Update app.py

app.py

@@ -2,7 +2,13 @@ import streamlit as st
 import requests
 import time
 
-def infer(prompt, model_name, max_new_tokens=10, temperature=0.0, top_p=1.0):
+def infer(prompt, 
+          model_name, 
+          max_new_tokens=10, 
+          temperature=0.0, 
+          top_p=1.0,
+          num_completions=1,
+          seed=42,):
 
     model_name_map = {
         "GPT-JT-6B-v1": "Together-gpt-JT-6B-v1",
@@ -12,7 +18,7 @@ def infer(prompt, model_name, max_new_tokens=10, temperature=0.0, top_p=1.0):
         "type": "general",
         "payload": {
             "max_tokens": int(max_new_tokens),
-            "n": 1,
+            "n": int(num_completions),
             "temperature": float(temperature),
             "top_p": float(top_p),
             "model": model_name_map[model_name],
@@ -21,7 +27,7 @@ def infer(prompt, model_name, max_new_tokens=10, temperature=0.0, top_p=1.0):
             "stop": None,
             "best_of": 1,
             "echo": False,
-            "seed": 42,
+            "seed": int(seed),
             "prompt_embedding": False,
         },
         "returned_payload": {},
@@ -47,6 +53,14 @@ st.title("TOMA Application")
 
 col1, col2 = st.columns([1, 3])
 
+with col1:
+    model_name = st.selectbox("Model", ["GPT-JT-6B-v1"])
+    max_new_tokens = st.text_input('Max new tokens', "10")
+    temperature = st.text_input('temperature', "1.0")
+    top_p = st.text_input('top_p', "1.0")
+    num_completions = st.text_input('num_completions', "1")
+    seed = st.text_input('seed', "42")
+
 with col2:
     s_example = "Please answer the following question:\n\nQuestion: Where is Zurich?\nAnswer:"
     prompt = st.text_area(
@@ -63,11 +77,8 @@ with col2:
 
     if button_submit:
     with st.spinner(text="In progress.."):
-        report_text = infer(prompt, model_name=model_name, max_new_tokens=max_new_tokens, temperature=temperature, top_p=top_p)
+        report_text = infer(
+            prompt, model_name=model_name, max_new_tokens=max_new_tokens, temperature=temperature, top_p=top_p,
+            num_completions=num_completions, seed=seed,
+        )
         generated_area.markdown(report_text)
-
-with col1:
-    model_name = st.selectbox("Model", ["GPT-JT-6B-v1"])
-    max_new_tokens = st.text_input('Max new tokens', "10")
-    temperature = st.text_input('temperature', "0.0")
-    top_p = st.text_input('top_p', "1.0")