thanks to christophschuhmann ❤

LICENSE

@@ -0,0 +1,201 @@
+ Apache License
+ Version 2.0, January 2004
+ http://www.apache.org/licenses/
+
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+ 1. Definitions.
+
+ "License" shall mean the terms and conditions for use, reproduction,
+ and distribution as defined by Sections 1 through 9 of this document.
+
+ "Licensor" shall mean the copyright owner or entity authorized by
+ the copyright owner that is granting the License.
+
+ "Legal Entity" shall mean the union of the acting entity and all
+ other entities that control, are controlled by, or are under common
+ control with that entity. For the purposes of this definition,
+ "control" means (i) the power, direct or indirect, to cause the
+ direction or management of such entity, whether by contract or
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
+ outstanding shares, or (iii) beneficial ownership of such entity.
+
+ "You" (or "Your") shall mean an individual or Legal Entity
+ exercising permissions granted by this License.
+
+ "Source" form shall mean the preferred form for making modifications,
+ including but not limited to software source code, documentation
+ source, and configuration files.
+
+ "Object" form shall mean any form resulting from mechanical
+ transformation or translation of a Source form, including but
+ not limited to compiled object code, generated documentation,
+ and conversions to other media types.
+
+ "Work" shall mean the work of authorship, whether in Source or
+ Object form, made available under the License, as indicated by a
+ copyright notice that is included in or attached to the work
+ (an example is provided in the Appendix below).
+
+ "Derivative Works" shall mean any work, whether in Source or Object
+ form, that is based on (or derived from) the Work and for which the
+ editorial revisions, annotations, elaborations, or other modifications
+ represent, as a whole, an original work of authorship. For the purposes
+ of this License, Derivative Works shall not include works that remain
+ separable from, or merely link (or bind by name) to the interfaces of,
+ the Work and Derivative Works thereof.
+
+ "Contribution" shall mean any work of authorship, including
+ the original version of the Work and any modifications or additions
+ to that Work or Derivative Works thereof, that is intentionally
+ submitted to Licensor for inclusion in the Work by the copyright owner
+ or by an individual or Legal Entity authorized to submit on behalf of
+ the copyright owner. For the purposes of this definition, "submitted"
+ means any form of electronic, verbal, or written communication sent
+ to the Licensor or its representatives, including but not limited to
+ communication on electronic mailing lists, source code control systems,
+ and issue tracking systems that are managed by, or on behalf of, the
+ Licensor for the purpose of discussing and improving the Work, but
+ excluding communication that is conspicuously marked or otherwise
+ designated in writing by the copyright owner as "Not a Contribution."
+
+ "Contributor" shall mean Licensor and any individual or Legal Entity
+ on behalf of whom a Contribution has been received by Licensor and
+ subsequently incorporated within the Work.
+
+ 2. Grant of Copyright License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ copyright license to reproduce, prepare Derivative Works of,
+ publicly display, publicly perform, sublicense, and distribute the
+ Work and such Derivative Works in Source or Object form.
+
+ 3. Grant of Patent License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ (except as stated in this section) patent license to make, have made,
+ use, offer to sell, sell, import, and otherwise transfer the Work,
+ where such license applies only to those patent claims licensable
+ by such Contributor that are necessarily infringed by their
+ Contribution(s) alone or by combination of their Contribution(s)
+ with the Work to which such Contribution(s) was submitted. If You
+ institute patent litigation against any entity (including a
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
+ or a Contribution incorporated within the Work constitutes direct
+ or contributory patent infringement, then any patent licenses
+ granted to You under this License for that Work shall terminate
+ as of the date such litigation is filed.
+
+ 4. Redistribution. You may reproduce and distribute copies of the
+ Work or Derivative Works thereof in any medium, with or without
+ modifications, and in Source or Object form, provided that You
+ meet the following conditions:
+
+ (a) You must give any other recipients of the Work or
+ Derivative Works a copy of this License; and
+
+ (b) You must cause any modified files to carry prominent notices
+ stating that You changed the files; and
+
+ (c) You must retain, in the Source form of any Derivative Works
+ that You distribute, all copyright, patent, trademark, and
+ attribution notices from the Source form of the Work,
+ excluding those notices that do not pertain to any part of
+ the Derivative Works; and
+
