diff --git a/config.pbtxt b/config.pbtxt index 65c0ee0..130a973 100644 --- a/config.pbtxt +++ b/config.pbtxt @@ -1,5 +1,6 @@ -# Triton Backend for Python. -backend: "python" +# Triton Backend for vLLM. +backend: "vllm" +platform: "vllm" max_batch_size: 0 input [ @@ -59,12 +60,6 @@ output [ ] -parameters: [ - { - key: "model_path", - value: {string_value: "/cheetah/input/model/groupuser/TinyLlama-1.1B-Chat-v1.0"} - } -] instance_group [