Update app.py

app.py

@@ -49,7 +49,7 @@ for sent in context.split("."):
         
         
         
-corpus_embeddings = np.load('task_embeddings_msmarco-distilbert-base-v4.npy')
+corpus_embeddings = np.load('task_embeddings_updated_msmarco-distilbert-base-v4.npy')
 
 
 
@@ -57,25 +57,11 @@ def find_sentences(query):
     query_embedding = model.encode(query)
     hits = util.semantic_search(query_embedding, corpus_embeddings, top_k=1)
     hit = hits[0][0]
-    message(hit)
     corpus_id = hit['corpus_id']
-    message(corpus_id)
     saved = corpus[corpus_id]
-    message(saved)
-    # Find source document based on sentence index
-    count = 0
-    for idx, c in enumerate(sentence_count):
-        count+=c
-        if (corpus_id > count-1):
-            continue
-        else:
-            doc = corpus[idx]
-            return doc
     return saved
 
 
-
-
 message_history = [{"text":"Let's find out the best task for your use case! Tell me about your use case :)", "is_user":False}]
 
 st.subheader("If you don't know how to build your machine learning product for your use case, Taskmaster is here to help you! 🪄✨")