+ (d) If the Work includes a "NOTICE" text file as part of its
+ distribution, then any Derivative Works that You distribute must
+ include a readable copy of the attribution notices contained
+ within such NOTICE file, excluding those notices that do not
+ pertain to any part of the Derivative Works, in at least one
+ of the following places: within a NOTICE text file distributed
+ as part of the Derivative Works; within the Source form or
+ documentation, if provided along with the Derivative Works; or,
+ within a display generated by the Derivative Works, if and
+ wherever such third-party notices normally appear. The contents
+ of the NOTICE file are for informational purposes only and
+ do not modify the License. You may add Your own attribution
+ notices within Derivative Works that You distribute, alongside
+ or as an addendum to the NOTICE text from the Work, provided
+ that such additional attribution notices cannot be construed
+ as modifying the License.
+
+ You may add Your own copyright statement to Your modifications and
+ may provide additional or different license terms and conditions
+ for use, reproduction, or distribution of Your modifications, or
+ for any such Derivative Works as a whole, provided Your use,
+ reproduction, and distribution of the Work otherwise complies with
+ the conditions stated in this License.
+
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
+ any Contribution intentionally submitted for inclusion in the Work
+ by You to the Licensor shall be under the terms and conditions of
+ this License, without any additional terms or conditions.
+ Notwithstanding the above, nothing herein shall supersede or modify
+ the terms of any separate license agreement you may have executed
+ with Licensor regarding such Contributions.
+
+ 6. Trademarks. This License does not grant permission to use the trade
+ names, trademarks, service marks, or product names of the Licensor,
+ except as required for reasonable and customary use in describing the
+ origin of the Work and reproducing the content of the NOTICE file.
+
+ 7. Disclaimer of Warranty. Unless required by applicable law or
+ agreed to in writing, Licensor provides the Work (and each
+ Contributor provides its Contributions) on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+ implied, including, without limitation, any warranties or conditions
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+ PARTICULAR PURPOSE. You are solely responsible for determining the
+ appropriateness of using or redistributing the Work and assume any
+ risks associated with Your exercise of permissions under this License.
+
+ 8. Limitation of Liability. In no event and under no legal theory,
+ whether in tort (including negligence), contract, or otherwise,
+ unless required by applicable law (such as deliberate and grossly
+ negligent acts) or agreed to in writing, shall any Contributor be
+ liable to You for damages, including any direct, indirect, special,
+ incidental, or consequential damages of any character arising as a
+ result of this License or out of the use or inability to use the
+ Work (including but not limited to damages for loss of goodwill,
+ work stoppage, computer failure or malfunction, or any and all
+ other commercial damages or losses), even if such Contributor
+ has been advised of the possibility of such damages.
+
+ 9. Accepting Warranty or Additional Liability. While redistributing
+ the Work or Derivative Works thereof, You may choose to offer,
+ and charge a fee for, acceptance of support, warranty, indemnity,
+ or other liability obligations and/or rights consistent with this
+ License. However, in accepting such obligations, You may act only
+ on Your own behalf and on Your sole responsibility, not on behalf
+ of any other Contributor, and only if You agree to indemnify,
+ defend, and hold each Contributor harmless for any liability
+ incurred by, or claims asserted against, such Contributor by reason
+ of your accepting any such warranty or additional liability.
+
+ END OF TERMS AND CONDITIONS
+
+ APPENDIX: How to apply the Apache License to your work.
+
+ To apply the Apache License to your work, attach the following
+ boilerplate notice, with the fields enclosed by brackets "[]"
+ replaced with your own identifying information. (Don't include
+ the brackets!) The text should be enclosed in the appropriate
+ comment syntax for the file format. We also recommend that a
+ file or class name and description of purpose be included on the
+ same "printed page" as the copyright notice for easier
+ identification within third-party archives.
+
+ Copyright [yyyy] [name of copyright owner]
+
+ Licensed under the Apache License, Version 2.0 (the "License");
+ you may not use this file except in compliance with the License.
+ You may obtain a copy of the License at
+
+ http://www.apache.org/licenses/LICENSE-2.0
+
+ Unless required by applicable law or agreed to in writing, software
+ distributed under the License is distributed on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ See the License for the specific language governing permissions and
+ limitations under the License.

