use t5 relative positional bias in prior network causal transformer, since it makes more sense than rotary embeddings

README.md

@@ -22,9 +22,283 @@ For all of you emailing me (there is a lot), the best way to contribute is throu
 $ pip install dalle2-pytorch
 ```
 
-## Usage (work in progress)
+## Usage
 
-<a href="https://github.com/lucidrains/big-sleep">template</a>
+To train DALLE-2 is a 3 step process, with the training of CLIP being the most important
+
+To train CLIP, you can either use <a href="https://github.com/lucidrains/x-clip">x-clip</a> package, or join the LAION discord, where a lot of replication efforts are already <a href="https://github.com/mlfoundations/open_clip">underway</a>.
+
+This repository will demonstrate integration with `x-clip` for starters
+
+```python
+import torch
+from dalle2_pytorch import CLIP
+
+clip = CLIP(
+    dim_text = 512,
+    dim_image = 512,
+    dim_latent = 512,
+    num_text_tokens = 49408,
+    text_enc_depth = 1,
+    text_seq_len = 256,
+    text_heads = 8,
+    visual_enc_depth = 1,
+    visual_image_size = 256,
+    visual_patch_size = 32,
+    visual_heads = 8,
+    use_all_token_embeds = True,            # whether to use fine-grained contrastive learning (FILIP)
+    decoupled_contrastive_learning = True,  # use decoupled contrastive learning (DCL) objective function, removing positive pairs from the denominator of the InfoNCE loss (CLOOB + DCL)
+    extra_latent_projection = True,         # whether to use separate projections for text-to-image vs image-to-text comparisons (CLOOB)
+    use_visual_ssl = True,                  # whether to do self supervised learning on iages
+    visual_ssl_type = 'simclr',             # can be either 'simclr' or 'simsiam', depending on using DeCLIP or SLIP
+    use_mlm = False,                        # use masked language learning (MLM) on text (DeCLIP)
+    text_ssl_loss_weight = 0.05,            # weight for text MLM loss
+    image_ssl_loss_weight = 0.05            # weight for image self-supervised learning loss
+).cuda()
+
+# mock data
+
+text = torch.randint(0, 49408, (4, 256)).cuda()
+images = torch.randn(4, 3, 256, 256).cuda()
+
+# train
+
+loss = clip(
+    text,
+    images,
+    return_loss = True              # needs to be set to True to return contrastive loss
+)
+
+loss.backward()
+
+# do the above with as many texts and images as possible in a loop
+```
+
+Then, you will need to train the decoder, which learns to generate images based on the image embedding coming from the trained CLIP above
+
+```python
+import torch
+from dalle2_pytorch import Unet, Decoder, CLIP
+
+# trained clip from step 1
+
+clip = CLIP(
+    dim_text = 512,
+    dim_image = 512,
+    dim_latent = 512,
+    num_text_tokens = 49408,
+    text_enc_depth = 1,
+    text_seq_len = 256,
+    text_heads = 8,
+    visual_enc_depth = 1,
+    visual_image_size = 256,
+    visual_patch_size = 32,
+    visual_heads = 8
+).cuda()
+
+# unet for the decoder
+
+unet = Unet(
+    dim = 128,
+    image_embed_dim = 512,
+    cond_dim = 128,
+    channels = 3,
+    dim_mults=(1, 2, 4, 8)
+).cuda()
+
+# decoder, which contains the unet and clip
+
+decoder = Decoder(
+    net = unet,
+    clip = clip,
+    timesteps = 100,
+    cond_drop_prob = 0.2
+).cuda()
+
+# mock images (get a lot of this)
+
+images = torch.randn(4, 3, 256, 256).cuda()
+
+# feed images into decoder
+
+loss = decoder(images)
+loss.backward()
+
+# do the above for many many many many steps
+# then it will learn to generate images based on the CLIP image embeddings
+```
+
+Finally, the main contribution of the paper. The repository offers the diffusion prior network. It takes the CLIP text embeddings and tries to generate the CLIP image embeddings. Again, you will need the trained CLIP from the first step
+
+```python
+import torch
+from dalle2_pytorch import DiffusionPriorNetwork, DiffusionPrior, CLIP
+
+# get trained CLIP from step one
+
+clip = CLIP(
+    dim_text = 512,
+    dim_image = 512,
+    dim_latent = 512,
+    num_text_tokens = 49408,
+    text_enc_depth = 6,
+    text_seq_len = 256,
+    text_heads = 8,
+    visual_enc_depth = 6,
+    visual_image_size = 256,
+    visual_patch_size = 32,
+    visual_heads = 8,
+).cuda()
+
+# setup prior network, which contains an autoregressive transformer
+
+prior_network = DiffusionPriorNetwork(
+    dim = 512,
+    depth = 6,
+    dim_head = 64,
+    heads = 8
+).cuda()
+
+# diffusion prior network, which contains the CLIP and network (with transformer) above
+
+diffusion_prior = DiffusionPrior(
+    net = prior_network,
+    clip = clip,
+    timesteps = 100,
+    cond_drop_prob = 0.2
+).cuda()
+
+# mock data
+
+text = torch.randint(0, 49408, (4, 256)).cuda()
+images = torch.randn(4, 3, 256, 256).cuda()
+
+# feed text and images into diffusion prior network
+
+loss = diffusion_prior(text, images)
+loss.backward()
+
+# do the above for many many many steps
+# now the diffusion prior can generate image embeddings from the text embeddings
+```
+
+Finally, to generate the DALL-E2 images from text. Insert the trained `DiffusionPrior` as well as the `Decoder` (which both contains `CLIP`, a unet, and a causal transformer)
+
+```python
+from dalle2_pytorch import DALLE2
+
+dalle2 = DALLE2(
+    prior = diffusion_prior,
+    decoder = decoder
+)
+
+# send the text as a string if you want to use the simple tokenizer from DALLE v1
+# or you can do it as token ids, if you have your own tokenizer
+
+texts = ['glistening morning dew on a flower petal']
+images = dalle2(texts) # (1, 3, 256, 256)
+```
+
+That's it!
