Update handler.py

handler.py

@@ -8,6 +8,20 @@ import time
 import os
 import torch
 
+def max_pooling(model_output):    
+    # Get dimensions    
+    _, Z, Y = model_output.shape    
+    # Initialize an empty list with length Y (384 in your case)    
+    output_array = [0] * Y    
+    # Loop over secondary arrays (Z)    
+    for i in range(Z):        
+        # Loop over values in innermost arrays (Y)        
+        for j in range(Y):            
+            # If value is greater than current max, update max            
+            if model_output[0][i][j] > output_array[j]:                
+                output_array[j] = model_output[0][i][j]    
+    return output_array
+    
 #Mean Pooling - Take attention mask into account for correct averaging
 def mean_pooling(model_output, attention_mask):
     token_embeddings = model_output[0] #First element of model_output contains all token embeddings