saswat call

.gitignore

@@ -2,3 +2,4 @@
 *Logs*
 *word_vectors_cache*
 *dataset*
+*testfiles*

embeddings.py

+import torch
+from logger import Logger
+LOGGER = Logger().logger()
+LOGGER.info("Running the model cuda_available = "+str(torch.cuda.is_available()))
+from feature_maps import FeatureMaps
+from word_vectors import WordVectors
+from preprocess import Preprocessor
+from Utils import device
+#Author: @meetdoshi
+
+class Embeddings():
+    def __init__(self) -> None:
+        self.word_vectors = WordVectors()
+        self.feature_maps = FeatureMaps()
+        return
+
+    def get_embeddings(self,word,img):
+        '''
+        Function to get embeddings for a word
+        @params
+        word: str
+        img: numpy image f64
+        '''
+        word_vector = self.word_vectors.get_vector(word)
+        feature_maps = self.feature_maps.get_fmaps_content(img)
+        embedding = torch.cat((word_vector,feature_maps),dim=0)
+        return embedding
+
+    
+if __name__ == "__main__":
+    img = Preprocessor().process('./testfiles/style.jpg')
+    emb = Embeddings()
+    print(emb.get_embeddings('hello',img).shape)
\ No newline at end of file

feature_maps.py

@@ -2,6 +2,7 @@ import torch
 from logger import Logger
 LOGGER = Logger().logger()
 LOGGER.info("Running the model cuda_available = "+str(torch.cuda.is_available()))
+from Utils import device
 #Author: @meetdoshi
 class FeatureMaps():
     def __init__(self,arch="vgg19"):
@@ -13,6 +14,7 @@ class FeatureMaps():
         super()
         try:
             self.model = torch.hub.load('pytorch/vision:v0.10.0',arch,pretrained=True)
+            self.model = self.model.to(device)
         except Exception as e:
             LOGGER.error("Could not load model" + str(e))
         return

preprocess.py

@@ -6,10 +6,53 @@ from PIL import Image
 import numpy as np
 LOGGER = Logger().logger()
 #Author: @meetdoshi
+from Utils import device
 class Preprocessor:
-    def __init__(self) -> None:
-        pass
+    @staticmethod
+    def load_image(path):
+        '''
+        Function to load image
+        @params
+        path: os.path
+        '''
+        img = Image.open(path)
+        return img
 
+    @staticmethod
+    def subtract_mean(img):
+        '''
+        Function to subtract mean values of RGB channels computed over whole ImageNet dataset
+        @params
+        img: 3d numpy array
+        '''
+        mean = np.reshape([103.939, 116.779, 123.68],(1,1,3))#b,g,r
+        return img-mean
+
+    @staticmethod
+    def reshape_img(img):
+        '''
+        Function to reshpae image in 224x224xnum_of_channels shape
+        @params
+        img: 3d numpy array
+        '''
+        #loader = transforms.Compose([transforms.ToTensor(),transforms.Resize([224,224]),transforms.Normalize(mean=[0.485, 0.456, 0.406],std=[0.229, 0.224, 0.225],),])
+        loader = transforms.Compose([transforms.ToTensor(),transforms.Resize([224,224])])
+        img = loader(img).unsqueeze(0)
+        #assert img.shape == (1,3,224,224)
+        return img.to(device,torch.float)
+
+
+    @staticmethod
+    def process(path):
+        '''
+        Function to preprocess the image
+        @params
+        path: os.path
+        '''
+        img = Preprocessor.load_image(path)
+        img = Preprocessor.reshape_img(img)
+        #img = Preprocessor.subtract_mean(img)
+        return img
 
 '''
 if __name__=="__main__":

word_vectors.py

@@ -2,6 +2,7 @@ import torch
 from logger import Logger
 LOGGER = Logger().logger()
 from torchnlp.word_to_vector import GloVe, FastText, BPEmb, CharNGram
+from Utils import device
 #Author: @meetdoshi
 class WordVectors():
     def __init__(self,model='fasttext'):
@@ -41,7 +42,7 @@ class WordVectors():
             LOGGER.error("Model not found")
         return self.vectors[word]
 
-    
+''' 
 if __name__ == "__main__":
     wv = WordVectors()
     print(wv.get_vector('ababasba').shape)
@@ -51,3 +52,4 @@ if __name__ == "__main__":
     print(wv.get_vector('hbasbasbsello').shape)
     wv = WordVectors(model='chargram')
     print(wv.get_vector('hebasallo').shape)
+'''
\ No newline at end of file