+
+Let's see the whole script below
+
+```python
+import torch
+from dalle2_pytorch import DALLE2, DiffusionPriorNetwork, DiffusionPrior, Unet, Decoder, CLIP
+
+clip = CLIP(
+    dim_text = 512,
+    dim_image = 512,
+    dim_latent = 512,
+    num_text_tokens = 49408,
+    text_enc_depth = 6,
+    text_seq_len = 256,
+    text_heads = 8,
+    visual_enc_depth = 6,
+    visual_image_size = 256,
+    visual_patch_size = 32,
+    visual_heads = 8
+).cuda()
+
+# mock data
+
+text = torch.randint(0, 49408, (4, 256)).cuda()
+images = torch.randn(4, 3, 256, 256).cuda()
+
+# train
+
+loss = clip(
+    text,
+    images,
+    return_loss = True
+)
+
+loss.backward()
+
+# do above for many steps ...
+
+# prior networks (with transformer)
+
+prior_network = DiffusionPriorNetwork(
+    dim = 512,
+    depth = 6,
+    dim_head = 64,
+    heads = 8
+).cuda()
+
+diffusion_prior = DiffusionPrior(
+    net = prior_network,
+    clip = clip,
+    timesteps = 100,
+    cond_drop_prob = 0.2
+).cuda()
+
+loss = diffusion_prior(text, images)
+loss.backward()
+
+# do above for many steps ...
+
+# decoder (with unet)
+
+unet = Unet(
+    dim = 128,
+    image_embed_dim = 512,
+    cond_dim = 128,
+    channels = 3,
+    dim_mults=(1, 2, 4, 8)
+).cuda()
+
+decoder = Decoder(
+    net = unet,
+    clip = clip,
+    timesteps = 100,
+    cond_drop_prob = 0.2
+).cuda()
+
+loss = decoder(images) # this can optionally be decoder(images, text) if you wish to condition on the text encodings as well, though it was hinted in the paper it didn't do much
+loss.backward()
+
+# do above for many steps
+
+dalle2 = DALLE2(
+    prior = diffusion_prior,
+    decoder = decoder
+)
+
+images = dalle2(
+    ['cute puppy chasing after a squirrel'],
+    cond_scale = 2. # classifier free guidance strength (> 1 would strengthen the condition)
+)
+
+# save your image
+```
+
+Everything in this readme should run without error
+
+For the layperson, no worries, training will all be automated into a CLI tool, at least for small scale training.
+
+## CLI Usage (work in progress)
 
 ```bash
 $ dream 'sharing a sunset at the summit of mount everest with my dog'
@@ -32,22 +306,25 @@ $ dream 'sharing a sunset at the summit of mount everest with my dog'
 
 Once built, images will be saved to the same directory the command is invoked
 
-## Training (work in progress, will offer both in code and as command-line)
+## Training wrapper (wip)
+
+Offer training wrappers
+
+## Training CLI (wip)
 
 <a href="https://github.com/lucidrains/stylegan2-pytorch">template</a>
 
-Todo
-
 ## Todo
 
 - [x] finish off gaussian diffusion class for latent embedding - allow for prediction of epsilon
 - [x] add what was proposed in the paper, where DDPM objective for image latent embedding predicts x0 directly (reread vq-diffusion paper and get caught up on that line of work)
-- [ ] make sure it works end to end to produce an output tensor, taking a single gradient step
-- [ ] augment unet so that it can also be conditioned on text encodings (although in paper they hinted this didn't make much a difference)
+- [x] make sure it works end to end to produce an output tensor, taking a single gradient step
+- [x] augment unet so that it can also be conditioned on text encodings (although in paper they hinted this didn't make much a difference)
 - [ ] look into Jonathan Ho's cascading DDPM for the decoder, as that seems to be what they are using. get caught up on DDPM literature
 - [ ] figure out all the current bag of tricks needed to make DDPMs great (starting with the blur trick mentioned in paper)
 - [ ] train on a toy task, offer in colab
 - [ ] add attention to unet - apply some personal tricks with efficient attention
+- [ ] figure out the big idea behind latent diffusion and what can be ported over
 
 ## Citations
 
@@ -95,3 +372,5 @@ Todo
     primaryClass = {cs.LG}
 }
 ```
+
+*Creating noise from data is easy; creating data from noise is generative modeling.* - Yang Song's <a href="https://arxiv.org/abs/2011.13456">paper</a>

dalle2_pytorch/__init__.py

@@ -1 +1,2 @@
-from dalle2_pytorch.dalle2_pytorch import DALLE2
+from dalle2_pytorch.dalle2_pytorch import DALLE2, DiffusionPriorNetwork, DiffusionPrior, Unet, Decoder
+from x_clip import CLIP

dalle2_pytorch/dalle2_pytorch.py

@@ -1,10 +1,19 @@
-import tqdm
+import math
+from tqdm import tqdm
+from inspect import isfunction
+
 import torch
 import torch.nn.functional as F
 from torch import nn, einsum
 
-from einops import rearrange
-from einops_exts import rearrange_many, repeat_many
+from einops import rearrange, repeat
+from einops.layers.torch import Rearrange
+from einops_exts import rearrange_many, repeat_many, check_shape
+from einops_exts.torch import EinopsToAndFrom
+
+from kornia.filters import filter2d
+
+from dalle2_pytorch.tokenizer import tokenizer
 
 # use x-clip
 
@@ -16,7 +25,9 @@ def exists(val):
     return val is not None
 
 def default(val, d):
-    return val if exists(val) else d
+    if exists(val):
+        return val
+    return d() if isfunction(d) else d
 
 def eval_decorator(fn):
     def inner(model, *args, **kwargs):
@@ -27,6 +38,11 @@ def eval_decorator(fn):
         return out
     return inner
 
+def is_list_str(x):
+    if not isinstance(x, (list, tuple)):
+        return False
+    return all([type(el) == str for el in x])
+
 # for controlling freezing of CLIP
 
 def set_module_requires_grad_(module, requires_grad):
@@ -43,6 +59,11 @@ def freeze_model_and_make_eval_(model):
     model.eval()
     freeze_all_layers_(model)
 
