From 4bb6be215e267ff18e7f9e87abef69b50ff33f84 Mon Sep 17 00:00:00 2001 From: cheetahadmin Date: Mon, 20 Oct 2025 04:41:24 +0000 Subject: [PATCH] Update config.pbtxt --- config.pbtxt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/config.pbtxt b/config.pbtxt index 332019f..1f889ab 100644 --- a/config.pbtxt +++ b/config.pbtxt @@ -1,7 +1,7 @@ # Triton backend to use name: "base-gemma-3-1b-it" backend: "python" -max_batch_size: 4 +max_batch_size: 8 # Triton should expect as input a single string # input of variable length named 'text_input' @@ -119,7 +119,7 @@ parameters: [ { key: "quantization", - value: {string_value: "int4"} + value: {string_value: "int8"} } ]