Update README.md

README.md

@@ -1,4 +1,5 @@
 ---
+inference: false
 pipeline_tag: sentence-similarity
 tags:
 - sentence-transformers
@@ -91,22 +92,22 @@ import torch
 model = SentenceTransformer('yseop/roberta-base-finance-hypernym-identification')
 # Our corpus containing the list of hypernym labels
 hypernyms = ['Bonds',
-			'Forward',
-			'Funds',
-			'Future',
-			'MMIs',
-			'Option',
-			'Stocks',
-			'Swap',
-			'Equity Index',
-			'Credit Index',
-			'Securities restrictions',
-			'Parametric schedules',
-			'Debt pricing and yields',
-			'Credit Events',
-			'Stock Corporation',
-			'Central Securities Depository',
-			'Regulatory Agency']
+\t\t\t'Forward',
+\t\t\t'Funds',
+\t\t\t'Future',
+\t\t\t'MMIs',
+\t\t\t'Option',
+\t\t\t'Stocks',
+\t\t\t'Swap',
+\t\t\t'Equity Index',
+\t\t\t'Credit Index',
+\t\t\t'Securities restrictions',
+\t\t\t'Parametric schedules',
+\t\t\t'Debt pricing and yields',
+\t\t\t'Credit Events',
+\t\t\t'Stock Corporation',
+\t\t\t'Central Securities Depository',
+\t\t\t'Regulatory Agency']
 hypernym_embeddings = model.encode(hypernyms, convert_to_tensor=True)
 # Query sentences are financial terms to match to the predefined labels
 queries = ['Convertible bond', 'weighted average coupon', 'Restriction 144-A']
@@ -117,9 +118,14 @@ for query in queries:
     # We use cosine-similarity and torch.topk to find the highest 5 scores
     cos_scores = util.pytorch_cos_sim(query_embedding, hypernym_embeddings)[0]
     top_results = torch.topk(cos_scores, k=top_k)
-    print("\n\n======================\n\n")
+    print("\
+\
+======================\
+\
+")
     print("Query:", query)
-    print("\nTop 5 most similar hypernyms:")
+    print("\
+Top 5 most similar hypernyms:")
     for score, idx in zip(top_results[0], top_results[1]):
         print(hypernyms[idx], "(Score: {:.4f})".format(score))
 ```