README.md

@@ -0,0 +1,13 @@
+# CLIP+MLP Aesthetic Score Predictor
+
+Train, use and visualize an aesthetic score predictor ( how much people like on average an image ) based on a simple neural net that takes CLIP embeddings as inputs.
+
+
+Link to the AVA training data ( already prepared) :
+https://drive.google.com/drive/folders/186XiniJup5Rt9FXsHiAGWhgWz-nmCK_r?usp=sharing
+
+
+Visualizations of all images from LAION 5B (english subset with 2.37B images) in 40 buckets with the model sac+logos+ava1-l14-linearMSE.pth:
+http://captions.christoph-schuhmann.de/aesthetic_viz_laion_sac+logos+ava1-l14-linearMSE-en-2.37B.html
+
+

ava+logos-l14-linearMSE.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:390a3aafaf3b37d57148f9b22f30556de38343064b7d915acfa80d3812b4c9ff
+size 3714759

ava+logos-l14-reluMSE.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0af3254c651d55b7ea851429c20f26ec880bb0169805a4df85b814bd7966f3e4
+size 3714887

prepare-data-for-training.py

@@ -0,0 +1,76 @@
+
+# This script prepares the training images and ratings for the training.
+# It assumes that all images are stored as files that PIL can read.
+# It also assumes that the paths to the images files and the average ratings are in a .parquet files that can be read into a dataframe ( df ).
+
+from datasets import load_dataset
+import pandas as pd
+import statistics
+from torch.utils.data import Dataset, DataLoader
+import clip
+import torch
+from PIL import Image, ImageFile
+import numpy as np
+import time
+
+def normalized(a, axis=-1, order=2):
+ import numpy as np # pylint: disable=import-outside-toplevel
+
+ l2 = np.atleast_1d(np.linalg.norm(a, order, axis))
+ l2[l2 == 0] = 1
+ return a / np.expand_dims(l2, axis)
+
+
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+model, preprocess = clip.load("ViT-L/14", device=device)
+
+ 
+f = "trainingdata.parquet"
+df = pd.read_parquet(f) #assumes that the df has the columns IMAGEPATH & AVERAGE_RATING
+
+
+x = []
+y = []
+c= 0
+
+for idx, row in df.iterrows():
+ start = time.time()
+
+ average_rating = float(row.AVERAGE_RATING)
+ print(average_rating)
+ if average_rating <1:
+ continue
+
+ img= row.IMAGEPATH #assumes that the df has the column IMAGEPATH
+ print(img)
+
+ try:
+ image = preprocess(Image.open(img)).unsqueeze(0).to(device)
+ except:
+ continue
+
+ with torch.no_grad():
+ image_features = model.encode_image(image)
+
+ im_emb_arr = image_features.cpu().detach().numpy() 
+ x.append(normalized ( im_emb_arr) ) # all CLIP embeddings are getting normalized. This also has to be done when inputting an embedding later for inference
+ y_ = np.zeros((1, 1))
+ y_[0][0] = average_rating
+ #y_[0][1] = stdev # I initially considered also predicting the standard deviation, but then didn't do it
+
+ y.append(y_)
+
+
+ print(c)
+ c+=1
+
+
+
+
+x = np.vstack(x)
+y = np.vstack(y)
+print(x.shape)
+print(y.shape)
+np.save('x_OpenAI_CLIP_L14_embeddings.npy', x)
+np.save('y_ratings.npy', y)