+# tensor helpers
+
+def l2norm(t):
+    return F.normalize(t, dim = -1)
+
 # classifier free guidance functions
 
 def prob_mask_like(shape, prob, device):
@@ -91,29 +112,121 @@ class RMSNorm(nn.Module):
         inv_norm = torch.rsqrt(squared_sum + self.eps)
         return x * inv_norm * self.gamma * self.scale
 
-class PreNormResidual(nn.Module):
-    def __init__(self, dim, fn):
+class ChanRMSNorm(RMSNorm):
+    def forward(self, x):
+        squared_sum = (x ** 2).sum(dim = 1, keepdim = True)
+        inv_norm = torch.rsqrt(squared_sum + self.eps)
+        return x * inv_norm * rearrange(self.gamma, 'c -> 1 c 1 1') * self.scale
+
+class Residual(nn.Module):
+    def __init__(self, fn):
         super().__init__()
-        self.norm = RMSNorm(dim)
+        self.fn = fn
 
     def forward(self, x, **kwargs):
-        return self.fn(self.norm(x), **kwargs) + x
+        return self.fn(x, **kwargs) + x
+
+# mlp
+
+class MLP(nn.Module):
+    def __init__(
+        self,
+        dim_in,
+        dim_out,
+        *,
+        expansion_factor = 2.,
+        depth = 2,
+        norm = False,
+    ):
+        super().__init__()
+        hidden_dim = int(expansion_factor * dim_out)
+        norm_fn = lambda: nn.LayerNorm(hidden_dim) if norm else nn.Identity()
+
+        layers = [nn.Sequential(
+            nn.Linear(dim_in, hidden_dim),
+            nn.SiLU(),
+            norm_fn()
+        )]
+
+        for _ in range(depth - 1):
+            layers.append(nn.Sequential(
+                nn.Linear(hidden_dim, hidden_dim),
+                nn.SiLU(),
+                norm_fn()
+            ))
+
+        layers.append(nn.Linear(hidden_dim, dim_out))
+        self.net = nn.Sequential(*layers)
+
+    def forward(self, x):
+        return self.net(x.float())
+
+# relative positional bias for causal transformer
+
+class RelPosBias(nn.Module):
+    def __init__(
+        self,
+        heads = 8,
+        num_buckets = 32,
+        max_distance = 128,
+    ):
+        super().__init__()
+        self.num_buckets = num_buckets
+        self.max_distance = max_distance
+        self.relative_attention_bias = nn.Embedding(num_buckets, heads)
+
+    @staticmethod
+    def _relative_position_bucket(
+        relative_position,
+        num_buckets = 32,
+        max_distance = 128
+    ):
+        n = -relative_position
+        n = torch.max(n, torch.zeros_like(n))
+
+        max_exact = num_buckets // 2
+        is_small = n < max_exact
+
+        val_if_large = max_exact + (torch.log(n.float() / max_exact) / math.log(max_distance / max_exact) * (num_buckets - max_exact)).long()
+        val_if_large = torch.min(val_if_large, torch.full_like(val_if_large, num_buckets - 1))
+        return torch.where(is_small, n, val_if_large)
+
+    def forward(self, i, j, *, device):
+        q_pos = torch.arange(i, dtype = torch.long, device = device)
+        k_pos = torch.arange(j, dtype = torch.long, device = device)
+        rel_pos = rearrange(k_pos, 'j -> 1 j') - rearrange(q_pos, 'i -> i 1')
+        rp_bucket = self._relative_position_bucket(rel_pos, num_buckets = self.num_buckets, max_distance = self.max_distance)
+        values = self.relative_attention_bias(rp_bucket)
+        return rearrange(values, 'i j h -> h i j')
+
+# feedforward
+
+class SwiGLU(nn.Module):
+    """ used successfully in https://arxiv.org/abs/2204.0231 """
+    def forward(self, x):
+        x, gate = x.chunk(2, dim = -1)
+        return x * F.silu(gate)
+
+def FeedForward(dim, mult = 4, dropout = 0., post_activation_norm = False):
+    """ post-activation norm https://arxiv.org/abs/2110.09456 """
 
-def FeedForward(dim, mult = 4, dropout = 0.):
     inner_dim = int(mult * dim)
     return nn.Sequential(
         RMSNorm(dim),
-        nn.Linear(dim, inner_dim, bias = False),
-        nn.GELU(),
+        nn.Linear(dim, inner_dim * 2, bias = False),
+        SwiGLU(),
+        RMSNorm(inner_dim) if post_activation_norm else nn.Identity(),
         nn.Dropout(dropout),
         nn.Linear(inner_dim, dim, bias = False)
     )
 
+# attention
+
 class Attention(nn.Module):
     def __init__(
         self,
-        *,
         dim,
+        *,
         dim_head = 64,
         heads = 8,
         dropout = 0.,
@@ -121,6 +234,7 @@ class Attention(nn.Module):
     ):
         super().__init__()
         self.scale = dim_head ** -0.5
+        self.heads = heads
         inner_dim = dim_head * heads
 
         self.causal = causal
@@ -128,17 +242,17 @@ class Attention(nn.Module):
         self.dropout = nn.Dropout(dropout)
 
         self.null_kv = nn.Parameter(torch.randn(2, dim_head))
-        self.to_qkv = nn.Linear(dim, inner_dim, bias = False)
+        self.to_q = nn.Linear(dim, inner_dim, bias = False)
         self.to_kv = nn.Linear(dim, dim_head * 2, bias = False)
         self.to_out = nn.Linear(inner_dim, dim, bias = False)
 
-    def forward(self, x, mask = None):
-        b, n, device = x.shape[:2], x.device
+    def forward(self, x, mask = None, attn_bias = None):
+        b, n, device = *x.shape[:2], x.device
 
         x = self.norm(x)
-        qkv = self.to_qkv(x).chunk(3, dim = -1)
+        q, k, v = (self.to_q(x), *self.to_kv(x).chunk(2, dim = -1))
 
-        q = rearrange(q, 'b n (h d) -> b h n d')
+        q = rearrange(q, 'b n (h d) -> b h n d', h = self.heads)
 
         # add null key / value for classifier free guidance in prior net
 
@@ -148,7 +262,15 @@ class Attention(nn.Module):
 
         q = q * self.scale
 
-        sim = einsum('b h i d, b j d -> b h i j')
