updated

Dockerfile

@@ -13,6 +13,9 @@ RUN useradd -m -u 1000 user
 USER user
 ENV HOME /home/user
 ENV PATH $HOME/.local/bin:$PATH
+ENV TIKTOKEN_CACHE_DIR $HOME/.cache/tiktoken
+
+RUN mkdir -p $HOME/.cache/tiktoken
 
 WORKDIR $HOME
 RUN mkdir app

agent.py

@@ -12,6 +12,7 @@ load_dotenv(override=True)
 from vectara_agentic.agent import Agent
 from vectara_agentic.agent_config import AgentConfig
 from vectara_agentic.tools import ToolsFactory, VectaraToolFactory
+from vectara_agentic.types import ModelProvider, AgentType
 
 class AgentTools:
     def __init__(self, _cfg, agent_config):
@@ -21,13 +22,12 @@ class AgentTools:
 
 
     def get_tools(self):
-        class QueryElectricCars(BaseModel):
-            query: str = Field(description="The user query.")
-
-        vec_factory_1 = VectaraToolFactory(vectara_api_key=self.cfg.api_keys[0],
-                                            vectara_corpus_key=self.cfg.corpus_keys[0])
+        vec_factory_1 = VectaraToolFactory(
+            vectara_api_key=self.cfg.api_keys[0],
+            vectara_corpus_key=self.cfg.corpus_keys[0]
+        )
         
-        summarizer = 'vectara-experimental-summary-ext-2023-12-11-med-omni'
+        summarizer = 'vectara-summary-table-md-query-ext-jan-2025-gpt-4o'
 
         ask_vehicles = vec_factory_1.create_rag_tool(
             tool_name = "ask_vehicles",
@@ -35,12 +35,11 @@ class AgentTools:
             Given a user query, 
             returns a response to a user question about electric vehicles.
             """,
-            tool_args_schema = QueryElectricCars,
             reranker = "chain", rerank_k = 100,
             rerank_chain = [
                 {
                     "type": "slingshot",
-                    "cutoff": 0.2
+                    "cutoff": 0.3
                 },
                 {
                     "type": "mmr",
@@ -48,24 +47,29 @@ class AgentTools:
                 }
             ],
             n_sentences_before = 2, n_sentences_after = 2, lambda_val = 0.005,
-            summary_num_results = 5,
+            summary_num_results = 10,
             vectara_summarizer = summarizer,
-            include_citations = False,
+            max_tokens = 4096, max_response_chars = 8192,
+            include_citations = True,
+            save_history = True,
+            verbose = False,        
         )
 
-        vec_factory_2 = VectaraToolFactory(vectara_api_key=self.cfg.api_keys[1],
-                                        vectara_corpus_key=self.cfg.corpus_keys[1])
+        vec_factory_2 = VectaraToolFactory(
+            vectara_api_key=self.cfg.api_keys[1],
+            vectara_corpus_key=self.cfg.corpus_keys[1]
+        )
         
 
         class QueryEVLaws(BaseModel):
-            query: str = Field(description="The user query")
-            state: Optional[str] = Field(default=None,
-                                        description="The two digit state code. Optional.",
-                                        examples=['CA', 'US', 'WA'])
-            policy_type: Optional[str] = Field(default=None,
-                                            description="The type of policy. Optional",
-                                            examples = ['Laws and Regulations', 'State Incentives', 'Incentives', 'Utility / Private Incentives', 'Programs'])
-            
+            state: Optional[str] = Field(
+                default=None, description="The two digit state code. Optional.",
+                examples=['CA', 'US', 'WA']
+            )
+            policy_type: Optional[str] = Field(
+                default=None, description="The type of policy. Optional",
+                examples = ['Laws and Regulations', 'State Incentives', 'Incentives', 'Utility / Private Incentives', 'Programs']
+            )
 
