Update README.md

README.md

@@ -41,18 +41,6 @@ model_name = 'vamsibanda/sbert-onnx-all-MiniLM-L6-v2'
 cache_folder = './'
 model_path =  os.path.join(cache_folder, model_name.replace("/", "_"))
 
-def generate_embedding(text):
-    token = tokenizer(text, return_tensors='pt')
-    embeddings = model(input_ids=token['input_ids'], attention_mask=token['attention_mask'])
-    sbert_embeddings = mean_pooling(embeddings, token['attention_mask'])
-    sbert_embeddings = F.normalize(sbert_embeddings, p=2, dim=1)
-    return sbert_embeddings.tolist()[0]
-    
-def mean_pooling(model_output, attention_mask):
-    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
-    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
-    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
-
 def download_onnx_model(model_name, cache_folder, model_path, force_download = False):
     if force_download and os.path.exists(model_path):
        shutil.rmtree(model_path)
@@ -63,8 +51,19 @@ def download_onnx_model(model_name, cache_folder, model_path, force_download = F
                      library_name='sentence-transformers'
                      )
     return
+      
+def mean_pooling(model_output, attention_mask):
+    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
+    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
+    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
+
+def generate_embedding(text):
+    token = tokenizer(text, return_tensors='pt')
+    embedding = model(input_ids=token['input_ids'], attention_mask=token['attention_mask'])
+    embedding = mean_pooling(embedding, token['attention_mask'])
+    embedding = F.normalize(embedding, p=2, dim=1)
+    return embedding.tolist()[0]
                     
-     
 _ = download_onnx_model(model_name, cache_folder, model_path)                
 tokenizer = AutoTokenizer.from_pretrained(model_path)
 model = ORTModelForFeatureExtraction.from_pretrained(model_path, force_download=False)