sac+logos+ava1-l14-linearMSE.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21dd590f3ccdc646f0d53120778b296013b096a035a2718c9cb0d511bff0f1e0
+size 3714759

simple_inference.py

@@ -0,0 +1,122 @@
+import webdataset as wds
+from PIL import Image
+import io
+import matplotlib.pyplot as plt
+import os
+import json
+
+from warnings import filterwarnings
+
+
+# os.environ["CUDA_VISIBLE_DEVICES"] = "0" # choose GPU if you are on a multi GPU server
+import numpy as np
+import torch
+import pytorch_lightning as pl
+import torch.nn as nn
+from torchvision import datasets, transforms
+import tqdm
+
+from os.path import join
+from datasets import load_dataset
+import pandas as pd
+from torch.utils.data import Dataset, DataLoader
+import json
+
+import clip
+
+
+from PIL import Image, ImageFile
+
+
+##### This script will predict the aesthetic score for this image file:
+
+img_path = "test.jpg"
+
+
+
+
+
+# if you changed the MLP architecture during training, change it also here:
+class MLP(pl.LightningModule):
+ def __init__(self, input_size, xcol='emb', ycol='avg_rating'):
+ super().__init__()
+ self.input_size = input_size
+ self.xcol = xcol
+ self.ycol = ycol
+ self.layers = nn.Sequential(
+ nn.Linear(self.input_size, 1024),
+ #nn.ReLU(),
+ nn.Dropout(0.2),
+ nn.Linear(1024, 128),
+ #nn.ReLU(),
+ nn.Dropout(0.2),
+ nn.Linear(128, 64),
+ #nn.ReLU(),
+ nn.Dropout(0.1),
+
+ nn.Linear(64, 16),
+ #nn.ReLU(),
+
+ nn.Linear(16, 1)
+ )
+
+ def forward(self, x):
+ return self.layers(x)
+
+ def training_step(self, batch, batch_idx):
+ x = batch[self.xcol]
+ y = batch[self.ycol].reshape(-1, 1)
+ x_hat = self.layers(x)
+ loss = F.mse_loss(x_hat, y)
+ return loss
+ 
+ def validation_step(self, batch, batch_idx):
+ x = batch[self.xcol]
+ y = batch[self.ycol].reshape(-1, 1)
+ x_hat = self.layers(x)
+ loss = F.mse_loss(x_hat, y)
+ return loss
+
+ def configure_optimizers(self):
+ optimizer = torch.optim.Adam(self.parameters(), lr=1e-3)
+ return optimizer
+
+def normalized(a, axis=-1, order=2):
+ import numpy as np # pylint: disable=import-outside-toplevel
+
+ l2 = np.atleast_1d(np.linalg.norm(a, order, axis))
+ l2[l2 == 0] = 1
+ return a / np.expand_dims(l2, axis)
+
+
+model = MLP(768) # CLIP embedding dim is 768 for CLIP ViT L 14
+
+s = torch.load("sac+logos+ava1-l14-linearMSE.pth") # load the model you trained previously or the model available in this repo
+
+model.load_state_dict(s)
+
+model.to("cuda")
+model.eval()
+
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+model2, preprocess = clip.load("ViT-L/14", device=device) #RN50x64 
+
+
+pil_image = Image.open(img_path)
+
+image = preprocess(pil_image).unsqueeze(0).to(device)
+
+
+
+with torch.no_grad():
+ image_features = model2.encode_image(image)
+
+im_emb_arr = normalized(image_features.cpu().detach().numpy() )
+
+prediction = model(torch.from_numpy(im_emb_arr).to(device).type(torch.cuda.FloatTensor))
+
+print( "Aesthetic score predicted by the model:")
+print( prediction )
+
+

