aljaz/DALLE2-pytorch
1
1
Fork 0
mirror of https://github.com/lucidrains/DALLE2-pytorch.git synced 2025-12-19 09:44:19 +01:00
Code Issues Packages Projects Releases Wiki Activity

make sure openai clip adapter outputs l2normed embeddings

Browse Source
This commit is contained in:
Phil Wang
2022-05-06 10:12:03 -07:00
parent 14e63a3f67
commit 28e944f328
2 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
dalle2_pytorch/dalle2_pytorch.py
View File

@@ -264,7 +264,7 @@ class OpenAIClipAdapter(BaseClipAdapter):
text_embed = self.clip.encode_text(text) text_embed = self.clip.encode_text(text)
text_encodings = self.text_encodings text_encodings = self.text_encodings
del self.text_encodings del self.text_encodings
return EmbeddedText(text_embed.float(), text_encodings.float(), text_mask) return EmbeddedText(l2norm(text_embed.float()), text_encodings.float(), text_mask)
@torch.no_grad() @torch.no_grad()
def embed_image(self, image): def embed_image(self, image):
@@ -272,7 +272,7 @@ class OpenAIClipAdapter(BaseClipAdapter):
image = resize_image_to(image, self.image_size) image = resize_image_to(image, self.image_size)
image = self.clip_normalize(unnormalize_img(image)) image = self.clip_normalize(unnormalize_img(image))
image_embed = self.clip.encode_image(image) image_embed = self.clip.encode_image(image)
return EmbeddedImage(image_embed.float(), None) return EmbeddedImage(l2norm(image_embed.float()), None)
# classifier free guidance functions # classifier free guidance functions

2
setup.py
View File

@@ -10,7 +10,7 @@ setup(
'dream = dalle2_pytorch.cli:dream' 'dream = dalle2_pytorch.cli:dream'
], ],
}, },
version = '0.1.2', version = '0.1.4',
license='MIT', license='MIT',
description = 'DALL-E 2', description = 'DALL-E 2',
author = 'Phil Wang', author = 'Phil Wang',