make sure gradient accumulation feature works even if all arguments passed in are keyword arguments

The default config for clip from openai should be ViT-B/32

README.md

@@ -508,7 +508,7 @@ To use a pretrained OpenAI CLIP, simply import `OpenAIClipAdapter` and pass it i
 import torch
 from dalle2_pytorch import DALLE2, DiffusionPriorNetwork, DiffusionPrior, Unet, Decoder, OpenAIClipAdapter
 
-# openai pretrained clip - defaults to ViT/B-32
+# openai pretrained clip - defaults to ViT-B/32
 
 clip = OpenAIClipAdapter()
 
@@ -732,8 +732,8 @@ clip = CLIP(
 
 # mock data
 
-text = torch.randint(0, 49408, (4, 256)).cuda()
-images = torch.randn(4, 3, 256, 256).cuda()
+text = torch.randint(0, 49408, (32, 256)).cuda()
+images = torch.randn(32, 3, 256, 256).cuda()
 
 # decoder (with unet)
 
@@ -774,8 +774,12 @@ decoder_trainer = DecoderTrainer(
 )
 
 for unet_number in (1, 2):
-    loss = decoder_trainer(images, text = text, unet_number = unet_number)  # use the decoder_trainer forward
-    loss.backward()
+    loss = decoder_trainer(
+        images,
+        text = text,
+        unet_number = unet_number, # which unet to train on
+        max_batch_size = 4         # gradient accumulation - this sets the maximum batch size in which to do forward and backwards pass - for this example 32 / 4 == 8 times
+    )
 
     decoder_trainer.update(unet_number) # update the specific unet as well as its exponential moving average
 
@@ -810,8 +814,8 @@ clip = CLIP(
 
 # mock data
 
-text = torch.randint(0, 49408, (4, 256)).cuda()
-images = torch.randn(4, 3, 256, 256).cuda()
+text = torch.randint(0, 49408, (32, 256)).cuda()
+images = torch.randn(32, 3, 256, 256).cuda()
 
 # prior networks (with transformer)
 
@@ -838,8 +842,7 @@ diffusion_prior_trainer = DiffusionPriorTrainer(
     ema_update_every = 10,
 )
 
-loss = diffusion_prior_trainer(text, images)
-loss.backward()
+loss = diffusion_prior_trainer(text, images, max_batch_size = 4)
 diffusion_prior_trainer.update()  # this will update the optimizer as well as the exponential moving averaged diffusion prior
 
 # after much of the above three lines in a loop
@@ -1004,6 +1007,7 @@ Once built, images will be saved to the same directory the command is invoked
 - [x] make sure the cascading ddpm in the repository can be trained unconditionally, offer a one-line CLI tool for training on a folder of images
 - [x] bring in cross-scale embedding from iclr paper https://github.com/lucidrains/vit-pytorch/blob/main/vit_pytorch/crossformer.py#L14
 - [x] cross embed layers for downsampling, as an option
+- [x] use an experimental tracker agnostic setup, as done <a href="https://github.com/lucidrains/tf-bind-transformer#simple-trainer-class-for-fine-tuning">here</a>
 - [ ] become an expert with unets, cleanup unet code, make it fully configurable, port all learnings over to https://github.com/lucidrains/x-unet (test out unet² in ddpm repo) - consider https://github.com/lucidrains/uformer-pytorch attention-based unet
 - [ ] transcribe code to Jax, which lowers the activation energy for distributed training, given access to TPUs
 - [ ] train on a toy task, offer in colab
@@ -1011,12 +1015,12 @@ Once built, images will be saved to the same directory the command is invoked
 - [ ] extend diffusion head to use diffusion-gan (potentially using lightweight-gan) to speed up inference
 - [ ] figure out if possible to augment with external memory, as described in https://arxiv.org/abs/2204.11824
 - [ ] test out grid attention in cascading ddpm locally, decide whether to keep or remove
-- [ ] use an experimental tracker agnostic setup, as done <a href="https://github.com/lucidrains/tf-bind-transformer#simple-trainer-class-for-fine-tuning">here</a>
 - [ ] interface out the vqgan-vae so a pretrained one can be pulled off the shelf to validate latent diffusion + DALL-E2
 - [ ] make sure FILIP works with DALL-E2 from x-clip https://arxiv.org/abs/2111.07783
 - [ ] offer save / load methods on the trainer classes to automatically take care of state dicts for scalers / optimizers / saving versions and checking for breaking changes
 - [ ] bring in skip-layer excitatons (from lightweight gan paper) to see if it helps for either decoder of unet or vqgan-vae training
 - [ ] decoder needs one day worth of refactor for tech debt
+- [ ] allow for unet to be able to condition non-cross attention style as well
 
 ## Citations
 

dalle2_pytorch/dalle2_pytorch.py

@@ -1,7 +1,7 @@
 import math
 from tqdm import tqdm
 from inspect import isfunction
-from functools import partial
+from functools import partial, wraps
 from contextlib import contextmanager
 from collections import namedtuple
 from pathlib import Path
@@ -33,6 +33,10 @@ from rotary_embedding_torch import RotaryEmbedding
 from x_clip import CLIP
 from coca_pytorch import CoCa
 
+# constants
+
+NAT = 1. / math.log(2.)
+
 # helper functions
 
 def exists(val):
@@ -41,6 +45,14 @@ def exists(val):
 def identity(t, *args, **kwargs):
     return t
 
+def maybe(fn):
+    @wraps(fn)
+    def inner(x):
+        if not exists(x):
+            return x
+        return fn(x)
+    return inner
+
 def default(val, d):
     if exists(val):
         return val
@@ -91,6 +103,9 @@ def freeze_model_and_make_eval_(model):
 
 # tensor helpers
 
+def log(t, eps = 1e-12):
+    return torch.log(t.clamp(min = eps))
+
 def l2norm(t):
     return F.normalize(t, dim = -1)
 
@@ -107,10 +122,10 @@ def resize_image_to(image, target_image_size):
 # ddpms expect images to be in the range of -1 to 1
 # but CLIP may otherwise
 
-def normalize_img(img):
+def normalize_neg_one_to_one(img):
     return img * 2 - 1
 
-def unnormalize_img(normed_img):
+def unnormalize_zero_to_one(normed_img):
     return (normed_img + 1) * 0.5
 
 # clip related adapters
@@ -271,7 +286,7 @@ class OpenAIClipAdapter(BaseClipAdapter):
     def embed_image(self, image):
         assert not self.cleared
         image = resize_image_to(image, self.image_size)
-        image = self.clip_normalize(unnormalize_img(image))
+        image = self.clip_normalize(image)
         image_embed = self.clip.encode_image(image)
         return EmbeddedImage(l2norm(image_embed.float()), None)
 
@@ -297,6 +312,36 @@ def noise_like(shape, device, repeat=False):
     noise = lambda: torch.randn(shape, device=device)
     return repeat_noise() if repeat else noise()
 
