From 645a6f806846cea1b51403c900a67d694ebde989 Mon Sep 17 00:00:00 2001 From: drbh Date: Fri, 2 Aug 2024 21:48:04 +0000 Subject: [PATCH] fix: typo tweak --- server/text_generation_server/layers/attention/cuda.py | 1 - 1 file changed, 1 deletion(-) diff --git a/server/text_generation_server/layers/attention/cuda.py b/server/text_generation_server/layers/attention/cuda.py index 7e42eddf..e74c9ba9 100644 --- a/server/text_generation_server/layers/attention/cuda.py +++ b/server/text_generation_server/layers/attention/cuda.py @@ -261,7 +261,6 @@ else: raise NotImplementedError( "window_size_left is only available with flash attn v2" ) - if softcap is not None: raise NotImplementedError("softcap is only available with flash attn v2")