jhenhong's picture
examples : fix n_gpu_layers usage in talk-llama (#1441)
e0ea7d1 unverified