From 25c91efbb669170992a4242a480e38ce8e2ee97b Mon Sep 17 00:00:00 2001 From: Ryan Dick Date: Thu, 22 Aug 2024 16:18:33 +0000 Subject: [PATCH] Rename field positive_prompt -> prompt. --- invokeai/app/invocations/flux_text_encoder.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/invokeai/app/invocations/flux_text_encoder.py b/invokeai/app/invocations/flux_text_encoder.py index d3585c66a6..ae3456be06 100644 --- a/invokeai/app/invocations/flux_text_encoder.py +++ b/invokeai/app/invocations/flux_text_encoder.py @@ -36,7 +36,7 @@ class FluxTextEncoderInvocation(BaseInvocation): t5_max_seq_len: Literal[256, 512] = InputField( description="Max sequence length for the T5 encoder. Expected to be 256 for FLUX schnell models and 512 for FLUX dev models." ) - positive_prompt: str = InputField(description="Positive prompt for text-to-image generation.") + prompt: str = InputField(description="Text prompt to encode.") # TODO(ryand): Should we create a new return type for this invocation? This ConditioningOutput is clearly not # compatible with other ConditioningOutputs. @@ -73,7 +73,7 @@ class FluxTextEncoderInvocation(BaseInvocation): clip_encoder = HFEncoder(clip_text_encoder, clip_tokenizer, True, 77) t5_encoder = HFEncoder(t5_text_encoder, t5_tokenizer, False, self.t5_max_seq_len) - prompt = [self.positive_prompt] + prompt = [self.prompt] prompt_embeds = t5_encoder(prompt) pooled_prompt_embeds = clip_encoder(prompt)