Skip to content

Distribute and run LLMs with a single file.

Remove n-gpu-layer limitation (#534)

BIGWONG01b8d49443ea
diff --git a/llama.cpp/llama.cpp b/llama.cpp/llama.cpp
index 18da7731..8187c5ae 100644
--- a/llama.cpp/llama.cpp
+++ b/llama.cpp/llama.cpp
@@ -5682,7 +5682,7 @@ static bool llm_load_tensors(
auto & hparams = model.hparams;
- n_gpu_layers = std::min(n_gpu_layers, (int)hparams.n_layer); // [jart]
+ // n_gpu_layers = std::min(n_gpu_layers, (int)hparams.n_layer); // [jart]
n_gpu_layers = std::max(n_gpu_layers, 0); // [jart]
model.split_mode = split_mode;