Update app.py

app.py

@@ -1,6 +1,6 @@
 import gradio as gr
 from transformers import AutoTokenizer, AutoModel
-from openai import OpenAI
+import openai
 import os
 import numpy as np
 from sklearn.metrics.pairwise import cosine_similarity
@@ -12,7 +12,7 @@ bi_model = AutoModel.from_pretrained(bi_encoder_model_name)
 
 # Set up OpenAI client
 api_key = os.getenv('OPENAI_API_KEY')
-client = OpenAI(api_key=api_key)
+openai.api_key = api_key
 
 # Define a system message to introduce Exos
 system_message = "You are Exos, a helpful assistant specializing in Exoplanet research. Provide detailed and accurate responses related to Exoplanet research."
@@ -30,33 +30,33 @@ def retrieve_relevant_context(user_input, context_texts):
     most_relevant_idx = np.argmax(similarities)
     return context_texts[most_relevant_idx]
 
-def generate_response(user_input, relevant_context=""):
+def generate_response(user_input, relevant_context="", max_tokens=150, temperature=0.7, top_p=0.9, frequency_penalty=0.5, presence_penalty=0.0):
     if relevant_context:
         combined_input = f"Context: {relevant_context}\nQuestion: {user_input}\nAnswer:"
     else:
         combined_input = f"Question: {user_input}\nAnswer:"
     
-    response = client.chat.completions.create(
+    response = openai.ChatCompletion.create(
         model="gpt-4",
         messages=[
             {"role": "system", "content": system_message},
             {"role": "user", "content": combined_input}
         ],
-        max_tokens=150,
-        temperature=0.5,
-        top_p=0.9,
-        frequency_penalty=0.5,
-        presence_penalty=0.0
+        max_tokens=max_tokens,
+        temperature=temperature,
+        top_p=top_p,
+        frequency_penalty=frequency_penalty,
+        presence_penalty=presence_penalty
     )
     return response.choices[0].message.content.strip()
 
-def chatbot(user_input, context="", use_encoder=False):
+def chatbot(user_input, context="", use_encoder=False, max_tokens=150, temperature=0.7, top_p=0.9, frequency_penalty=0.5, presence_penalty=0.0):
     if use_encoder and context:
         context_texts = context.split("\n")
         relevant_context = retrieve_relevant_context(user_input, context_texts)
     else:
         relevant_context = ""
-    response = generate_response(user_input, relevant_context)
+    response = generate_response(user_input, relevant_context, max_tokens, temperature, top_p, frequency_penalty, presence_penalty)
     return response
 
 # Create the Gradio interface
@@ -65,7 +65,12 @@ iface = gr.Interface(
     inputs=[
         gr.Textbox(lines=2, placeholder="Enter your message here...", label="Your Question"),
         gr.Textbox(lines=5, placeholder="Enter context here, separated by new lines...", label="Context (Optional)"),
-        gr.Checkbox(label="Use Bi-Encoder for Context")
+        gr.Checkbox(label="Use Bi-Encoder for Context"),
+        gr.Slider(50, 500, value=150, step=10, label="Max Tokens"),
+        gr.Slider(0.0, 1.0, value=0.7, step=0.1, label="Temperature"),
+        gr.Slider(0.0, 1.0, value=0.9, step=0.1, label="Top-p"),
+        gr.Slider(0.0, 1.0, value=0.5, step=0.1, label="Frequency Penalty"),
+        gr.Slider(0.0, 1.0, value=0.0, step=0.1, label="Presence Penalty")
     ],
     outputs="text",
     title="Exos - Your Exoplanet Research Assistant",
@@ -85,3 +90,4 @@ iface.launch(share=True)
 
 
 
+