+def meanflat(x):
+    return x.mean(dim = tuple(range(1, len(x.shape))))
+
+def normal_kl(mean1, logvar1, mean2, logvar2):
+    return 0.5 * (-1.0 + logvar2 - logvar1 + torch.exp(logvar1 - logvar2) + ((mean1 - mean2) ** 2) * torch.exp(-logvar2))
+
+def approx_standard_normal_cdf(x):
+    return 0.5 * (1.0 + torch.tanh(((2.0 / math.pi) ** 0.5) * (x + 0.044715 * (x ** 3))))
+
+def discretized_gaussian_log_likelihood(x, *, means, log_scales, thres = 0.999):
+    assert x.shape == means.shape == log_scales.shape
+
+    centered_x = x - means
+    inv_stdv = torch.exp(-log_scales)
+    plus_in = inv_stdv * (centered_x + 1. / 255.)
+    cdf_plus = approx_standard_normal_cdf(plus_in)
+    min_in = inv_stdv * (centered_x - 1. / 255.)
+    cdf_min = approx_standard_normal_cdf(min_in)
+    log_cdf_plus = log(cdf_plus)
+    log_one_minus_cdf_min = log(1. - cdf_min)
+    cdf_delta = cdf_plus - cdf_min
+
+    log_probs = torch.where(x < -thres,
+        log_cdf_plus,
+        torch.where(x > thres,
+            log_one_minus_cdf_min,
+            log(cdf_delta)))
+
+    return log_probs
+
 def cosine_beta_schedule(timesteps, s = 0.008):
     """
     cosine schedule
@@ -398,12 +443,6 @@ class BaseGaussianDiffusion(nn.Module):
         register_buffer('posterior_mean_coef1', betas * torch.sqrt(alphas_cumprod_prev) / (1. - alphas_cumprod))
         register_buffer('posterior_mean_coef2', (1. - alphas_cumprod_prev) * torch.sqrt(alphas) / (1. - alphas_cumprod))
 
-    def q_mean_variance(self, x_start, t):
-        mean = extract(self.sqrt_alphas_cumprod, t, x_start.shape) * x_start
-        variance = extract(1. - self.alphas_cumprod, t, x_start.shape)
-        log_variance = extract(self.log_one_minus_alphas_cumprod, t, x_start.shape)
-        return mean, variance, log_variance
-
     def q_posterior(self, x_start, x_t, t):
         posterior_mean = (
             extract(self.posterior_mean_coef1, t, x_t.shape) * x_start +
@@ -575,7 +614,6 @@ class Attention(nn.Module):
         heads = 8,
         dropout = 0.,
         causal = False,
-        post_norm = False,
         rotary_emb = None
     ):
         super().__init__()
@@ -585,7 +623,6 @@ class Attention(nn.Module):
 
         self.causal = causal
         self.norm = LayerNorm(dim)
-        self.post_norm = LayerNorm(dim)     # sandwich norm from Coqview paper + Normformer
         self.dropout = nn.Dropout(dropout)
 
         self.null_kv = nn.Parameter(torch.randn(2, dim_head))
@@ -596,7 +633,7 @@ class Attention(nn.Module):
 
         self.to_out = nn.Sequential(
             nn.Linear(inner_dim, dim, bias = False),
-            LayerNorm(dim) if post_norm else nn.Identity()
+            LayerNorm(dim)
         )
 
     def forward(self, x, mask = None, attn_bias = None):
@@ -653,8 +690,7 @@ class Attention(nn.Module):
         out = einsum('b h i j, b j d -> b h i d', attn, v)
 
         out = rearrange(out, 'b h n d -> b n (h d)')
-        out = self.to_out(out)
-        return self.post_norm(out)
+        return self.to_out(out)
 
 class CausalTransformer(nn.Module):
     def __init__(
@@ -680,7 +716,7 @@ class CausalTransformer(nn.Module):
         self.layers = nn.ModuleList([])
         for _ in range(depth):
             self.layers.append(nn.ModuleList([
-                Attention(dim = dim, causal = True, dim_head = dim_head, heads = heads, dropout = attn_dropout, post_norm = normformer, rotary_emb = rotary_emb),
+                Attention(dim = dim, causal = True, dim_head = dim_head, heads = heads, dropout = attn_dropout, rotary_emb = rotary_emb),
                 FeedForward(dim = dim, mult = ff_mult, dropout = ff_dropout, post_activation_norm = normformer)
             ]))
 
@@ -831,7 +867,7 @@ class DiffusionPrior(BaseGaussianDiffusion):
         image_channels = 3,
         timesteps = 1000,
         cond_drop_prob = 0.,
-        loss_type = "l1",
+        loss_type = "l2",
         predict_x_start = True,
         beta_schedule = "cosine",
         condition_on_text_encodings = True, # the paper suggests this is needed, but you can turn it off for your CLIP preprocessed text embed -> image embed training
@@ -1127,6 +1163,7 @@ class CrossAttention(nn.Module):
         dim_head = 64,
         heads = 8,
         dropout = 0.,
+        norm_context = False
     ):
         super().__init__()
         self.scale = dim_head ** -0.5
@@ -1136,13 +1173,17 @@ class CrossAttention(nn.Module):
         context_dim = default(context_dim, dim)
 
         self.norm = LayerNorm(dim)
-        self.norm_context = LayerNorm(context_dim)
+        self.norm_context = LayerNorm(context_dim) if norm_context else nn.Identity()
         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)
+
+        self.to_out = nn.Sequential(
+            nn.Linear(inner_dim, dim, bias = False),
+            LayerNorm(dim)
+        )
 
     def forward(self, x, context, mask = None):
         b, n, device = *x.shape[:2], x.device
@@ -1272,6 +1313,7 @@ class Unet(nn.Module):
         out_dim = None,
         dim_mults=(1, 2, 4, 8),
         channels = 3,
+        channels_out = None,
         attn_dim_head = 32,
         attn_heads = 16,
         lowres_cond = False, # for cascading diffusion - https://cascaded-diffusion.github.io/
@@ -1302,6 +1344,7 @@ class Unet(nn.Module):
         # determine dimensions
 
         self.channels = channels
+        self.channels_out = default(channels_out, channels)
 
         init_channels = channels if not lowres_cond else channels * 2 # in cascading diffusion, one concats the low resolution image, blurred, for conditioning the higher resolution synthesis
         init_dim = default(init_dim, dim // 3 * 2)
@@ -1336,6 +1379,9 @@ class Unet(nn.Module):
             Rearrange('b (n d) -> b n d', n = num_image_tokens)
         ) if image_embed_dim != cond_dim else nn.Identity()
 
+        self.norm_cond = nn.LayerNorm(cond_dim)
+        self.norm_mid_cond = nn.LayerNorm(cond_dim)
+
         # text encoding conditioning (optional)
 
         self.text_to_cond = None
@@ -1407,11 +1453,9 @@ class Unet(nn.Module):
                 Upsample(dim_in)
             ]))
 
-        out_dim = default(out_dim, channels)
-
         self.final_conv = nn.Sequential(
             ResnetBlock(dim, dim, groups = resnet_groups[0]),
-            nn.Conv2d(dim, out_dim, 1)
+            nn.Conv2d(dim, self.channels_out, 1)
         )
 
