Update app.py

app.py

@@ -9,7 +9,8 @@ def infer(prompt,
           temperature=0.0, 
           top_p=1.0,
           num_completions=1,
-          seed=42,):
+          seed=42,
+          stop="\n"):
 
     model_name_map = {
         "GPT-JT-6B-v1": "Together-gpt-JT-6B-v1",
@@ -25,7 +26,7 @@ def infer(prompt,
             "model": model_name_map[model_name],
             "prompt": [prompt],
             "request_type": "language-model-inference",
-            "stop": None,
+            "stop": stop.split(";"),
             "best_of": 1,
             "echo": False,
             "seed": int(seed),
@@ -59,7 +60,8 @@ with col1:
     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")
-    num_completions = st.text_input('num_completions', "1")
+    num_completions = st.text_input('num_completions (only the best one will be returend)', "1")
+    stop = st.text_input('stop, split by;', "\n")
     seed = st.text_input('seed', "42")
 
 with col2:
@@ -77,10 +79,9 @@ with col2:
     button_submit = st.button("Submit")
 
     if button_submit:
-        with st.spinner(text="In progress.."):
-            generated_area.markdown("...")
-            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)
+        generated_area.markdown(prompt)
+        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, stop=stop,
+        )
+        generated_area.markdown(prompt + "**" + report_text + "**")