From ba6248eec083b5af3b36fad32d057f860333c6cd Mon Sep 17 00:00:00 2001 From: gakada Date: Sat, 17 May 2025 19:10:49 +0900 Subject: [PATCH] Exl3: fix add_bos in generator --- backends/exllamav3/model.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/backends/exllamav3/model.py b/backends/exllamav3/model.py index 60cf5d0..7b39a33 100644 --- a/backends/exllamav3/model.py +++ b/backends/exllamav3/model.py @@ -844,7 +844,7 @@ class ExllamaV3Container(BaseModelContainer): job = AsyncJob( self.generator, sampler=sampler, - input_ids=self.tokenizer.encode(prompt, add_bos=False), + input_ids=input_ids, max_new_tokens=max_tokens, stop_conditions=stop_conditions, banned_strings=params.banned_strings,