train_predictor.py

@@ -0,0 +1,178 @@
+import os
+# os.environ['CUDA_VISIBLE_DEVICES'] = "0" # in case you are using a multi GPU workstation, choose your GPU here
+import tqdm
+import pytorch_lightning as pl
+import torch
+import torch.nn as nn
+from torch.utils.data import DataLoader
+import torch.nn.functional as F
+import pandas as pd
+from datasets import load_dataset
+from torch.utils.data import TensorDataset, DataLoader
+
+import numpy as np
+
+#define your neural net here:
+
+class MLP(pl.LightningModule):
+ def __init__(self, input_size, xcol='emb', ycol='avg_rating'):
+ super().__init__()
+ self.input_size = input_size
+ self.xcol = xcol
+ self.ycol = ycol
+ self.layers = nn.Sequential(
+ nn.Linear(self.input_size, 1024),
+ #nn.ReLU(),
+ nn.Dropout(0.2),
+ nn.Linear(1024, 128),
+ #nn.ReLU(),
+ nn.Dropout(0.2),
+ nn.Linear(128, 64),
+ #nn.ReLU(),
+ nn.Dropout(0.1),
+
+ nn.Linear(64, 16),
+ #nn.ReLU(),
+
+ nn.Linear(16, 1)
+ )
+
+ def forward(self, x):
+ return self.layers(x)
+
+ def training_step(self, batch, batch_idx):
+ x = batch[self.xcol]
+ y = batch[self.ycol].reshape(-1, 1)
+ x_hat = self.layers(x)
+ loss = F.mse_loss(x_hat, y)
+ return loss
+ 
+ def validation_step(self, batch, batch_idx):
+ x = batch[self.xcol]
+ y = batch[self.ycol].reshape(-1, 1)
+ x_hat = self.layers(x)
+ loss = F.mse_loss(x_hat, y)
+ return loss
+
+ def configure_optimizers(self):
+ optimizer = torch.optim.Adam(self.parameters(), lr=1e-3)
+ return optimizer
+
+
+
+# load the training data 
+
+x = np.load ("/mnt/spirit/ava_x.npy")
+
+y = np.load ("/mnt/spirit/ava_y.npy")
+
+val_percentage = 0.05 # 5% of the trainingdata will be used for validation
+
+train_border = int(x.shape()[0] * (1 - val_percentage) )
+
+train_tensor_x = torch.Tensor(x[:train_border]) # transform to torch tensor
+train_tensor_y = torch.Tensor(y[:train_border])
+
+train_dataset = TensorDataset(train_tensor_x,train_tensor_y) # create your datset
+train_loader = DataLoader(train_dataset, batch_size=256, shuffle=True, num_workers=16) # create your dataloader
+
+
+val_tensor_x = torch.Tensor(x[train_border:]) # transform to torch tensor
+val_tensor_y = torch.Tensor(y[train_border:])
+
+'''
+print(train_tensor_x.size())
+print(val_tensor_x.size())
+print( val_tensor_x.dtype)
+print( val_tensor_x[0].dtype)
+'''
+