+        sim = einsum('b h i d, b j d -> b h i j', q, k)
+
+        # relative positional encoding (T5 style)
+
+        if exists(attn_bias):
+            sim = sim + attn_bias
+
+        # masking
+
         max_neg_value = -torch.finfo(sim.dtype).max
 
         if exists(mask):
@@ -157,11 +279,17 @@ class Attention(nn.Module):
             sim = sim.masked_fill(~mask, max_neg_value)
 
         if self.causal:
-            causal_mask = torch.ones((n, n), dtype = torch.bool, device = device).triu(1)
+            i, j = sim.shape[-2:]
+            causal_mask = torch.ones((i, j), dtype = torch.bool, device = device).triu(j - i + 1)
             sim = sim.masked_fill(causal_mask, max_neg_value)
 
+        # attention
+
         sim = sim - sim.amax(dim = -1, keepdim = True)
         attn = sim.softmax(dim = -1)
+        attn = self.dropout(attn)
+
+        # aggregate values
 
         out = einsum('b h i j, b j d -> b h i d', attn, v)
 
@@ -182,7 +310,7 @@ class CausalTransformer(nn.Module):
         ff_dropout = 0.
     ):
         super().__init__()
-        # todo - bring in rotary embeddings or alibi
+        self.rel_pos_bias = RelPosBias(heads = heads)
 
         self.layers = nn.ModuleList([])
         for _ in range(depth):
@@ -198,8 +326,12 @@ class CausalTransformer(nn.Module):
         x,
         mask = None    # we will need a mask here, due to variable length of the text encodings - also offer dalle1 strategy with padding token embeddings
     ):
+        n, device = x.shape[1], x.device
+
+        attn_bias = self.rel_pos_bias(n, n + 1, device = device)
+
         for attn, ff in self.layers:
-            x = attn(x, mask = mask) + x
+            x = attn(x, mask = mask, attn_bias = attn_bias) + x
             x = ff(x) + x
 
         return self.norm(x)
@@ -208,26 +340,26 @@ class DiffusionPriorNetwork(nn.Module):
     def __init__(
         self,
         dim,
-        num_timesteps = 1000,
+        num_timesteps = None,
         **kwargs
     ):
         super().__init__()
-        self.time_embeddings = nn.Embedding(num_timesteps, dim)  # also offer a continuous version of timestep embeddings, with a 2 layer MLP
+        self.time_embeddings = nn.Embedding(num_timesteps, dim) if exists(num_timesteps) else nn.Sequential(Rearrange('b -> b 1'), MLP(1, dim)) # also offer a continuous version of timestep embeddings, with a 2 layer MLP
         self.learned_query = nn.Parameter(torch.randn(dim))
-        self.causal_transformer = CausalTransformer(**kwargs)
+        self.causal_transformer = CausalTransformer(dim = dim, **kwargs)
 
     def forward_with_cond_scale(
         self,
-        x,
-        *,
+        *args,
         cond_scale = 1.,
         **kwargs
     ):
-        if cond_scale == 1:
-            return self.forward(x, **kwargs)
+        logits = self.forward(*args, **kwargs)
 
-        logits = self.forward(x, **kwargs)
-        null_logits = self.forward(x, cond_prob_drop = 1., **kwargs)
+        if cond_scale == 1:
+            return logits
+
+        null_logits = self.forward(*args, cond_drop_prob = 1., **kwargs)
         return null_logits + (logits - null_logits) * cond_scale
 
     def forward(
@@ -247,10 +379,18 @@ class DiffusionPriorNetwork(nn.Module):
 
         text_embed, image_embed = rearrange_many((text_embed, image_embed), 'b d -> b 1 d')
 
+        # whether text embedding is used for conditioning depends on whether text encodings are available for attention (for classifier free guidance, even though it seems from the paper it was not used in the prior ddpm, as the objective is different)
+        # but let's just do it right
+
         if exists(mask):
-            mask = F.pad(mask, (0, 4), value = True) # extend mask for text embedding, noised image embedding, time step embedding, and learned query
+            not_all_masked_out = mask.any(dim = -1)
+            mask = torch.cat((mask, rearrange(not_all_masked_out, 'b -> b 1')), dim = 1)
+
+        if exists(mask):
+            mask = F.pad(mask, (0, 2), value = True) # extend mask for text embedding, noised image embedding, time step embedding, and learned query
 
         time_embed = self.time_embeddings(diffusion_timesteps)
+        time_embed = rearrange(time_embed, 'b d -> b 1 d')
 
         learned_queries = repeat(self.learned_query, 'd -> b 1 d', b = batch)
 
@@ -268,7 +408,7 @@ class DiffusionPriorNetwork(nn.Module):
 
         # classifier free guidance
 
-        cond_prob_mask = prob_mask_like((batch_size,), cond_prob_drop, device = device)
+        cond_prob_mask = prob_mask_like((batch,), cond_drop_prob, device = device)
         mask &= rearrange(cond_prob_mask, 'b -> b 1')
 
         # attend
@@ -288,19 +428,20 @@ class DiffusionPrior(nn.Module):
         *,
         clip,
         timesteps = 1000,
-        cond_prob_drop = 0.2,
+        cond_drop_prob = 0.2,
         loss_type = 'l1',
         predict_x0 = True
     ):
         super().__init__()
         assert isinstance(clip, CLIP)
         freeze_model_and_make_eval_(clip)
+        self.clip = clip
 
         self.net = net
         self.image_embed_dim = clip.dim_latent
         self.channels = clip.image_channels
         self.image_size = clip.image_size
-        self.cond_prob_drop = cond_prob_drop
+        self.cond_drop_prob = cond_drop_prob
 
         self.predict_x0 = predict_x0
         # in paper, they do not predict the noise, but predict x0 directly for image embedding, claiming empirically better results. I'll just offer both.