     # if the current settings for the unet are not correct
@@ -1421,13 +1465,25 @@ class Unet(nn.Module):
         *,
         lowres_cond,
         channels,
+        channels_out,
         cond_on_image_embeds,
         cond_on_text_encodings
     ):
-        if lowres_cond == self.lowres_cond and channels == self.channels and cond_on_image_embeds == self.cond_on_image_embeds and cond_on_text_encodings == self.cond_on_text_encodings:
+        if lowres_cond == self.lowres_cond and \
+            channels == self.channels and \
+            cond_on_image_embeds == self.cond_on_image_embeds and \
+            cond_on_text_encodings == self.cond_on_text_encodings and \
+            channels_out == self.channels_out:
             return self
 
-        updated_kwargs = {'lowres_cond': lowres_cond, 'channels': channels, 'cond_on_image_embeds': cond_on_image_embeds, 'cond_on_text_encodings': cond_on_text_encodings}
+        updated_kwargs = dict(
+            lowres_cond = lowres_cond,
+            channels = channels,
+            channels_out = channels_out,
+            cond_on_image_embeds = cond_on_image_embeds,
+            cond_on_text_encodings = cond_on_text_encodings
+        )
+
         return self.__class__(**{**self._locals, **updated_kwargs})
 
     def forward_with_cond_scale(
@@ -1541,6 +1597,11 @@ class Unet(nn.Module):
 
         mid_c = c if not exists(text_tokens) else torch.cat((c, text_tokens), dim = -2)
 
+        # normalize conditioning tokens
+
+        c = self.norm_cond(c)
+        mid_c = self.norm_mid_cond(mid_c)
+
         # go through the layers of the unet, down and up
 
         hiddens = []
@@ -1614,7 +1675,7 @@ class Decoder(BaseGaussianDiffusion):
         timesteps = 1000,
         image_cond_drop_prob = 0.1,
         text_cond_drop_prob = 0.5,
-        loss_type = 'l1',
+        loss_type = 'l2',
         beta_schedule = 'cosine',
         predict_x_start = False,
         predict_x_start_for_latent_diffusion = False,
@@ -1627,6 +1688,8 @@ class Decoder(BaseGaussianDiffusion):
         clip_denoised = True,
         clip_x_start = True,
         clip_adapter_overrides = dict(),
+        learned_variance = True,
+        vb_loss_weight = 0.001,
         unconditional = False
     ):
         super().__init__(
@@ -1665,10 +1728,18 @@ class Decoder(BaseGaussianDiffusion):
         unets = cast_tuple(unet)
         vaes = pad_tuple_to_length(cast_tuple(vae), len(unets), fillvalue = NullVQGanVAE(channels = self.channels))
 
+        # whether to use learned variance, defaults to True for the first unet in the cascade, as in paper
+
+        learned_variance = pad_tuple_to_length(cast_tuple(learned_variance), len(unets), fillvalue = False)
+        self.learned_variance = learned_variance
+        self.vb_loss_weight = vb_loss_weight
+
+        # construct unets and vaes
+
         self.unets = nn.ModuleList([])
         self.vaes = nn.ModuleList([])
 
-        for ind, (one_unet, one_vae) in enumerate(zip(unets, vaes)):
+        for ind, (one_unet, one_vae, one_unet_learned_var) in enumerate(zip(unets, vaes, learned_variance)):
             assert isinstance(one_unet, Unet)
             assert isinstance(one_vae, (VQGanVAE, NullVQGanVAE))
 
@@ -1676,12 +1747,14 @@ class Decoder(BaseGaussianDiffusion):
             latent_dim = one_vae.encoded_dim if exists(one_vae) else None
 
             unet_channels = default(latent_dim, self.channels)
+            unet_channels_out = unet_channels * (1 if not one_unet_learned_var else 2)
 
             one_unet = one_unet.cast_model_parameters(
                 lowres_cond = not is_first,
                 cond_on_image_embeds = is_first and not unconditional,
                 cond_on_text_encodings = one_unet.cond_on_text_encodings and not unconditional,
-                channels = unet_channels
+                channels = unet_channels,
+                channels_out = unet_channels_out
             )
 
             self.unets.append(one_unet)
@@ -1744,8 +1817,11 @@ class Decoder(BaseGaussianDiffusion):
         yield
         unet.cpu()
 
-    def p_mean_variance(self, unet, x, t, image_embed, text_encodings = None, text_mask = None, lowres_cond_img = None, clip_denoised = True, predict_x_start = False, cond_scale = 1.):
-        pred = unet.forward_with_cond_scale(x, t, image_embed = image_embed, text_encodings = text_encodings, text_mask = text_mask, cond_scale = cond_scale, lowres_cond_img = lowres_cond_img)
+    def p_mean_variance(self, unet, x, t, image_embed, text_encodings = None, text_mask = None, lowres_cond_img = None, clip_denoised = True, predict_x_start = False, learned_variance = False, cond_scale = 1., model_output = None):
+        pred = default(model_output, lambda: unet.forward_with_cond_scale(x, t, image_embed = image_embed, text_encodings = text_encodings, text_mask = text_mask, cond_scale = cond_scale, lowres_cond_img = lowres_cond_img))
+
+        if learned_variance:
+            pred, var_interp_frac_unnormalized = pred.chunk(2, dim = 1)
 
         if predict_x_start:
             x_recon = pred
@@ -1756,24 +1832,38 @@ class Decoder(BaseGaussianDiffusion):
             x_recon.clamp_(-1., 1.)
 
         model_mean, posterior_variance, posterior_log_variance = self.q_posterior(x_start=x_recon, x_t=x, t=t)
+
+        if learned_variance:
+            # if learned variance, posterio variance and posterior log variance are predicted by the network
+            # by an interpolation of the max and min log beta values
+            # eq 15 - https://arxiv.org/abs/2102.09672
+            min_log = extract(self.posterior_log_variance_clipped, t, x.shape)
+            max_log = extract(torch.log(self.betas), t, x.shape)
+            var_interp_frac = unnormalize_zero_to_one(var_interp_frac_unnormalized)
+
+            posterior_log_variance = var_interp_frac * max_log + (1 - var_interp_frac) * min_log
+            posterior_variance = posterior_log_variance.exp()
+
         return model_mean, posterior_variance, posterior_log_variance
 
     @torch.inference_mode()
-    def p_sample(self, unet, x, t, image_embed, text_encodings = None, text_mask = None, cond_scale = 1., lowres_cond_img = None, predict_x_start = False, clip_denoised = True, repeat_noise = False):
+    def p_sample(self, unet, x, t, image_embed, text_encodings = None, text_mask = None, cond_scale = 1., lowres_cond_img = None, predict_x_start = False, learned_variance = False, clip_denoised = True, repeat_noise = False):
         b, *_, device = *x.shape, x.device
-        model_mean, _, model_log_variance = self.p_mean_variance(unet, x = x, t = t, image_embed = image_embed, text_encodings = text_encodings, text_mask = text_mask, cond_scale = cond_scale, lowres_cond_img = lowres_cond_img, clip_denoised = clip_denoised, predict_x_start = predict_x_start)
+        model_mean, _, model_log_variance = self.p_mean_variance(unet, x = x, t = t, image_embed = image_embed, text_encodings = text_encodings, text_mask = text_mask, cond_scale = cond_scale, lowres_cond_img = lowres_cond_img, clip_denoised = clip_denoised, predict_x_start = predict_x_start, learned_variance = learned_variance)
         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.inference_mode()
-    def p_sample_loop(self, unet, shape, image_embed, predict_x_start = False, clip_denoised = True, lowres_cond_img = None, text_encodings = None, text_mask = None, cond_scale = 1):
+    def p_sample_loop(self, unet, shape, image_embed, predict_x_start = False, learned_variance = False, clip_denoised = True, lowres_cond_img = None, text_encodings = None, text_mask = None, cond_scale = 1):
         device = self.betas.device
 
         b = shape[0]
         img = torch.randn(shape, device = device)
 
