From e952e4cfd3edd7fee9717e4f940793e229d2a553 Mon Sep 17 00:00:00 2001 From: OlivierDehaene <23298448+OlivierDehaene@users.noreply.github.com> Date: Thu, 18 Jul 2024 15:59:36 +0200 Subject: [PATCH] fix(server): fix cohere --- .../models/custom_modeling/flash_cohere_modeling.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/server/text_generation_server/models/custom_modeling/flash_cohere_modeling.py b/server/text_generation_server/models/custom_modeling/flash_cohere_modeling.py index 25719b99..49f5a81b 100644 --- a/server/text_generation_server/models/custom_modeling/flash_cohere_modeling.py +++ b/server/text_generation_server/models/custom_modeling/flash_cohere_modeling.py @@ -259,8 +259,8 @@ class FlashCohereAttention(torch.nn.Module): cu_seqlen_prefill, kv_cache, block_tables, - input_lengths, slots, + input_lengths, max_s, ): qkv = self.query_key_value(hidden_states)