add utils

app.py

@@ -1,43 +1,10 @@
 import os
-import random
-import time
-from langchain.schema.messages import HumanMessage, SystemMessage
 import pinecone
 from langchain.vectorstores import Pinecone
 from langchain.embeddings.openai import OpenAIEmbeddings
 from langchain.chat_models import ChatOpenAI
 import gradio as gr
-
-def retrieve_knowledge(query, k=10, randomize=True):
-    knowledge = [d.page_content.strip() for d in vectorstore.similarity_search(query, k=k)]
-    
-    if randomize:
-        knowledge = random.sample(knowledge, k)
-
-    knowledge = "\n\n\n".join(knowledge)
-
-    return knowledge
-
-
-def generate_workout(query, knowledge):
-    messages = [
-        SystemMessage(content=system_prompt.format(workout_context=knowledge)),
-        HumanMessage(content=query)
-    ]
-
-    response = llm.invoke(messages).content.strip()
-
-    return response
-
-
-def run(gender, muscle_group, equipment, level, duration, k=5, randomize=True):
-    query = f"{duration}-minute {muscle_group} workout for {gender} {level} level {equipment}"
-    knowledge = retrieve_knowledge(query, k, randomize)
-    response = generate_workout(query, knowledge)
-
-    for i in range(len(response)):
-        time.sleep(0.01)
-        yield response[:i+1]
+from utils import *
 
 
 # embedding model

utils.py

@@ -0,0 +1,36 @@
+
+import random
+import time
+from langchain.schema.messages import HumanMessage, SystemMessage
+
+
+def retrieve_knowledge(query, k=10, randomize=True):
+    knowledge = [d.page_content.strip() for d in vectorstore.similarity_search(query, k=k)]
+    
+    if randomize:
+        knowledge = random.sample(knowledge, k)
+
+    knowledge = "\n\n\n".join(knowledge)
+
+    return knowledge
+
+
+def generate_workout(query, knowledge):
+    messages = [
+        SystemMessage(content=system_prompt.format(workout_context=knowledge)),
+        HumanMessage(content=query)
+    ]
+
+    response = llm.invoke(messages).content.strip()
+
+    return response
+
+
+def run(gender, muscle_group, equipment, level, duration, k=5, randomize=True):
+    query = f"{duration}-minute {muscle_group} workout for {gender} {level} level {equipment}"
+    knowledge = retrieve_knowledge(query, k, randomize)
+    response = generate_workout(query, knowledge)
+
+    for i in range(len(response)):
+        time.sleep(0.01)
+        yield response[:i+1]