+        lowres_cond_img = maybe(normalize_neg_one_to_one)(lowres_cond_img)
+
         for i in tqdm(reversed(range(0, self.num_timesteps)), desc = 'sampling loop time step', total = self.num_timesteps):
             img = self.p_sample(
                 unet,
@@ -1785,17 +1875,26 @@ class Decoder(BaseGaussianDiffusion):
                 cond_scale = cond_scale,
                 lowres_cond_img = lowres_cond_img,
                 predict_x_start = predict_x_start,
+                learned_variance = learned_variance,
                 clip_denoised = clip_denoised
             )
 
-        return img
+        unnormalize_img = unnormalize_zero_to_one(img)
+        return unnormalize_img
 
-    def p_losses(self, unet, x_start, times, *, image_embed, lowres_cond_img = None, text_encodings = None, text_mask = None, predict_x_start = False, noise = None):
+    def p_losses(self, unet, x_start, times, *, image_embed, lowres_cond_img = None, text_encodings = None, text_mask = None, predict_x_start = False, noise = None, learned_variance = False, clip_denoised = False):
         noise = default(noise, lambda: torch.randn_like(x_start))
 
+        # normalize to [-1, 1]
+
+        x_start = normalize_neg_one_to_one(x_start)
+        lowres_cond_img = maybe(normalize_neg_one_to_one)(lowres_cond_img)
+
+        # get x_t
+
         x_noisy = self.q_sample(x_start = x_start, t = times, noise = noise)
 
-        pred = unet(
+        model_output = unet(
             x_noisy,
             times,
             image_embed = image_embed,
@@ -1806,10 +1905,48 @@ class Decoder(BaseGaussianDiffusion):
             text_cond_drop_prob = self.text_cond_drop_prob,
         )
 
+        if learned_variance:
+            pred, _ = model_output.chunk(2, dim = 1)
+        else:
+            pred = model_output
+
         target = noise if not predict_x_start else x_start
 
         loss = self.loss_fn(pred, target)
-        return loss
+
+        if not learned_variance:
+            # return simple loss if not using learned variance
+            return loss
+
+        # most of the code below is transcribed from
+        # https://github.com/hojonathanho/diffusion/blob/master/diffusion_tf/diffusion_utils_2.py
+        # the Improved DDPM paper then further modified it so that the mean is detached (shown a couple lines before), and weighted to be smaller than the l1 or l2 "simple" loss
+        # it is questionable whether this is really needed, looking at some of the figures in the paper, but may as well stay faithful to their implementation
+
+        # if learning the variance, also include the extra weight kl loss
+
+        true_mean, _, true_log_variance_clipped = self.q_posterior(x_start = x_start, x_t = x_noisy, t = times)
+        model_mean, _, model_log_variance = self.p_mean_variance(unet, x = x_noisy, t = times, image_embed = image_embed, clip_denoised = clip_denoised, learned_variance = True, model_output = model_output)
+
+        # kl loss with detached model predicted mean, for stability reasons as in paper
+
+        detached_model_mean = model_mean.detach()
+
+        kl = normal_kl(true_mean, true_log_variance_clipped, detached_model_mean, model_log_variance)
+        kl = meanflat(kl) * NAT
+
+        decoder_nll = -discretized_gaussian_log_likelihood(x_start, means = detached_model_mean, log_scales = 0.5 * model_log_variance)
+        decoder_nll = meanflat(decoder_nll) * NAT
+
+        # at the first timestep return the decoder NLL, otherwise return KL(q(x_{t-1}|x_t,x_0) || p(x_{t-1}|x_t))
+
+        vb_losses = torch.where(times == 0, decoder_nll, kl)
+
+        # weight the vb loss smaller, for stability, as in the paper (recommended 0.001)
+
+        vb_loss = vb_losses.mean() * self.vb_loss_weight
+
+        return loss + vb_loss
 
     @torch.inference_mode()
     @eval_decorator
@@ -1836,7 +1973,7 @@ class Decoder(BaseGaussianDiffusion):
         img = None
         is_cuda = next(self.parameters()).is_cuda
 
-        for unet_number, unet, vae, channel, image_size, predict_x_start in tqdm(zip(range(1, len(self.unets) + 1), self.unets, self.vaes, self.sample_channels, self.image_sizes, self.predict_x_start)):
+        for unet_number, unet, vae, channel, image_size, predict_x_start, learned_variance in tqdm(zip(range(1, len(self.unets) + 1), self.unets, self.vaes, self.sample_channels, self.image_sizes, self.predict_x_start, self.learned_variance)):
 
             context = self.one_unet_in_gpu(unet = unet) if is_cuda else null_context()
 
@@ -1862,6 +1999,7 @@ class Decoder(BaseGaussianDiffusion):
                     text_mask = text_mask,
                     cond_scale = cond_scale,
                     predict_x_start = predict_x_start,
+                    learned_variance = learned_variance,
                     clip_denoised = not is_latent_diffusion,
                     lowres_cond_img = lowres_cond_img
                 )
@@ -1891,6 +2029,7 @@ class Decoder(BaseGaussianDiffusion):
         target_image_size   = self.image_sizes[unet_index]
         predict_x_start     = self.predict_x_start[unet_index]
         random_crop_size    = self.random_crop_sizes[unet_index]
+        learned_variance    = self.learned_variance[unet_index]
         b, c, h, w, device, = *image.shape, image.device
 
         check_shape(image, 'b c h w', c = self.channels)
@@ -1928,7 +2067,7 @@ class Decoder(BaseGaussianDiffusion):
             if exists(lowres_cond_img):
                 lowres_cond_img = vae.encode(lowres_cond_img)
 
-        return self.p_losses(unet, image, times, image_embed = image_embed, text_encodings = text_encodings, text_mask = text_mask, lowres_cond_img = lowres_cond_img, predict_x_start = predict_x_start)
+        return self.p_losses(unet, image, times, image_embed = image_embed, text_encodings = text_encodings, text_mask = text_mask, lowres_cond_img = lowres_cond_img, predict_x_start = predict_x_start, learned_variance = learned_variance)
 
 # main class
 
@@ -1978,4 +2117,3 @@ class DALLE2(nn.Module):
             return images[0]
 
         return images
-

dalle2_pytorch/optimizer.py

@@ -7,16 +7,17 @@ def separate_weight_decayable_params(params):
 
 def get_optimizer(
     params,
-    lr = 3e-4,
+    lr = 2e-5,
     wd = 1e-2,
     betas = (0.9, 0.999),
+    eps = 1e-8,
     filter_by_requires_grad = False
 ):
     if filter_by_requires_grad:
         params = list(filter(lambda t: t.requires_grad, params))
 
     if wd == 0:
-        return Adam(params, lr = lr, betas = betas)
+        return Adam(params, lr = lr, betas = betas, eps = eps)
 
     params = set(params)
     wd_params, no_wd_params = separate_weight_decayable_params(params)
@@ -26,4 +27,4 @@ def get_optimizer(
         {'params': list(no_wd_params), 'weight_decay': 0},
     ]
 
