2024-04-23 20:50:00 +08:00
parent 141eb5107f
commit e0a3679aeb
2 changed files with 5 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

2
llama.cpp
View file

@ -13618,7 +13618,7 @@ static void llama_model_quantize_internal(const std::string & fname_inp, const s
// Set split info if needed // Set split info if needed
if (n_split > 1) { if (n_split > 1) {
for (int i = 0; i < ctx_outs.size(); ++i) { for (size_t i = 0; i < ctx_outs.size(); ++i) {
gguf_set_val_u16(ctx_outs[i], ml.llm_kv(LLM_KV_SPLIT_NO).c_str(), i); gguf_set_val_u16(ctx_outs[i], ml.llm_kv(LLM_KV_SPLIT_NO).c_str(), i);
gguf_set_val_u16(ctx_outs[i], ml.llm_kv(LLM_KV_SPLIT_COUNT).c_str(), n_split); gguf_set_val_u16(ctx_outs[i], ml.llm_kv(LLM_KV_SPLIT_COUNT).c_str(), n_split);
gguf_set_val_i32(ctx_outs[i], ml.llm_kv(LLM_KV_SPLIT_TENSORS_COUNT).c_str(), ml.n_tensors); gguf_set_val_i32(ctx_outs[i], ml.llm_kv(LLM_KV_SPLIT_TENSORS_COUNT).c_str(), ml.n_tensors);