add toggle

app.py

@@ -43,7 +43,7 @@ def add_text(history, text):
 
 
 def bot(history, api_kind,
-            num_docs, model_kind, sub_vector_size, chunk_size, splitter_type):
+            num_docs, model_kind, sub_vector_size, chunk_size, splitter_type, all_at_once):
     query = history[-1][0]
 
     if not query:
@@ -72,22 +72,23 @@ def bot(history, api_kind,
     
 
     history[-1][1] = ""
-    # for character in generate_fn(prompt, history[:-1]):
-    #     history[-1][1] = character
-    #     yield history, prompt_html
-    
-    for model_name, doc, size, sub_vector in combinations:
-        documents_i = retrieve(query, int(num_docs), model_name, sub_vector, size, doc)
-        prompt_i = template.render(documents=documents_i, query=query)
-        prompt_html = template_html.render(documents=documents, query=query)
-        
-        hist_chunk = ""
-        prev_hist = history[-1][1] + f"\nmodel {model_name}, splitter {doc}, size {size}, sub vector {sub_vector}\n"
-        for character in generate_fn(prompt_i, history[:-1]):
-            hist_chunk = character
-            history[-1][1] = prev_hist + hist_chunk
+    if all_at_once:
+        for model_name, doc, size, sub_vector in combinations:
+            documents_i = retrieve(query, int(num_docs), model_name, sub_vector, size, doc)
+            prompt_i = template.render(documents=documents_i, query=query)
+            prompt_html = template_html.render(documents=documents, query=query)
+            
+            hist_chunk = ""
+            prev_hist = history[-1][1] + f"\nmodel {model_name}, splitter {doc}, size {size}, sub vector {sub_vector}\n"
+            for character in generate_fn(prompt_i, history[:-1]):
+                hist_chunk = character
+                history[-1][1] = prev_hist + hist_chunk
+                yield history, prompt_html
+    else:
+        for character in generate_fn(prompt, history[:-1]):
+            history[-1][1] = character
             yield history, prompt_html
-
+    
 
 with gr.Blocks() as demo:
     chatbot = gr.Chatbot(
@@ -114,6 +115,7 @@ with gr.Blocks() as demo:
         num_docs = gr.Slider(1, 20, label="number of docs", step=1, value=4)
         model_kind = gr.Radio(choices=emb_models, value="bge", label="embedding model")
         sub_vector_size = gr.Radio(choices=sub_vectors, value="32", label="sub-vector size")
+        all_at_once = gr.Checkbox(value=False, label="Run all at once")
     with gr.Row():
         api_kind = gr.Radio(choices=["HuggingFace", "OpenAI"], value="HuggingFace", label="Chat model engine")
         chunk_size = gr.Radio(choices=chunk_sizes, value="2000", label="chunk size")
@@ -123,7 +125,7 @@ with gr.Blocks() as demo:
     # Turn off interactivity while generating if you click
     txt_msg = txt_btn.click(add_text, [chatbot, txt], [chatbot, txt], queue=False).then(
             bot, [chatbot, api_kind,
-            num_docs, model_kind, sub_vector_size, chunk_size, splitter_type
+            num_docs, model_kind, sub_vector_size, chunk_size, splitter_type, all_at_once
             ], [chatbot, prompt_html])
 
     # Turn it back on