+val_dataset = TensorDataset(val_tensor_x,val_tensor_y) # create your datset
+val_loader = DataLoader(val_dataset, batch_size=512, num_workers=16) # create your dataloader
+
+
+
+
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+
+model = MLP(768).to(device) # CLIP embedding dim is 768 for CLIP ViT L 14
+
+optimizer = torch.optim.Adam(model.parameters()) 
+
+# choose the loss you want to optimze for
+criterion = nn.MSELoss()
+criterion2 = nn.L1Loss()
+
+epochs = 50
+
+model.train()
+best_loss =999
+save_name = "linear_predictor_L14_MSE.pth"
+
+
+for epoch in range(epochs):
+ losses = []
+ losses2 = []
+ for batch_num, input_data in enumerate(train_loader):
+ optimizer.zero_grad()
+ x, y = input_data
+ x = x.to(device).float()
+ y = y.to(device)
+
+ output = model(x)
+ loss = criterion(output, y)
+ loss.backward()
+ losses.append(loss.item())
+
+
+ optimizer.step()
+
+ if batch_num % 1000 == 0:
+ print('\tEpoch %d | Batch %d | Loss %6.2f' % (epoch, batch_num, loss.item()))
+ #print(y)
+
+ print('Epoch %d | Loss %6.2f' % (epoch, sum(losses)/len(losses)))
+ losses = []
+ losses2 = []
+ 
+ for batch_num, input_data in enumerate(val_loader):
+ optimizer.zero_grad()
+ x, y = input_data
+ x = x.to(device).float()
+ y = y.to(device)
+
+ output = model(x)
+ loss = criterion(output, y)
+ lossMAE = criterion2(output, y)
+ #loss.backward()
+ losses.append(loss.item())
+ losses2.append(lossMAE.item())
+ #optimizer.step()
+
+ if batch_num % 1000 == 0:
+ print('\tValidation - Epoch %d | Batch %d | MSE Loss %6.2f' % (epoch, batch_num, loss.item()))
+ print('\tValidation - Epoch %d | Batch %d | MAE Loss %6.2f' % (epoch, batch_num, lossMAE.item()))
+ 
+ #print(y)
+
+ print('Validation - Epoch %d | MSE Loss %6.2f' % (epoch, sum(losses)/len(losses)))
+ print('Validation - Epoch %d | MAE Loss %6.2f' % (epoch, sum(losses2)/len(losses2)))
+ if sum(losses)/len(losses) < best_loss:
+ print("Best MAE Val loss so far. Saving model")
+ best_loss = sum(losses)/len(losses)
+ print( best_loss ) 
+
+ torch.save(model.state_dict(), save_name )
+
+
+torch.save(model.state_dict(), save_name)
+
+print( best_loss ) 
+
+print("training done")
+# inferece test with dummy samples from the val set, sanity check
+print( "inferece test with dummy samples from the val set, sanity check")
+model.eval()
+output = model(x[:5].to(device))
+print(output.size())
+print(output)

