From 98684d58149a45938faeeee02743f60bc85bfc86 Mon Sep 17 00:00:00 2001 From: groupuser Date: Wed, 23 Apr 2025 12:51:12 +0000 Subject: [PATCH] Automatically created by CHEETAH --- config.pbtxt | 86 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 86 insertions(+) create mode 100644 config.pbtxt diff --git a/config.pbtxt b/config.pbtxt new file mode 100644 index 0000000..cc3a984 --- /dev/null +++ b/config.pbtxt @@ -0,0 +1,86 @@ +name: "gemma-3-1b-it" +max_batch_size: 0 +backend: "python" + +# Triton should expect as input a single string +# input of variable length named 'text_input' +input [ + { + name: "text_input" + data_type: TYPE_STRING + dims: [ -1 ] + }, + { + name: "max_length" + data_type: TYPE_INT32 + dims: [ 1 ] + optional: true + }, + { + name: "max_new_tokens" + data_type: TYPE_INT32 + dims: [ 1 ] + optional: true + }, + { + name: "do_sample" + data_type: TYPE_BOOL + dims: [ 1 ] + optional: true + }, + { + name: "top_k" + data_type: TYPE_INT32 + dims: [ 1 ] + optional: true + }, + { + name: "top_p" + data_type: TYPE_FP32 + dims: [ 1 ] + optional: true + }, + { + name: "temperature" + data_type: TYPE_FP32 + dims: [ 1 ] + optional: true + }, + { + name: "repetition_penalty" + data_type: TYPE_FP32 + dims: [ 1 ] + optional: true + }, + { + name: "stream" + data_type: TYPE_BOOL + dims: [ 1 ] + optional: true + } +] + + +# Triton should expect to respond with a single string +# output of variable length named 'text_output' +output [ + { + name: "text_output" + data_type: TYPE_STRING + dims: [ -1 ] + } +] + +parameters: [ + { + key: "enable_inference_trace", + value: {string_value: "False"} + } +] + +instance_group [ + { + kind: KIND_AUTO, + count: 1 + } +] \ No newline at end of file