-    return AdamW(param_groups, lr = lr, weight_decay = wd, betas = betas)
+    return AdamW(param_groups, lr = lr, weight_decay = wd, betas = betas, eps = eps)

dalle2_pytorch/trackers.py

@@ -0,0 +1,49 @@
+import os
+import torch
+from torch import nn
+
+# helper functions
+
+def exists(val):
+    return val is not None
+
+# base class
+
+class BaseTracker(nn.Module):
+    def __init__(self):
+        super().__init__()
+
+    def init(self, config, **kwargs):
+        raise NotImplementedError
+
+    def log(self, log, **kwargs):
+        raise NotImplementedError
+
+# basic stdout class
+
+class ConsoleTracker(BaseTracker):
+    def init(self, **config):
+        print(config)
+
+    def log(self, log, **kwargs):
+        print(log)
+
+# basic wandb class
+
+class WandbTracker(BaseTracker):
+    def __init__(self):
+        super().__init__()
+        try:
+            import wandb
+        except ImportError as e:
+            print('`pip install wandb` to use the wandb experiment tracker')
+            raise e
+
+        os.environ["WANDB_SILENT"] = "true"
+        self.wandb = wandb
+
+    def init(self, **config):
+        self.wandb.init(**config)
+
+    def log(self, log, **kwargs):
+        self.wandb.log(log, **kwargs)

dalle2_pytorch/train.py

@@ -1,6 +1,8 @@
 import time
 import copy
+from math import ceil
 from functools import partial
+from collections.abc import Iterable
 
 import torch
 from torch import nn
@@ -14,6 +16,9 @@ from dalle2_pytorch.optimizer import get_optimizer
 def exists(val):
     return val is not None
 
+def default(val, d):
+    return val if exists(val) else d
+
 def cast_tuple(val, length = 1):
     return val if isinstance(val, tuple) else ((val,) * length)
 
@@ -40,6 +45,56 @@ def groupby_prefix_and_trim(prefix, d):
     kwargs_without_prefix = dict(map(lambda x: (x[0][len(prefix):], x[1]), tuple(kwargs_with_prefix.items())))
     return kwargs_without_prefix, kwargs
 
+# gradient accumulation functions
+
+def split_iterable(it, split_size):
+    accum = []
+    for ind in range(ceil(len(it) / split_size)):
+        start_index = ind * split_size
+        accum.append(it[start_index: (start_index + split_size)])
+    return accum
+
+def split(t, split_size = None):
+    if not exists(split_size):
+        return t
+
+    if isinstance(t, torch.Tensor):
+        return t.split(split_size, dim = 0)
+
+    if isinstance(t, Iterable):
+        return split_iterable(t, split_size)
+
+    return TypeError
+
+def find_first(cond, arr):
+    for el in arr:
+        if cond(el):
+            return el
+    return None
+
+def split_args_and_kwargs(*args, split_size = None, **kwargs):
+    all_args = (*args, *kwargs.values())
+    len_all_args = len(all_args)
+    first_tensor = find_first(lambda t: isinstance(t, torch.Tensor), all_args)
+    assert exists(first_tensor)
+
+    batch_size = len(first_tensor)
+    split_size = default(split_size, batch_size)
+    chunk_size = ceil(batch_size / split_size)
+
+    dict_len = len(kwargs)
+    dict_keys = kwargs.keys()
+    split_kwargs_index = len_all_args - dict_len
+
+    split_all_args = [split(arg, split_size = split_size) if exists(arg) and isinstance(arg, (torch.Tensor, Iterable)) else ((arg,) * chunk_size) for arg in all_args]
+    chunk_sizes = tuple(map(len, split_all_args[0]))
+
+    for (chunk_size, *chunked_all_args) in tuple(zip(chunk_sizes, *split_all_args)):
+        chunked_args, chunked_kwargs_values = chunked_all_args[:split_kwargs_index], chunked_all_args[split_kwargs_index:]
+        chunked_kwargs = dict(tuple(zip(dict_keys, chunked_kwargs_values)))
+        chunk_size_frac = chunk_size / batch_size
+        yield chunk_size_frac, (chunked_args, chunked_kwargs)
+
 # print helpers
 
 def print_ribbon(s, symbol = '=', repeat = 40):
@@ -71,7 +126,7 @@ def load_diffusion_model(dprior_path, device):
     # Load state dict from saved model
     diffusion_prior.load_state_dict(loaded_obj['model'])
 
-    return diffusion_prior
+    return diffusion_prior, loaded_obj
 
 def save_diffusion_model(save_path, model, optimizer, scaler, config, image_embed_dim):
     # Saving State Dict
@@ -90,7 +145,7 @@ class EMA(nn.Module):
     def __init__(
         self,
         model,
-        beta = 0.99,
+        beta = 0.9999,
         update_after_step = 1000,
         update_every = 10,
     ):
@@ -105,6 +160,10 @@ class EMA(nn.Module):
         self.register_buffer('initted', torch.Tensor([False]))
         self.register_buffer('step', torch.tensor([0.]))
 
+    def restore_ema_model_device(self):
+        device = self.initted.device
+        self.ema_model.to(device)
+
     def update(self):
         self.step += 1
 
@@ -143,6 +202,7 @@ class DiffusionPriorTrainer(nn.Module):
         use_ema = True,
         lr = 3e-4,
         wd = 1e-2,
+        eps = 1e-6,
         max_grad_norm = None,
         amp = False,
         **kwargs
@@ -169,6 +229,7 @@ class DiffusionPriorTrainer(nn.Module):
             diffusion_prior.parameters(),
             lr = lr,
             wd = wd,
+            eps = eps,
             **kwargs
         )
 
@@ -176,6 +237,8 @@ class DiffusionPriorTrainer(nn.Module):
 
         self.max_grad_norm = max_grad_norm
 
+        self.register_buffer('step', torch.tensor([0.]))
+
     def update(self):
         if exists(self.max_grad_norm):
             self.scaler.unscale_(self.optimizer)
@@ -188,6 +251,8 @@ class DiffusionPriorTrainer(nn.Module):
         if self.use_ema:
             self.ema_diffusion_prior.update()
 
+        self.step += 1
+
     @torch.inference_mode()
     def p_sample_loop(self, *args, **kwargs):
         return self.ema_diffusion_prior.ema_model.p_sample_loop(*args, **kwargs)
@@ -203,12 +268,20 @@ class DiffusionPriorTrainer(nn.Module):
     def forward(
         self,
         *args,
-        divisor = 1,
+        max_batch_size = None,
         **kwargs
     ):
-        with autocast(enabled = self.amp):
-            loss = self.diffusion_prior(*args, **kwargs)
-        return self.scaler.scale(loss / divisor)
+        total_loss = 0.
+
+        for chunk_size_frac, (chunked_args, chunked_kwargs) in split_args_and_kwargs(*args, split_size = max_batch_size, **kwargs):
+            with autocast(enabled = self.amp):
+                loss = self.diffusion_prior(*chunked_args, **chunked_kwargs)
+                loss = loss * chunk_size_frac
+
+            total_loss += loss.item()
+            self.scaler.scale(loss).backward()
+
+        return total_loss
 