visulaize_100k_from_LAION400M.py

@@ -0,0 +1,175 @@
+import webdataset as wds
+from PIL import Image
+import io
+import matplotlib.pyplot as plt
+import os
+import json
+
+from warnings import filterwarnings
+
+
+# os.environ["CUDA_VISIBLE_DEVICES"] = "0" # choose GPU if you are on a multi GPU server
+import numpy as np
+import torch
+import pytorch_lightning as pl
+import torch.nn as nn
+from torchvision import datasets, transforms
+import tqdm
+
+from os.path import join
+from datasets import load_dataset
+import pandas as pd
+from torch.utils.data import Dataset, DataLoader
+import json
+
+import clip
+#import open_clip
+
+from PIL import Image, ImageFile
+
+
+# if you changed the MLP architecture during training, change it also here:
+
+class MLP(pl.LightningModule):
+ def __init__(self, input_size, xcol='emb', ycol='avg_rating'):
+ super().__init__()
+ self.input_size = input_size
+ self.xcol = xcol
+ self.ycol = ycol
+ self.layers = nn.Sequential(
+ nn.Linear(self.input_size, 1024),
+ #nn.ReLU(),
+ nn.Dropout(0.2),
+ nn.Linear(1024, 128),
+ #nn.ReLU(),
+ nn.Dropout(0.2),
+ nn.Linear(128, 64),
+ #nn.ReLU(),
+ nn.Dropout(0.1),
+
+ nn.Linear(64, 16),
+ #nn.ReLU(),
+
+ nn.Linear(16, 1)
+ )
+
+ def forward(self, x):
+ return self.layers(x)
+
+ def training_step(self, batch, batch_idx):
+ x = batch[self.xcol]
+ y = batch[self.ycol].reshape(-1, 1)
+ x_hat = self.layers(x)
+ loss = F.mse_loss(x_hat, y)
+ return loss
+ 
+ def validation_step(self, batch, batch_idx):
+ x = batch[self.xcol]
+ y = batch[self.ycol].reshape(-1, 1)
+ x_hat = self.layers(x)
+ loss = F.mse_loss(x_hat, y)
+ return loss
+
+ def configure_optimizers(self):
+ optimizer = torch.optim.Adam(self.parameters(), lr=1e-3)
+ return optimizer
+
+def normalized(a, axis=-1, order=2):
+ import numpy as np # pylint: disable=import-outside-toplevel
+
+ l2 = np.atleast_1d(np.linalg.norm(a, order, axis))
+ l2[l2 == 0] = 1
+ return a / np.expand_dims(l2, axis)
+
+
+model = MLP(768) # CLIP embedding dim is 768 for CLIP ViT L 14
+
+s = torch.load("ava+logos-l14-linearMSE.pth") # load the model you trained previously or the model available in this repo
+
+model.load_state_dict(s)
+
+
+model.to("cuda")
+model.eval()
+
+
+
+
+
+device = "cuda" if torch.cuda.is_available() else "cpu"
+model2, preprocess = clip.load("ViT-L/14", device=device) #RN50x64 
+
+
+
+c=0
+urls= []
+predictions=[]
+
+# this will run inference over 10 webdataset tar files from LAION 400M and sort them into 20 categories
+# you can DL LAION 400M and convert it to wds tar files with img2dataset ( https://github.com/rom1504/img2dataset ) 
+
+
+for j in range(10):
+ if j<10:
+ # change the path to the tar files accordingly
+ dataset = wds.WebDataset("pipe:aws s3 cp s3://s-datasets/laion400m/laion400m-dat-release/0000"+str(j)+".tar -") #"pipe:aws s3 cp s3://s-datasets/laion400m/laion400m-dat-release/00625.tar -")
+ else:
+ dataset = wds.WebDataset("pipe:aws s3 cp s3://s-datasets/laion400m/laion400m-dat-release/000"+str(j)+".tar -") #"pipe:aws s3 cp s3://s-datasets/laion400m/laion400m-dat-release/00625.tar -")
+
+
+ for i, d in enumerate(dataset):
+ print(c)
+
+ metadata= json.loads(d['json']) 
+
+ pil_image = Image.open(io.BytesIO(d['jpg']))
+ c=c+1
+ try:
+ image = preprocess(pil_image).unsqueeze(0).to(device)
+
+ except:
+ continue
+
+ with torch.no_grad():
+ image_features = model2.encode_image(image)
+
+
+ im_emb_arr = normalized(image_features.cpu().detach().numpy() )
+
+ prediction = model(torch.from_numpy(im_emb_arr).to(device).type(torch.cuda.FloatTensor))
+ urls.append(metadata["url"])
+ predictions.append(prediction)
+
+
+df = pd.DataFrame(list(zip(urls, predictions)),
+ columns =['filepath', 'prediction'])
+
+
+buckets = [(i, i+1) for i in range(20)]
+
+
+html= "<h1>Aesthetic subsets in LAION 100k samples</h1>"
+
+i =0
+for [a,b] in buckets:
+ a = a/2
+ b = b/2
+ total_part = df[( (df["prediction"] ) *1>= a) & ( (df["prediction"] ) *1 <= b)]
+ print(a,b)
+ print(len(total_part) )
+ count_part = len(total_part) / len(df) * 100
+ estimated =int ( len(total_part) )
+ part = total_part[:50]
+
+ html+=f"<h2>In bucket {a} - {b} there is {count_part:.2f}% samples:{estimated:.2f} </h2> <div>"
+ for filepath in part["filepath"]:
+ html+='<img src="'+filepath +'" height="200" />'
+
+
+ html+="</div>"
+ i+=1
+ print(i)
+with open("./aesthetic_viz_laion_ava+logos_L14_100k-linearMSE.html", "w") as f:
+ f.write(html)
+ 
+