From 6c7148c0bfec31fc98d71cf22d0c10530b28482b Mon Sep 17 00:00:00 2001 From: cheetahadmin Date: Mon, 20 Oct 2025 05:47:30 +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 bbdcf2c..d8a83a2 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: 15 +max_batch_size: 150 # 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: "int8"} + value: {string_value: "int4"} } ]