         ask_policies = vec_factory_2.create_rag_tool(
             tool_name = "ask_policies",
@@ -79,7 +83,7 @@ class AgentTools:
             rerank_chain = [
                 {
                     "type": "slingshot",
-                    "cutoff": 0.2
+                    "cutoff": 0.3
                 },
                 {
                     "type": "mmr",
@@ -88,12 +92,15 @@ class AgentTools:
             ],
             n_sentences_before = 2, n_sentences_after = 2, lambda_val = 0.005,
             summary_num_results = 10,
+            max_tokens = 4096,
+            max_response_chars = 8192,
             vectara_summarizer = summarizer,
             include_citations = False,
+            save_history = True,
+            verbose = False,
         )
 
         tools_factory = ToolsFactory()
-
         db_tools = tools_factory.database_tools(
                     tool_name_prefix = "ev",
                     content_description = 'Electric Vehicles in the state of Washington and other population information',
@@ -111,8 +118,26 @@ def initialize_agent(_cfg, agent_progress_callback=None):
     - You are a helpful research assistant, with expertise in electric vehicles, in conversation with a user.
     - Never discuss politics, and always respond politely.
     """
+    agent_config = AgentConfig(
+        agent_type = os.getenv("VECTARA_AGENTIC_AGENT_TYPE", AgentType.OPENAI.value),
+        main_llm_provider = os.getenv("VECTARA_AGENTIC_MAIN_LLM_PROVIDER", ModelProvider.OPENAI.value),
+        main_llm_model_name = os.getenv("VECTARA_AGENTIC_MAIN_MODEL_NAME", ""),
+        tool_llm_provider = os.getenv("VECTARA_AGENTIC_TOOL_LLM_PROVIDER", ModelProvider.OPENAI.value),
+        tool_llm_model_name = os.getenv("VECTARA_AGENTIC_TOOL_MODEL_NAME", ""),
+        observer = os.getenv("VECTARA_AGENTIC_OBSERVER_TYPE", "NO_OBSERVER")
+    )
+    fallback_agent_config = AgentConfig(
+        agent_type = os.getenv("VECTARA_AGENTIC_FALLBACK_AGENT_TYPE", AgentType.OPENAI.value),
+        main_llm_provider = os.getenv("VECTARA_AGENTIC_FALLBACK_MAIN_LLM_PROVIDER", ModelProvider.OPENAI.value),
+        main_llm_model_name = os.getenv("VECTARA_AGENTIC_FALLBACK_MAIN_MODEL_NAME", ""),
+        tool_llm_provider = os.getenv("VECTARA_AGENTIC_FALLBACK_TOOL_LLM_PROVIDER", ModelProvider.OPENAI.value),
+        tool_llm_model_name = os.getenv("VECTARA_AGENTIC_FALLBACK_TOOL_MODEL_NAME", ""),
+        observer = os.getenv("VECTARA_AGENTIC_OBSERVER_TYPE", "NO_OBSERVER")
+    )
 
     agent = Agent(
+        agent_config=agent_config,
+        fallback_agent_config=fallback_agent_config,
         tools=AgentTools(_cfg, AgentConfig()).get_tools(),
         topic="Electric vehicles in the United States",
         custom_instructions=electric_vehicle_bot_instructions,

requirements.txt

@@ -1,10 +1,10 @@
 omegaconf==2.3.0
 python-dotenv==1.0.1
-streamlit==1.43.2
+streamlit==1.45.0
 streamlit-feedback==0.1.3
 langdetect==1.0.9
 langcodes==3.4.0
 datasets==2.19.2
 uuid==1.30
-vectara-agentic==0.2.9
+vectara-agentic==0.2.15
 torch==2.6.0

st_app.py

@@ -131,7 +131,7 @@ async def launch_bot():
     if st.session_state.prompt:
         with st.chat_message("assistant", avatar='🤖'):
             st.session_state.status = st.status('Processing...', expanded=False)
-            response = st.session_state.agent.chat(st.session_state.prompt)
+            response = await st.session_state.agent.achat(st.session_state.prompt)
             res = escape_dollars_outside_latex(response.response)
             message = {"role": "assistant", "content": res, "avatar": '🤖'}
             st.session_state.messages.append(message)