The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.
Ace step prompts match now. (#12376)
comfyanonymousa4be04c5d750
diff --git a/comfy/text_encoders/ace15.py b/comfy/text_encoders/ace15.pyindex 5dac644c2..73697b3c1 100644--- a/comfy/text_encoders/ace15.py+++ b/comfy/text_encoders/ace15.py@@ -145,7 +145,7 @@ class ACE15Tokenizer(sd1_clip.SD1Tokenizer):def _metas_to_cot(self, *, return_yaml: bool = False, **kwargs) -> str:user_metas = {k: kwargs.pop(k)- for k in ("bpm", "duration", "keyscale", "timesignature", "language")+ for k in ("bpm", "duration", "keyscale", "timesignature")if k in kwargs}timesignature = user_metas.get("timesignature")@@ -208,8 +208,8 @@ class ACE15Tokenizer(sd1_clip.SD1Tokenizer):if not kwargs.get("use_negative_caption"):_ = metas_negative.pop("caption", None)- cot_text = self._metas_to_cot(caption = text, **kwargs)- cot_text_negative = "<think>\n</think>" if not metas_negative else self._metas_to_cot(**metas_negative)+ cot_text = self._metas_to_cot(caption=text, **kwargs)+ cot_text_negative = "<think>\n\n</think>" if not metas_negative else self._metas_to_cot(**metas_negative)meta_cap = self._metas_to_cap(**kwargs)lm_template = "<|im_start|>system\n# Instruction\nGenerate audio semantic tokens based on the given conditions:\n\n<|im_end|>\n<|im_start|>user\n# Caption\n{}\n\n# Lyric\n{}\n<|im_end|>\n<|im_start|>assistant\n{}\n\n<|im_end|>\n"