nvidia nim update

app.py

@@ -28,6 +28,7 @@ from langchain_community.embeddings import HuggingFaceBgeEmbeddings
 from langchain_community.vectorstores.utils import DistanceStrategy
 from langchain_openai import ChatOpenAI
 from langchain_anthropic import ChatAnthropic
+from langchain_nvidia_ai_endpoints import ChatNVIDIA
 from langchain_pinecone import PineconeVectorStore
 from pinecone import Pinecone
 import streamlit as st
@@ -57,7 +58,6 @@ CONGRESS_GOV_TYPE_MAP = {
 }
 OPENAI_CHAT_MODELS = [
     "gpt-3.5-turbo-0125",
-#    "gpt-4-0125-preview",
     "gpt-4o",
 ]
 ANTHROPIC_CHAT_MODELS = [
@@ -65,7 +65,20 @@ ANTHROPIC_CHAT_MODELS = [
     "claude-3-sonnet-20240229",
     "claude-3-haiku-20240307",
 ]
-CHAT_MODELS = OPENAI_CHAT_MODELS + ANTHROPIC_CHAT_MODELS
+NVIDIA_NIM_CHAT_MODELS = [
+    "microsoft/phi-3-mini-128k-instruct",
+    "google/gemma-7b",
+    "meta/llama3-8b-instruct",
+    "meta/llama3-70b-instruct",
+    "mistralai/mixtral-8x22b-instruct-v0.1",
+]
+CHAT_MODELS = OPENAI_CHAT_MODELS + ANTHROPIC_CHAT_MODELS + NVIDIA_NIM_CHAT_MODELS
+
+PROVIDER_MODELS = {
+    "OpenAI": OPENAI_CHAT_MODELS,
+    "Anthropic": ANTHROPIC_CHAT_MODELS,
+    "Nvidia NIM": NVIDIA_NIM_CHAT_MODELS,
+}
 
 
 def get_sponsor_url(bioguide_id: str) -> str:
@@ -278,9 +291,10 @@ def render_sidebar():
     st.checkbox("add legis urls in answer", value=True, key="response_add_legis_urls")
 
     with st.expander("Generative Config"):
-        st.selectbox(label="model name", options=CHAT_MODELS, key="model_name")
+        st.selectbox(label="provider", options=PROVIDER_MODELS.keys(), key="provider")
+        st.selectbox(label="model name", options=PROVIDER_MODELS[SS["provider"]], key="model_name")
         st.slider(
-            "temperature", min_value=0.0, max_value=2.0, value=0.0, key="temperature"
+            "temperature", min_value=0.0, max_value=2.0, value=0.01, key="temperature"
         )
         st.slider(
             "max_output_tokens", min_value=512, max_value=1024, key="max_output_tokens"
@@ -315,7 +329,7 @@ def render_query_rag_tab():
 
     render_example_queries()
 
-    QUERY_TEMPLATE = """Use the following excerpts from US congressional legislation to respond to the user's query. The excerpts are formatted as a JSON list. Each JSON object has "legis_id", "title", "introduced_date", "sponsor", and "snippets" keys. If a snippet is useful in writing part of your response, then cite the "legis_id", "title", "introduced_date", and "sponsor" in the response. If you don't know how to respond, just tell the user.
+    QUERY_TEMPLATE = """You are an expert legislative analyst. Use the following excerpts from US congressional legislation to respond to the user's query. The excerpts are formatted as a JSON list. Each JSON object has "legis_id", "title", "introduced_date", "sponsor", and "snippets" keys. If a snippet is useful in writing part of your response, then cite the "legis_id", "title", "introduced_date", and "sponsor" in the response. If you don't know how to respond, just tell the user.
 
 ---
 
@@ -329,7 +343,6 @@ Query: {query}"""
 
     prompt = ChatPromptTemplate.from_messages(
         [
-            ("system", "You are an expert legislative analyst."),
             ("human", QUERY_TEMPLATE),
         ]
     )
@@ -398,7 +411,8 @@ def render_query_agent_tab():
 
     from langchain_community.tools import WikipediaQueryRun
     from langchain_community.utilities import WikipediaAPIWrapper
-    from langchain.agents import load_tools
+#    from langchain.agents import load_tools
+    from langchain_community.agent_toolkits.load_tools import load_tools
     from langchain.agents import create_react_agent
     from langchain import hub
 
@@ -497,6 +511,15 @@ elif SS["model_name"] in ANTHROPIC_CHAT_MODELS:
         top_p=SS["top_p"],
         max_tokens_to_sample=SS["max_output_tokens"],
     )
+elif SS["model_name"] in NVIDIA_NIM_CHAT_MODELS:
+    llm = ChatNVIDIA(
+        model=SS["model_name"],
+        temperature=SS["temperature"],
+        max_tokens=SS["max_output_tokens"],
+        top_p=SS["top_p"],
+        seed=SEED,
+        nvidia_api_key=st.secrets["nvidia_api_key"],
+    )
 else:
     raise ValueError()
 

requirements.txt

@@ -41,15 +41,16 @@ jsonpatch==1.33
 jsonpointer==2.4
 jsonschema==4.21.1
 jsonschema-specifications==2023.12.1
-langchain==0.1.13
+langchain==0.2.5
 langchain-anthropic==0.1.1
-langchain-community==0.0.29
-langchain-core==0.1.36
-langchain-openai==0.0.7
+langchain-community==0.2.5
+langchain-core==0.2.7
+langchain-nvidia-ai-endpoints==0.1.2
+langchain-openai==0.1.8
 langchain-pinecone==0.0.3
-langchain-text-splitters==0.0.1
+langchain-text-splitters==0.2.1
 langchainhub==0.1.15
-langsmith==0.1.38
+langsmith==0.1.77
 markdown-it-py==3.0.0
 MarkupSafe==2.1.5
 marshmallow==3.20.2
@@ -60,7 +61,7 @@ multidict==6.0.5
 mypy-extensions==1.0.0
 networkx==3.2.1
 numpy==1.26.4
-openai==1.12.0
+openai==1.34.0
 orjson==3.10.0
 packaging==23.2
 pandas==2.2.1
@@ -102,7 +103,7 @@ streamlit==1.31.1
 sympy==1.12
 tenacity==8.2.3
 threadpoolctl==3.3.0
-tiktoken==0.6.0
+tiktoken==0.7.0
 tokenizers==0.15.2
 toml==0.10.2
 tomli==2.0.1