@@ -345,12 +486,13 @@ 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 image_embed
+        return l2norm(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):
@@ -389,7 +531,7 @@ class DiffusionPrior(nn.Module):
         return model_mean, posterior_variance, posterior_log_variance
 
     @torch.no_grad()
-    def p_sample(self, x, t, image_embed, text_cond = None, clip_denoised = True, repeat_noise = False):
+    def p_sample(self, x, t, text_cond = None, clip_denoised = True, repeat_noise = False):
         b, *_, device = *x.shape, x.device
         model_mean, _, model_log_variance = self.p_mean_variance(x = x, t = t, text_cond = text_cond, clip_denoised = clip_denoised)
         noise = noise_like(x.shape, device, repeat_noise)
@@ -420,18 +562,18 @@ class DiffusionPrior(nn.Module):
         text_cond = self.get_text_cond(text)
 
         image_embeds = self.p_sample_loop((batch_size, image_embed_dim), text_cond = text_cond)
-        text_embeds = text_cond['text_embeds']
+        text_embeds = text_cond['text_embed']
 
         text_embeds = rearrange(text_embeds, '(b r) d -> b r d', r = num_samples_per_batch)
         image_embeds = rearrange(image_embeds, '(b r) d -> b r d', r = num_samples_per_batch)
 
-        text_image_sims = einsum('b r d, b r d -> b r')
+        text_image_sims = einsum('b r d, b r d -> b r', l2norm(text_embeds), l2norm(image_embeds))
         top_sim_indices = text_image_sims.topk(k = 1).indices
 
-        top_sim_indices = repeat(top_sim_indices, 'b 1 -> b d', d = image_embed_dim)
+        top_sim_indices = repeat(top_sim_indices, 'b 1 -> b 1 d', d = image_embed_dim)
 
         top_image_embeds = image_embeds.gather(1, top_sim_indices)
-        return top_image_embeds
+        return rearrange(top_image_embeds, 'b 1 d -> b d')
 
     def q_sample(self, x_start, t, noise=None):
         noise = default(noise, lambda: torch.randn_like(x_start))
@@ -442,14 +584,14 @@ class DiffusionPrior(nn.Module):
         )
 
     def p_losses(self, image_embed, t, text_cond, noise = None):
-        noise = default(noise, lambda: torch.randn_like(x_start))
+        noise = default(noise, lambda: torch.randn_like(image_embed))
 
         image_embed_noisy = self.q_sample(x_start = image_embed, t = t, noise = noise)
 
         x_recon = self.net(
             image_embed_noisy,
             t,
-            cond_prob_drop = self.cond_prob_drop,
+            cond_drop_prob = self.cond_drop_prob,
             **text_cond
         )
 
@@ -472,7 +614,7 @@ class DiffusionPrior(nn.Module):
         image_embed = self.get_image_embed(image)
         text_cond = self.get_text_cond(text)
 
-        loss = self.p_losses(x, times, image_embed = image_embed, text_cond = text_cond, *args, **kwargs)
+        loss = self.p_losses(image_embed, times, text_cond = text_cond, *args, **kwargs)
         return loss
 
 # decoder
@@ -483,6 +625,17 @@ def Upsample(dim):
 def Downsample(dim):
     return nn.Conv2d(dim, dim, 4, 2, 1)
 
+class Blur(nn.Module):
+    def __init__(self):
+        super().__init__()
+        filt = torch.Tensor([1, 2, 1])
+        self.register_buffer('filt', filt)
+
+    def forward(self, x):
+        filt = self.filt
+        filt = rearrange(filt, '... j -> ... 1 j') * rearrange(flit, '... i -> ... i 1')
+        return filter2d(x, filt, normalized = True)
+
 class SinusoidalPosEmb(nn.Module):
     def __init__(self, dim):
         super().__init__()
@@ -510,16 +663,23 @@ class ConvNextBlock(nn.Module):
         super().__init__()
         need_projection = dim != dim_out
 
-        self.mlp = nn.Sequential(
-            nn.GELU(),
-            nn.Linear(cond_dim, dim)
-        ) if exists(cond_dim) else None
+        self.cross_attn = None
+
+        if exists(cond_dim):
+            self.cross_attn = EinopsToAndFrom(
+                'b c h w',
+                'b (h w) c',
+                CrossAttention(
+                    dim = dim,
+                    context_dim = cond_dim
+                )
+            )
 
         self.ds_conv = nn.Conv2d(dim, dim, 7, padding = 3, groups = dim)
 
         inner_dim = int(dim_out * mult)
         self.net = nn.Sequential(
-            RMSNorm(dim) if norm else nn.Identity(),
+            ChanRMSNorm(dim) if norm else nn.Identity(),
             nn.Conv2d(dim, inner_dim, 3, padding = 1),
             nn.GELU(),
             nn.Conv2d(inner_dim, dim_out, 3, padding = 1)
@@ -530,28 +690,73 @@ class ConvNextBlock(nn.Module):
     def forward(self, x, cond = None):
         h = self.ds_conv(x)
 
-        if exists(self.mlp):
+        if exists(self.cross_attn):
             assert exists(cond)
-            condition = self.mlp(cond)
-            h = h + rearrange(condition, 'b c -> b c 1 1')
+            h = self.cross_attn(h, context = cond) + h
 
         h = self.net(h)
+
         return h + self.res_conv(x)
 
-class EinopsToAndFrom(nn.Module):
-    def __init__(self, from_einops, to_einops, fn):
+class CrossAttention(nn.Module):
+    def __init__(
+        self,
+        dim,
+        *,
+        context_dim = None,
+        dim_head = 64,
+        heads = 8,
+        dropout = 0.,
+    ):
         super().__init__()
-        self.from_einops = from_einops
-        self.to_einops = to_einops
-        self.fn = fn
+        self.scale = dim_head ** -0.5
+        self.heads = heads
