From 756b09f6bc54682c9221af62df1aaad2d2478c20 Mon Sep 17 00:00:00 2001 From: Masahiro Masuda Date: Wed, 10 Jan 2024 21:59:52 +0000 Subject: [PATCH] fix build --- mlc_llm/core.py | 1 + 1 file changed, 1 insertion(+) diff --git a/mlc_llm/core.py b/mlc_llm/core.py index 7a69562696..f7afbbb693 100644 --- a/mlc_llm/core.py +++ b/mlc_llm/core.py @@ -593,6 +593,7 @@ def mod_transform_before_build( # This is equivalent to prefill but without KV cache. It is used for # determining the number of paged cache blocks that can be allocated. model_names.append("evaluate") + model_names.append("evaluate_multi_query") if args.sep_embed: model_names = ["embed", "prefill_with_embed"] + model_names[1:]