From a7a223070beee190306894ee49303bbef52e1f10 Mon Sep 17 00:00:00 2001 From: Daniel Han Date: Fri, 21 Mar 2025 17:08:00 -0700 Subject: [PATCH] Update llama.py --- unsloth/models/llama.py | 1 + 1 file changed, 1 insertion(+) diff --git a/unsloth/models/llama.py b/unsloth/models/llama.py index d20e2b4b75..7557a96120 100644 --- a/unsloth/models/llama.py +++ b/unsloth/models/llama.py @@ -1575,6 +1575,7 @@ def unsloth_fast_generate( kwargs["pad_token_id"] = kwargs.pop("pad_token_id", model_eos_token_id) # Mixed precision autocast + print(args, kwargs) with torch.inference_mode(), torch.autocast(device_type = "cuda", dtype = dtype): output = self._old_generate(*args, **kwargs) pass