EdgeCraft RAG UI bug fix

EdgeCraftRAG/edgecraftrag/components/generator.py

@@ -66,6 +66,7 @@ def run(self, chat_request, retrieved_nodes, **kwargs):
             repetition_penalty=chat_request.repetition_penalty,
         )
         self.llm().generate_kwargs = generate_kwargs
+        self.llm().max_new_tokens = chat_request.max_tokens
         if chat_request.stream:
 
             async def stream_generator():
@@ -99,8 +100,10 @@ def run_vllm(self, chat_request, retrieved_nodes, **kwargs):
             max_tokens=chat_request.max_tokens,
             model=model_name,
             top_p=chat_request.top_p,
+            top_k=chat_request.top_k,
             temperature=chat_request.temperature,
             streaming=chat_request.stream,
+            repetition_penalty=chat_request.repetition_penalty,
         )
 
         if chat_request.stream:

EdgeCraftRAG/ui/gradio/default.yaml

@@ -14,7 +14,7 @@ name: "default"
 
 # Node parser
 node_parser: "simple"
-chunk_size: 192
+chunk_size: 400
 chunk_overlap: 48
 
 # Indexer

EdgeCraftRAG/ui/gradio/ecrag_client.py

@@ -67,7 +67,7 @@ def create_update_pipeline(
                 weight=llm_weights,
             ),
         ),
-        retriever=api_schema.RetrieverIn(retriever_type=retriever, retriever_topk=vector_search_top_k),
+        retriever=api_schema.RetrieverIn(retriever_type=retriever, retrieve_topk=vector_search_top_k),
         postprocessor=[
             api_schema.PostProcessorIn(
                 processor_type=postprocessor[0],

EdgeCraftRAG/ui/gradio/ecragui.py

@@ -87,15 +87,11 @@ async def bot(
         top_k,
         repetition_penalty,
         max_tokens,
-        hide_full_prompt,
         docs,
         chunk_size,
         chunk_overlap,
         vector_search_top_k,
-        vector_search_top_n,
-        run_rerank,
-        search_method,
-        score_threshold,
+        vector_rerank_top_n,
     ):
         """Callback function for running chatbot on submit button click.
 
@@ -108,8 +104,21 @@ async def bot(
         repetition_penalty: parameter for penalizing tokens based on how frequently they occur in the text.
         conversation_id: unique conversation identifier.
         """
+        if history[-1][0] == "" or len(history[-1][0]) == 0:
+            yield history[:-1]
+            return
+
         stream_opt = True
-        new_req = {"messages": history[-1][0], "stream": stream_opt, "max_tokens": max_tokens}
+        new_req = {
+            "messages": history[-1][0],
+            "stream": stream_opt,
+            "max_tokens": max_tokens,
+            "top_n": vector_rerank_top_n,
+            "temperature": temperature,
+            "top_p": top_p,
+            "top_k": top_k,
+            "repetition_penalty": repetition_penalty,
+        }
         server_addr = f"http://{MEGA_SERVICE_HOST_IP}:{MEGA_SERVICE_PORT}"
 
         # Async for streaming response
@@ -362,7 +371,7 @@ def get_pipeline_df():
                                     choices=avail_llm_inference_type, label="LLM Inference Type", value="local"
                                 )
 
-                                with gr.Accordion("LLM Configuration", open=True):
+                                with gr.Accordion("LLM Configuration", open=True) as accordion:
                                     u_llm_model_id = gr.Dropdown(
                                         choices=avail_llms,
                                         value=cfg.llm_model_id,
@@ -393,6 +402,12 @@ def get_pipeline_df():
         # RAG Settings Events
         # -------------------
         # Event handlers
+        def update_visibility(selected_value):  # Accept the event argument, even if not used
+            if selected_value == "vllm":
+                return gr.Accordion(visible=False)
+            else:
+                return gr.Accordion(visible=True)
+
         def show_pipeline_detail(evt: gr.SelectData):
             # get selected pipeline id
             # Dataframe: {'headers': '', 'data': [[x00, x01], [x10, x11]}
@@ -470,6 +485,8 @@ def create_update_pipeline(
             return res, get_pipeline_df()
 
         # Events
+        u_llm_infertype.change(update_visibility, inputs=u_llm_infertype, outputs=accordion)
+
         u_pipelines.select(
             show_pipeline_detail,
             inputs=None,
@@ -735,39 +752,9 @@ def delete_file():
                             with gr.Row():
                                 submit = gr.Button("Submit")
                                 clear = gr.Button("Clear")
-                    retriever_argument = gr.Accordion("Retriever Configuration", open=True)
+                    retriever_argument = gr.Accordion("Retriever Configuration", open=False)
                     with retriever_argument:
                         with gr.Row():
-                            with gr.Row():
-                                do_rerank = gr.Checkbox(
-                                    value=True,
-                                    label="Rerank searching result",
-                                    interactive=True,
-                                )
-                                hide_context = gr.Checkbox(
-                                    value=True,
-                                    label="Hide searching result in prompt",
-                                    interactive=True,
-                                )
-                            with gr.Row():
-                                search_method = gr.Dropdown(
-                                    ["similarity_score_threshold", "similarity", "mmr"],
-                                    value=cfg.search_method,
-                                    label="Searching Method",
-                                    info="Method used to search vector store",
-                                    multiselect=False,
-                                    interactive=True,
-                                )
-                            with gr.Row():
-                                score_threshold = gr.Slider(
-                                    0.01,
-                                    0.99,
-                                    value=cfg.score_threshold,
-                                    step=0.01,
-                                    label="Similarity Threshold",
-                                    info="Only working for 'similarity score threshold' method",
-                                    interactive=True,
-                                )
                             with gr.Row():
                                 vector_rerank_top_n = gr.Slider(
                                     1,
@@ -811,15 +798,11 @@ def delete_file():
                 top_k,
                 repetition_penalty,
                 u_max_tokens,
-                hide_context,
                 docs,
                 u_chunk_size,
                 u_chunk_overlap,
                 u_vector_search_top_k,
                 vector_rerank_top_n,
-                do_rerank,
-                search_method,
-                score_threshold,
             ],
             chatbot,
             queue=True,
@@ -833,15 +816,11 @@ def delete_file():
                 top_k,
                 repetition_penalty,
                 u_max_tokens,
-                hide_context,
                 docs,
                 u_chunk_size,
                 u_chunk_overlap,
                 u_vector_search_top_k,
                 vector_rerank_top_n,
-                do_rerank,
-                search_method,
-                score_threshold,
             ],
             chatbot,
             queue=True,