From f5b81c9767938f55d65415e6a3597b4211a6db39 Mon Sep 17 00:00:00 2001 From: Daniel Walmsley Date: Fri, 14 Jun 2024 16:01:26 -0700 Subject: [PATCH] Fix Stream Generator on MacOS --- TTS/tts/layers/xtts/stream_generator.py | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/TTS/tts/layers/xtts/stream_generator.py b/TTS/tts/layers/xtts/stream_generator.py index e12f8995cf..dd07e9dc07 100644 --- a/TTS/tts/layers/xtts/stream_generator.py +++ b/TTS/tts/layers/xtts/stream_generator.py @@ -183,10 +183,12 @@ def generate( requires_attention_mask = "encoder_outputs" not in model_kwargs if model_kwargs.get("attention_mask", None) is None and requires_attention_mask and accepts_attention_mask: + pad_token_tensor = torch.tensor([generation_config.pad_token_id], device=inputs_tensor.device) if generation_config.pad_token_id is not None else None + eos_token_tensor = torch.tensor([generation_config.eos_token_id], device=inputs_tensor.device) if generation_config.eos_token_id is not None else None model_kwargs["attention_mask"] = self._prepare_attention_mask_for_generation( inputs_tensor, - generation_config.pad_token_id, - generation_config.eos_token_id, + pad_token_tensor, + eos_token_tensor, ) # decoder-only models should use left-padding for generation