+        inner_dim = dim_head * heads
 
-    def forward(self, x, **kwargs):
-        shape = x.shape
-        reconstitute_kwargs = dict(tuple(zip(self.from_einops.split(' '), shape)))
-        x = rearrange(x, f'{self.from_einops} -> {self.to_einops}')
-        x = self.fn(x, **kwargs)
-        x = rearrange(x, f'{self.to_einops} -> {self.from_einops}', **reconstitute_kwargs)
-        return x
+        context_dim = default(context_dim, dim)
+
+        self.norm = RMSNorm(dim)
+        self.norm_context = RMSNorm(context_dim)
+        self.dropout = nn.Dropout(dropout)
+
+        self.null_kv = nn.Parameter(torch.randn(2, dim_head))
+        self.to_q = nn.Linear(dim, inner_dim, bias = False)
+        self.to_kv = nn.Linear(context_dim, inner_dim * 2, bias = False)
+        self.to_out = nn.Linear(inner_dim, dim, bias = False)
+
+    def forward(self, x, context, mask = None):
+        b, n, device = *x.shape[:2], x.device
+
+        x = self.norm(x)
+        context = self.norm_context(context)
+
+        q, k, v = (self.to_q(x), *self.to_kv(context).chunk(2, dim = -1))
+
+        q, k, v = rearrange_many((q, k, v), 'b n (h d) -> b h n d', h = self.heads)
+
+        # add null key / value for classifier free guidance in prior net
+
+        nk, nv = repeat_many(self.null_kv.unbind(dim = -2), 'd -> b h 1 d', h = self.heads,  b = b)
+
+        k = torch.cat((nk, k), dim = -2)
+        v = torch.cat((nv, v), dim = -2)
+
+        q = q * self.scale
+
+        sim = einsum('b h i d, b h j d -> b h i j', q, k)
+        max_neg_value = -torch.finfo(sim.dtype).max
+
+        if exists(mask):
+            mask = F.pad(mask, (1, 0), value = True)
+            mask = rearrange(mask, 'b j -> b 1 1 j')
+            sim = sim.masked_fill(~mask, max_neg_value)
+
+        sim = sim - sim.amax(dim = -1, keepdim = True)
+        attn = sim.softmax(dim = -1)
+
+        out = einsum('b h i j, b h j d -> b h i d', attn, v)
+        out = rearrange(out, 'b h n d -> b n (h d)')
+        return self.to_out(out)
 
 class Unet(nn.Module):
     def __init__(
@@ -559,7 +764,8 @@ class Unet(nn.Module):
         dim,
         *,
         image_embed_dim,
-        time_dim = None,
+        cond_dim = None,
+        num_image_tokens = 4,
         out_dim = None,
         dim_mults=(1, 2, 4, 8),
         channels = 3,
@@ -570,18 +776,31 @@ class Unet(nn.Module):
         dims = [channels, *map(lambda m: dim * m, dim_mults)]
         in_out = list(zip(dims[:-1], dims[1:]))
 
-        time_dim = default(time_dim, dim)
+        # time, image embeddings, and optional text encoding
+
+        cond_dim = default(cond_dim, dim)
 
         self.time_mlp = nn.Sequential(
             SinusoidalPosEmb(dim),
             nn.Linear(dim, dim * 4),
             nn.GELU(),
-            nn.Linear(dim * 4, dim)
+            nn.Linear(dim * 4, cond_dim),
+            Rearrange('b d -> b 1 d')
         )
 
-        self.null_image_embed = nn.Parameter(torch.randn(image_embed_dim))
+        self.image_to_cond = nn.Sequential(
+            nn.Linear(image_embed_dim, cond_dim * num_image_tokens),
+            Rearrange('b (n d) -> b n d', n = num_image_tokens)
+        ) if image_embed_dim != cond_dim else nn.Identity()
 
-        cond_dim = time_dim + image_embed_dim
+        self.text_to_cond = nn.LazyLinear(cond_dim)
+
+        # for classifier free guidance
+
+        self.null_image_embed = nn.Parameter(torch.randn(1, num_image_tokens, cond_dim))
+        self.null_text_embed = nn.Parameter(torch.randn(1, 1, cond_dim))
+
+        # layers
 
         self.downs = nn.ModuleList([])
         self.ups = nn.ModuleList([])
@@ -591,14 +810,15 @@ class Unet(nn.Module):
             is_last = ind >= (num_resolutions - 1)
 
             self.downs.append(nn.ModuleList([
-                ConvNextBlock(dim_in, dim_out, cond_dim = cond_dim, norm = ind != 0),
+                ConvNextBlock(dim_in, dim_out, norm = ind != 0),
                 ConvNextBlock(dim_out, dim_out, cond_dim = cond_dim),
                 Downsample(dim_out) if not is_last else nn.Identity()
             ]))
 
         mid_dim = dims[-1]
+
         self.mid_block1 = ConvNextBlock(mid_dim, mid_dim, cond_dim = cond_dim)
-        self.mid_attn = EinopsToAndFrom('b c h w', 'b (h w) c', PreNormResidual(mid_dim, Attention(mid_dim)))
+        self.mid_attn = EinopsToAndFrom('b c h w', 'b (h w) c', Residual(Attention(mid_dim)))
         self.mid_block2 = ConvNextBlock(mid_dim, mid_dim, cond_dim = cond_dim)
 
         for ind, (dim_in, dim_out) in enumerate(reversed(in_out[1:])):
@@ -618,16 +838,16 @@ class Unet(nn.Module):
 
     def forward_with_cond_scale(
         self,
-        x,
-        *,
+        *args,
         cond_scale = 1.,
         **kwargs
     ):
-        if cond_scale == 1:
-            return self.forward(x, **kwargs)
+        logits = self.forward(*args, **kwargs)
 
-        logits = self.forward(x, **kwargs)
-        null_logits = self.forward(x, cond_prob_drop = 1., **kwargs)
+        if cond_scale == 1:
+            return logits
+
+        null_logits = self.forward(*args, cond_drop_prob = 1., **kwargs)
         return null_logits + (logits - null_logits) * cond_scale
 
     def forward(
@@ -637,39 +857,62 @@ class Unet(nn.Module):
         *,
         image_embed,
         text_encodings = None,
-        cond_prob_drop = 0.