 # decoder trainer
 
@@ -217,8 +290,9 @@ class DecoderTrainer(nn.Module):
         self,
         decoder,
         use_ema = True,
-        lr = 3e-4,
+        lr = 2e-5,
         wd = 1e-2,
+        eps = 1e-8,
         max_grad_norm = None,
         amp = False,
         **kwargs
@@ -243,13 +317,14 @@ class DecoderTrainer(nn.Module):
         # be able to finely customize learning rate, weight decay
         # per unet
 
-        lr, wd = map(partial(cast_tuple, length = self.num_unets), (lr, wd))
+        lr, wd, eps = map(partial(cast_tuple, length = self.num_unets), (lr, wd, eps))
 
-        for ind, (unet, unet_lr, unet_wd) in enumerate(zip(self.decoder.unets, lr, wd)):
+        for ind, (unet, unet_lr, unet_wd, unet_eps) in enumerate(zip(self.decoder.unets, lr, wd, eps)):
             optimizer = get_optimizer(
                 unet.parameters(),
                 lr = unet_lr,
                 wd = unet_wd,
+                eps = unet_eps,
                 **kwargs
             )
 
@@ -265,6 +340,8 @@ class DecoderTrainer(nn.Module):
 
         self.max_grad_norm = max_grad_norm
 
+        self.register_buffer('step', torch.tensor([0.]))
+
     @property
     def unets(self):
         return nn.ModuleList([ema.ema_model for ema in self.ema_unets])
@@ -295,6 +372,8 @@ class DecoderTrainer(nn.Module):
             ema_unet = self.ema_unets[index]
             ema_unet.update()
 
+        self.step += 1
+
     @torch.no_grad()
     def sample(self, *args, **kwargs):
         if self.use_ema:
@@ -305,16 +384,28 @@ class DecoderTrainer(nn.Module):
 
         if self.use_ema:
             self.decoder.unets = trainable_unets             # restore original training unets
+
+        # cast the ema_model unets back to original device
+        for ema in self.ema_unets:
+            ema.restore_ema_model_device()
+
         return output
 
     def forward(
         self,
-        x,
-        *,
+        *args,
         unet_number,
-        divisor = 1,
+        max_batch_size = None,
         **kwargs
     ):
-        with autocast(enabled = self.amp):
-            loss = self.decoder(x, unet_number = unet_number, **kwargs)
-        return self.scale(loss / divisor, unet_number = unet_number)
+        total_loss = 0.
+
+        for chunk_size_frac, (chunked_args, chunked_kwargs) in split_args_and_kwargs(*args, split_size = max_batch_size, **kwargs):
+            with autocast(enabled = self.amp):
+                loss = self.decoder(*chunked_args, unet_number = unet_number, **chunked_kwargs)
+                loss = loss * chunk_size_frac
+
+            total_loss += loss.item()
+            self.scale(loss, unet_number = unet_number).backward()
+
+        return total_loss

setup.py

@@ -10,7 +10,7 @@ setup(
       'dream = dalle2_pytorch.cli:dream'
     ],
   },
-  version = '0.2.10',
+  version = '0.2.31',
   license='MIT',
   description = 'DALL-E 2',
   author = 'Phil Wang',

train_diffusion_prior.py

@@ -1,24 +1,42 @@
-import os
+from pathlib import Path
+import click
 import math
-import argparse
+import time
 import numpy as np
 
 import torch
 from torch import nn
-from embedding_reader import EmbeddingReader
-from dalle2_pytorch import DiffusionPrior, DiffusionPriorNetwork
-from dalle2_pytorch.train import load_diffusion_model, save_diffusion_model, print_ribbon
-from dalle2_pytorch.optimizer import get_optimizer
-from torch.cuda.amp import autocast,GradScaler
 
-import time
+from dalle2_pytorch import DiffusionPrior, DiffusionPriorNetwork
+from dalle2_pytorch.train import DiffusionPriorTrainer, load_diffusion_model, save_diffusion_model, print_ribbon
+from dalle2_pytorch.trackers import ConsoleTracker, WandbTracker
+
+from embedding_reader import EmbeddingReader
+
 from tqdm import tqdm
 
-import wandb
-os.environ["WANDB_SILENT"] = "true"
+# constants
+
 NUM_TEST_EMBEDDINGS = 100 # for cosine similarity reporting during training
 REPORT_METRICS_EVERY = 100 # for cosine similarity and other metric reporting during training
 
+tracker = WandbTracker()
+
+# helpers functions
+
+def exists(val):
+    val is not None
+
+class Timer:
+    def __init__(self):
+        self.reset()
+
+    def reset(self):
+        self.last_time = time.time()
+
+    def elapsed(self):
+        return time.time() - self.last_time
+# functions
 
 def eval_model(model,device,image_reader,text_reader,start,end,batch_size,loss_type,phase="Validation"):
     model.eval()
@@ -40,7 +58,7 @@ def eval_model(model,device,image_reader,text_reader,start,end,batch_size,loss_t
             total_samples += batches
 
         avg_loss = (total_loss / total_samples)
-        wandb.log({f'{phase} {loss_type}': avg_loss})
+        tracker.log({f'{phase} {loss_type}': avg_loss})
 
 def report_cosine_sims(diffusion_prior,image_reader,text_reader,train_set_size,NUM_TEST_EMBEDDINGS,device):
     diffusion_prior.eval()
@@ -87,7 +105,7 @@ def report_cosine_sims(diffusion_prior,image_reader,text_reader,train_set_size,N
            text_embed, predicted_unrelated_embeddings).cpu().numpy()
        predicted_img_similarity = cos(
            test_image_embeddings, predicted_image_embeddings).cpu().numpy()
-       wandb.log({"CosineSimilarity(text_embed,image_embed)": np.mean(original_similarity),
+       tracker.log({"CosineSimilarity(text_embed,image_embed)": np.mean(original_similarity),
             "CosineSimilarity(text_embed,predicted_image_embed)":np.mean(predicted_similarity),
             "CosineSimilarity(orig_image_embed,predicted_image_embed)":np.mean(predicted_img_similarity),
             "CosineSimilarity(text_embed,predicted_unrelated_embed)": np.mean(unrelated_similarity),
@@ -124,48 +142,67 @@ def train(image_embed_dim,
           dropout=0.05,
           amp=False):
 
-    # DiffusionPriorNetwork 
+    # diffusion prior network
+
     prior_network = DiffusionPriorNetwork( 
-            dim = image_embed_dim, 
-            depth = dpn_depth, 
-            dim_head = dpn_dim_head, 
-            heads = dpn_heads,
-            attn_dropout = dropout,
-            ff_dropout = dropout,
-            normformer = dp_normformer).to(device)
+        dim = image_embed_dim,
+        depth = dpn_depth,
+        dim_head = dpn_dim_head,
+        heads = dpn_heads,
+        attn_dropout = dropout,
+        ff_dropout = dropout,
+        normformer = dp_normformer
+    )
     
