fix everything and make sure it runs end to end, document everything in readme for public

README.md

@@ -212,7 +212,8 @@ Let's see the whole script below
 
 ```python
 import torch
-from dalle2_pytorch import DALLE2, DiffusionPriorNetwork, DiffusionPrior, Unet, Decoder, CLIP
+from dalle2_pytorch.dalle2_pytorch import DALLE2
+from dalle2_pytorch import DiffusionPriorNetwork, DiffusionPrior, Unet, Decoder, CLIP
 
 import torch
 

dalle2_pytorch/dalle2_pytorch.py

@@ -374,13 +374,12 @@ class DiffusionPrior(nn.Module):
         image_encoding = self.clip.visual_transformer(image)
         image_cls = image_encoding[:, 0]
         image_embed = self.clip.to_visual_latent(image_cls)
-        return l2norm(image_embed)
+        return image_embed
 
     def get_text_cond(self, text):
         text_encodings = self.clip.text_transformer(text)
         text_cls, text_encodings = text_encodings[:, 0], text_encodings[:, 1:]
         text_embed = self.clip.to_text_latent(text_cls)
-        text_embed = l2norm(text_embed)
         return dict(text_encodings = text_encodings, text_embed = text_embed, mask = text != 0)
 
     def q_mean_variance(self, x_start, t):
@@ -751,7 +750,7 @@ class Decoder(nn.Module):
         image_encoding = self.clip.visual_transformer(image)
         image_cls = image_encoding[:, 0]
         image_embed = self.clip.to_visual_latent(image_cls)
-        return l2norm(image_embed)
+        return image_embed
 
     def q_mean_variance(self, x_start, t):
         mean = extract(self.sqrt_alphas_cumprod, t, x_start.shape) * x_start

setup.py

@@ -10,7 +10,7 @@ setup(
       'dream = dalle2_pytorch.cli:dream'
     ],
   },
-  version = '0.0.5',
+  version = '0.0.4',
   license='MIT',
   description = 'DALL-E 2',
   author = 'Phil Wang',