+        cond_drop_prob = 0.
     ):
-        t = self.time_mlp(time)
+        batch_size, device = x.shape[0], x.device
+        time_tokens = self.time_mlp(time)
 
-        cond_prob_mask = prob_mask_like((batch_size,), cond_prob_drop, device = device)
+        cond_prob_mask = prob_mask_like((batch_size,), cond_drop_prob, device = device)
+        cond_prob_mask = rearrange(cond_prob_mask, 'b -> b 1 1')
 
         # mask out image embedding depending on condition dropout
         # for classifier free guidance
 
-        image_embed = torch.where(
-            rearrange(cond_prob_mask, 'b -> b 1'),
-            image_embed,
-            rearrange(self.null_image_embed, 'd -> 1 d')
+        image_tokens = self.image_to_cond(image_embed)
+
+        image_tokens = torch.where(
+            cond_prob_mask,
+            image_tokens,
+            self.null_image_embed
         )
 
-        cond = torch.cat((t, image_embed), dim = -1)
+        # take care of text encodings (optional)
+
+        if exists(text_encodings):
+            text_tokens = self.text_to_cond(text_encodings)
+            text_tokens = torch.where(
+                cond_prob_mask,
+                text_tokens,
+                self.null_text_embed
+            )
+
+        # main conditioning tokens (c)
+
+        c = torch.cat((time_tokens, image_tokens), dim = -2)
+
+        # text and image conditioning tokens (mid_c)
+        # to save on compute, only do cross attention based conditioning on the inner most layers of the Unet
+
+        mid_c = c if not exists(text_encodings) else torch.cat((c, text_tokens), dim = -2)
+
+        # go through the layers of the unet, down and up
 
         hiddens = []
 
         for convnext, convnext2, downsample in self.downs:
-            x = convnext(x, t)
-            x = convnext2(x, t)
+            x = convnext(x, c)
+            x = convnext2(x, c)
             hiddens.append(x)
             x = downsample(x)
 
-        x = self.mid_block1(x, t)
-        x = self.attn(x)
-        x = self.mid_block2(x, t)
+        x = self.mid_block1(x, mid_c)
+        x = self.mid_attn(x)
+        x = self.mid_block2(x, mid_c)
 
         for convnext, convnext2, upsample in self.ups:
             x = torch.cat((x, hiddens.pop()), dim=1)
-            x = convnext(x, t)
-            x = convnext2(x, t)
+            x = convnext(x, c)
+            x = convnext2(x, c)
             x = upsample(x)
 
         return self.final_conv(x)
@@ -681,17 +924,18 @@ class Decoder(nn.Module):
         *,
         clip,
         timesteps = 1000,
-        cond_prob_drop = 0.2,
+        cond_drop_prob = 0.2,
         loss_type = 'l1'
     ):
         super().__init__()
         assert isinstance(clip, CLIP)
         freeze_model_and_make_eval_(clip)
+        self.clip = clip
 
         self.net = net
         self.channels = clip.image_channels
         self.image_size = clip.image_size
-        self.cond_prob_drop = cond_prob_drop
+        self.cond_drop_prob = cond_drop_prob
 
         betas = cosine_beta_schedule(timesteps)
 
@@ -729,11 +973,15 @@ class Decoder(nn.Module):
         self.register_buffer('posterior_mean_coef1', betas * torch.sqrt(alphas_cumprod_prev) / (1. - alphas_cumprod))
         self.register_buffer('posterior_mean_coef2', (1. - alphas_cumprod_prev) * torch.sqrt(alphas) / (1. - alphas_cumprod))
 
+    def get_text_encodings(self, text):
+        text_encodings = self.clip.text_transformer(text)
+        return text_encodings[:, 1:]
+
     def get_image_embed(self, image):
         image_encoding = self.clip.visual_transformer(image)
         image_cls = image_encoding[:, 0]
         image_embed = self.clip.to_visual_latent(image_cls)
-        return image_embed
+        return l2norm(image_embed)
 
     def q_mean_variance(self, x_start, t):
         mean = extract(self.sqrt_alphas_cumprod, t, x_start.shape) * x_start
@@ -756,8 +1004,8 @@ class Decoder(nn.Module):
         posterior_log_variance_clipped = extract(self.posterior_log_variance_clipped, t, x_t.shape)
         return posterior_mean, posterior_variance, posterior_log_variance_clipped
 
-    def p_mean_variance(self, x, t, image_embed, clip_denoised: bool):
-        x_recon = self.predict_start_from_noise(x, t = t, noise = self.net(x, t, image_embed = image_embed))
+    def p_mean_variance(self, x, t, image_embed, text_encodings = None, clip_denoised = True, cond_scale = 1.):
+        x_recon = self.predict_start_from_noise(x, t = t, noise = self.net.forward_with_cond_scale(x, t, image_embed = image_embed, text_encodings = text_encodings, cond_scale = cond_scale))
 
         if clip_denoised:
             x_recon.clamp_(-1., 1.)