-    # DiffusionPrior with text embeddings and image embeddings pre-computed
+    # diffusion prior with text embeddings and image embeddings pre-computed
+
     diffusion_prior = DiffusionPrior( 
-            net = prior_network, 
-            clip = clip, 
-            image_embed_dim = image_embed_dim, 
-            timesteps = dp_timesteps,
-            cond_drop_prob = dp_cond_drop_prob, 
-            loss_type = dp_loss_type, 
-            condition_on_text_encodings = dp_condition_on_text_encodings).to(device)
+        net = prior_network,
+        clip = clip,
+        image_embed_dim = image_embed_dim,
+        timesteps = dp_timesteps,
+        cond_drop_prob = dp_cond_drop_prob,
+        loss_type = dp_loss_type,
+        condition_on_text_encodings = dp_condition_on_text_encodings
+    )
 
     # Load pre-trained model from DPRIOR_PATH
+
     if RESUME:
-        diffusion_prior=load_diffusion_model(DPRIOR_PATH,device)   
-        wandb.init( entity=wandb_entity, project=wandb_project, config=config) 
+        diffusion_prior, loaded_obj = load_diffusion_model(DPRIOR_PATH, device)
+        tracker.init(entity = wandb_entity, project = wandb_project, config = config)
+
+    # diffusion prior trainer
+
+    trainer = DiffusionPriorTrainer(
+        diffusion_prior = diffusion_prior,
+        lr = learning_rate,
+        wd = weight_decay,
+        max_grad_norm = max_grad_norm,
+        amp = amp,
+    ).to(device)
+
+    # load optimizer and scaler
+
+    if RESUME:
+        trainer.optimizer.load_state_dict(loaded_obj['optimizer'])
+        trainer.scaler.load_state_dict(loaded_obj['scaler'])
 
     # Create save_path if it doesn't exist
-    if not os.path.exists(save_path):
-        os.makedirs(save_path)
+
+    Path(save_path).mkdir(exist_ok = True, parents = True)
 
     # Get image and text embeddings from the servers
+
     print_ribbon("Downloading embeddings - image and text")
     image_reader = EmbeddingReader(embeddings_folder=image_embed_url, file_format="npy")
     text_reader  = EmbeddingReader(embeddings_folder=text_embed_url, file_format="npy")
     num_data_points = text_reader.count
 
     ### Training code ###
-    scaler = GradScaler(enabled=amp)
-    optimizer = get_optimizer(diffusion_prior.net.parameters(), wd=weight_decay, lr=learning_rate)
-    epochs = num_epochs
 
-    step = 0
-    t = time.time()
+    timer = Timer()
+    epochs = num_epochs
 
     train_set_size = int(train_percent*num_data_points)
     val_set_size = int(val_percent*num_data_points)
@@ -176,32 +213,31 @@ def train(image_embed_dim,
         for emb_images,emb_text in zip(image_reader(batch_size=batch_size, start=0, end=train_set_size),
                 text_reader(batch_size=batch_size, start=0, end=train_set_size)):
 
-            diffusion_prior.train()
+            trainer.train()
             
             emb_images_tensor = torch.tensor(emb_images[0]).to(device)
             emb_text_tensor = torch.tensor(emb_text[0]).to(device)
 
-            with autocast(enabled=amp):
-                loss = diffusion_prior(text_embed = emb_text_tensor,image_embed = emb_images_tensor)
-                scaler.scale(loss).backward()
+            loss = trainer(text_embed = emb_text_tensor, image_embed = emb_images_tensor)
 
             # Samples per second
-            step+=1
-            samples_per_sec = batch_size*step/(time.time()-t)
+
+            samples_per_sec = batch_size * step / timer.elapsed()
+
             # Save checkpoint every save_interval minutes
-            if(int(time.time()-t) >= 60*save_interval):
-                t = time.time()
+            if(int(timer.elapsed()) >= 60 * save_interval):
+                timer.reset()
 
                 save_diffusion_model(
                     save_path,
                     diffusion_prior,
-                    optimizer,
-                    scaler,
+                    trainer.optimizer,
+                    trainer.scaler,
                     config,
                     image_embed_dim)
 
             # Log to wandb
-            wandb.log({"Training loss": loss.item(),
+            tracker.log({"Training loss": loss.item(),
                         "Steps": step,
                         "Samples per second": samples_per_sec})
             # Log cosineSim(text_embed,predicted_image_embed) - cosineSim(text_embed,image_embed)
@@ -225,91 +261,109 @@ def train(image_embed_dim,
                         dp_loss_type,
                         phase="Validation")
 
-            scaler.unscale_(optimizer)
-            nn.utils.clip_grad_norm_(diffusion_prior.parameters(), max_grad_norm)
-
-            scaler.step(optimizer)
-            scaler.update()
-            optimizer.zero_grad()
+            trainer.update()
 
     ### Test run ###
     test_set_size = int(test_percent*train_set_size) 
-    start=train_set_size+val_set_size
-    end=num_data_points
+    start = train_set_size+val_set_size
+    end = num_data_points
     eval_model(diffusion_prior,device,image_reader,text_reader,start,end,batch_size,dp_loss_type,phase="Test")
 