@@ -766,31 +1014,32 @@ class Decoder(nn.Module):
         return model_mean, posterior_variance, posterior_log_variance
 
     @torch.no_grad()
-    def p_sample(self, x, t, image_embed, clip_denoised = True, repeat_noise = False):
+    def p_sample(self, x, t, image_embed, text_encodings = None, cond_scale = 1., clip_denoised = True, repeat_noise = False):
         b, *_, device = *x.shape, x.device
-        model_mean, _, model_log_variance = self.p_mean_variance(x = x, t = t, clip_denoised = clip_denoised)
+        model_mean, _, model_log_variance = self.p_mean_variance(x = x, t = t, image_embed = image_embed, text_encodings = text_encodings, cond_scale = cond_scale, clip_denoised = clip_denoised)
         noise = noise_like(x.shape, device, repeat_noise)
         # no noise when t == 0
         nonzero_mask = (1 - (t == 0).float()).reshape(b, *((1,) * (len(x.shape) - 1)))
         return model_mean + nonzero_mask * (0.5 * model_log_variance).exp() * noise
 
     @torch.no_grad()
-    def p_sample_loop(self, shape, image_embed):
+    def p_sample_loop(self, shape, image_embed, text_encodings = None, cond_scale = 1):
         device = self.betas.device
 
         b = shape[0]
         img = torch.randn(shape, device=device)
 
         for i in tqdm(reversed(range(0, self.num_timesteps)), desc='sampling loop time step', total=self.num_timesteps):
-            img = self.p_sample(img, torch.full((b,), i, device = device, dtype = torch.long), image_embed = image_embed)
+            img = self.p_sample(img, torch.full((b,), i, device = device, dtype = torch.long), image_embed = image_embed, text_encodings = text_encodings, cond_scale = cond_scale)
         return img
 
     @torch.no_grad()
-    def sample(self, image_embed):
+    def sample(self, image_embed, text = None, cond_scale = 1.):
         batch_size = image_embed.shape[0]
         image_size = self.image_size
         channels = self.channels
-        return self.p_sample_loop((batch_size, channels, image_size, image_size), image_embed = image_embed)
+        text_encodings = self.get_text_encodings(text) if exists(text) else None
+        return self.p_sample_loop((batch_size, channels, image_size, image_size), image_embed = image_embed, text_encodings = text_encodings, cond_scale = cond_scale)
 
     def q_sample(self, x_start, t, noise=None):
         noise = default(noise, lambda: torch.randn_like(x_start))
@@ -800,7 +1049,7 @@ class Decoder(nn.Module):
             extract(self.sqrt_one_minus_alphas_cumprod, t, x_start.shape) * noise
         )
 
-    def p_losses(self, x_start, image_embed, t, noise = None):
+    def p_losses(self, x_start, t, *, image_embed, text_encodings = None, noise = None):
         noise = default(noise, lambda: torch.randn_like(x_start))
 
         x_noisy = self.q_sample(x_start = x_start, t = t, noise = noise)
@@ -809,7 +1058,8 @@ class Decoder(nn.Module):
             x_noisy,
             t,
             image_embed = image_embed,
-            cond_prob_drop = self.cond_prob_drop
+            text_encodings = text_encodings,
+            cond_drop_prob = self.cond_drop_prob
         )
 
         if self.loss_type == 'l1':
@@ -821,14 +1071,16 @@ class Decoder(nn.Module):
 
         return loss
 
-    def forward(self, image, *args, **kwargs):
+    def forward(self, image, text = None):
         b, device, img_size, = image.shape[0], image.device, self.image_size
         check_shape(image, 'b c h w', h = img_size, w = img_size, c = self.channels)
 
         times = torch.randint(0, self.num_timesteps, (b,), device = device, dtype = torch.long)
-        image_embed = self.get_image_embed(image)
 
-        loss = self.p_losses(x, times, image_embed = image_embed, *args, **kwargs)
+        image_embed = self.get_image_embed(image)
+        text_encodings = self.get_text_encodings(text) if exists(text) else None
+
+        loss = self.p_losses(image, times, image_embed = image_embed, text_encodings = text_encodings)
         return loss
 
 # main class
@@ -839,23 +1091,28 @@ class DALLE2(nn.Module):
         *,
         prior,
         decoder,
-        tokenizer = None
+        prior_num_samples = 2
     ):
         super().__init__()
-        assert isinstance(prior), DiffusionPrior
-        assert isinstance(decoder), Decoder
-        self.tokenizer = tokenizer
+        assert isinstance(prior, DiffusionPrior)
+        assert isinstance(decoder, Decoder)
+        self.prior = prior
+        self.decoder = decoder
+        self.prior_num_samples = prior_num_samples
 
     @torch.no_grad()
+    @eval_decorator
     def forward(
         self,
-        *,
-        text
+        text,
+        cond_scale = 1.
     ):
-        if isinstance(text, str):
-            assert exists(self.tokenizer), 'tokenizer must be passed in if you were to pass in the text as a string'
-            text = self.tokenizer.encode(text)
+        device = next(self.parameters()).device
 
-        image_embed = prior.sample(text, num_samples_per_batch = 2)
-        images = decoder.sample(image_embed)
+        if isinstance(text, str) or is_list_str(text):
+            text = [text] if not isinstance(text, (list, tuple)) else text
+            text = tokenizer.tokenize(text).to(device)
+
+        image_embed = self.prior.sample(text, num_samples_per_batch = self.prior_num_samples)
+        images = self.decoder.sample(image_embed, cond_scale = cond_scale)
         return images

setup.py

@@ -10,7 +10,7 @@ setup(
       'dream = dalle2_pytorch.cli:dream'
     ],
   },
-  version = '0.0.3',
+  version = '0.0.15',
   license='MIT',
   description = 'DALL-E 2',
   author = 'Phil Wang',
@@ -24,7 +24,8 @@ setup(
   install_requires=[
     'click',
     'einops>=0.4',
-    'einops-exts',
+    'einops-exts>=0.0.3',
+    'kornia>=0.5.4',
     'pillow',
     'torch>=1.10',
     'torchvision',