-def main():
-    parser = argparse.ArgumentParser()
-    # Logging
-    parser.add_argument("--wandb-entity", type=str, default="laion")
-    parser.add_argument("--wandb-project", type=str, default="diffusion-prior")
-    parser.add_argument("--wandb-dataset", type=str, default="LAION-5B")
-    parser.add_argument("--wandb-arch", type=str, default="DiffusionPrior")
-    # URLs for embeddings 
-    parser.add_argument("--image-embed-url", type=str, default="https://mystic.the-eye.eu/public/AI/cah/laion5b/embeddings/laion2B-en/img_emb/")
-    parser.add_argument("--text-embed-url", type=str, default="https://mystic.the-eye.eu/public/AI/cah/laion5b/embeddings/laion2B-en/text_emb/")
-    # Hyperparameters
-    parser.add_argument("--learning-rate", type=float, default=1.1e-4)
-    parser.add_argument("--weight-decay", type=float, default=6.02e-2)
-    parser.add_argument("--dropout", type=float, default=5e-2)
-    parser.add_argument("--max-grad-norm", type=float, default=0.5)
-    parser.add_argument("--batch-size", type=int, default=10**4)
-    parser.add_argument("--num-epochs", type=int, default=5)
-    # Image embed dimension
-    parser.add_argument("--image-embed-dim", type=int, default=768)
-    # Train-test split
-    parser.add_argument("--train-percent", type=float, default=0.7)
-    parser.add_argument("--val-percent", type=float, default=0.2)
-    parser.add_argument("--test-percent", type=float, default=0.1)
-    # LAION training(pre-computed embeddings)
-    # DiffusionPriorNetwork(dpn) parameters
-    parser.add_argument("--dpn-depth", type=int, default=6)
-    parser.add_argument("--dpn-dim-head", type=int, default=64)
-    parser.add_argument("--dpn-heads", type=int, default=8)
-    # DiffusionPrior(dp) parameters
-    parser.add_argument("--dp-condition-on-text-encodings", type=bool, default=False)
-    parser.add_argument("--dp-timesteps", type=int, default=100)
-    parser.add_argument("--dp-normformer", type=bool, default=False)
-    parser.add_argument("--dp-cond-drop-prob", type=float, default=0.1)
-    parser.add_argument("--dp-loss-type", type=str, default="l2")
-    parser.add_argument("--clip", type=str, default=None)
-    parser.add_argument("--amp", type=bool, default=False)
-    # Model checkpointing interval(minutes)
-    parser.add_argument("--save-interval", type=int, default=30)
-    parser.add_argument("--save-path", type=str, default="./diffusion_prior_checkpoints")
-    # Saved model path 
-    parser.add_argument("--pretrained-model-path", type=str, default=None)
+@click.command()
+@click.option("--wandb-entity", default="laion")
+@click.option("--wandb-project", default="diffusion-prior")
+@click.option("--wandb-dataset", default="LAION-5B")
+@click.option("--wandb-arch", default="DiffusionPrior")
+@click.option("--image-embed-url", default="https://mystic.the-eye.eu/public/AI/cah/laion5b/embeddings/laion2B-en/img_emb/")
+@click.option("--text-embed-url", default="https://mystic.the-eye.eu/public/AI/cah/laion5b/embeddings/laion2B-en/text_emb/")
+@click.option("--learning-rate", default=1.1e-4)
+@click.option("--weight-decay", default=6.02e-2)
+@click.option("--dropout", default=5e-2)
+@click.option("--max-grad-norm", default=0.5)
+@click.option("--batch-size", default=10**4)
+@click.option("--num-epochs", default=5)
+@click.option("--image-embed-dim", default=768)
+@click.option("--train-percent", default=0.7)
+@click.option("--val-percent", default=0.2)
+@click.option("--test-percent", default=0.1)
+@click.option("--dpn-depth", default=6)
+@click.option("--dpn-dim-head", default=64)
+@click.option("--dpn-heads", default=8)
+@click.option("--dp-condition-on-text-encodings", default=False)
+@click.option("--dp-timesteps", default=100)
+@click.option("--dp-normformer", default=False)
+@click.option("--dp-cond-drop-prob", default=0.1)
+@click.option("--dp-loss-type", default="l2")
+@click.option("--clip", default=None)
+@click.option("--amp", default=False)
+@click.option("--save-interval", default=30)
+@click.option("--save-path", default="./diffusion_prior_checkpoints")
+@click.option("--pretrained-model-path", default=None)
+def main(
+    wandb_entity,
+    wandb_project,
+    wandb_dataset,
+    wandb_arch,
+    image_embed_url,
+    text_embed_url,
+    learning_rate,
+    weight_decay,
+    dropout,
+    max_grad_norm,
+    batch_size,
+    num_epochs,
+    image_embed_dim,
+    train_percent,
+    val_percent,
+    test_percent,
+    dpn_depth,
+    dpn_dim_head,
+    dpn_heads,
+    dp_condition_on_text_encodings,
+    dp_timesteps,
+    dp_normformer,
+    dp_cond_drop_prob,
+    dp_loss_type,
+    clip,
+    amp,
+    save_interval,
+    save_path,
+    pretrained_model_path
+):
+    config = {
+        "learning_rate": learning_rate,
+        "architecture": wandb_arch,
+        "dataset": wandb_dataset,
+        "weight_decay": weight_decay,
+        "max_gradient_clipping_norm": max_grad_norm,
+        "batch_size": batch_size,
+        "epochs": num_epochs,
+        "diffusion_prior_network": {
+            "depth": dpn_depth,
+            "dim_head": dpn_dim_head,
+            "heads": dpn_heads,
+            "normformer": dp_normformer
+        },
+        "diffusion_prior": {
+            "condition_on_text_encodings": dp_condition_on_text_encodings,
+            "timesteps": dp_timesteps,
+            "cond_drop_prob": dp_cond_drop_prob,
+            "loss_type": dp_loss_type,
+            "clip": clip
+        }
+    }
 
-    args = parser.parse_args()
-
-    config = ({"learning_rate": args.learning_rate,
-        "architecture": args.wandb_arch,
-        "dataset": args.wandb_dataset,
-        "weight_decay":args.weight_decay,
-        "max_gradient_clipping_norm":args.max_grad_norm,
-        "batch_size":args.batch_size,
-        "epochs": args.num_epochs,
-        "diffusion_prior_network":{"depth":args.dpn_depth,
-        "dim_head":args.dpn_dim_head,
-        "heads":args.dpn_heads,
-        "normformer":args.dp_normformer},
-        "diffusion_prior":{"condition_on_text_encodings": args.dp_condition_on_text_encodings,
-        "timesteps": args.dp_timesteps,
-        "cond_drop_prob":args.dp_cond_drop_prob,
-        "loss_type":args.dp_loss_type,
-        "clip":args.clip}
-        })
-
-    RESUME = False
     # Check if DPRIOR_PATH exists(saved model path)
+
     DPRIOR_PATH = args.pretrained_model_path
-    if(DPRIOR_PATH is not None):
-        RESUME = True
-    else:
-        wandb.init(
-          entity=args.wandb_entity,
-          project=args.wandb_project,
-          config=config)
+    RESUME = exists(DPRIOR_PATH)
+
+    if not RESUME:
+        tracker.init(
+            entity = wandb_entity,
+            project = wandb_project,
+            config = config
+        )
 
     # Obtain the utilized device.
 
@@ -319,36 +373,36 @@ def main():
         torch.cuda.set_device(device)
 
     # Training loop
-    train(args.image_embed_dim,
-          args.image_embed_url,
-          args.text_embed_url,
-          args.batch_size,
-          args.train_percent,
-          args.val_percent,
-          args.test_percent,
-          args.num_epochs,
-          args.dp_loss_type,
-          args.clip,
-          args.dp_condition_on_text_encodings,
-          args.dp_timesteps,
-          args.dp_normformer,
-          args.dp_cond_drop_prob,
-          args.dpn_depth,
-          args.dpn_dim_head,
-          args.dpn_heads,
-          args.save_interval,
-          args.save_path,
+    train(image_embed_dim,
+          image_embed_url,
+          text_embed_url,
+          batch_size,
+          train_percent,
+          val_percent,
+          test_percent,
+          num_epochs,
+          dp_loss_type,
+          clip,
+          dp_condition_on_text_encodings,
+          dp_timesteps,
+          dp_normformer,
+          dp_cond_drop_prob,
+          dpn_depth,
+          dpn_dim_head,
+          dpn_heads,
+          save_interval,
+          save_path,
           device,
           RESUME,
           DPRIOR_PATH,
           config,
-          args.wandb_entity,
-          args.wandb_project,
-          args.learning_rate,
-          args.max_grad_norm,
-          args.weight_decay,
-          args.dropout,
-          args.amp)
+          wandb_entity,
+          wandb_project,
+          learning_rate,
+          max_grad_norm,
+          weight_decay,
+          dropout,
+          amp)
 
 if __name__ == "__main__":
-  main()
+    main()