Update app.py

app.py

@@ -1,57 +1,46 @@
-import os
-
-import octoai
-octoai_client = octoai.client.Client(token=os.getenv('OCTOML_KEY'))
+from langchain_voyageai import VoyageAIEmbeddings
+embed_model = VoyageAIEmbeddings(
+    voyage_api_key=os.getenv('VOYAGE_API_KEY'), model="voyage-large-2-instruct", # input_type="document",
+)
 
 from pinecone import Pinecone, ServerlessSpec
-pc = Pinecone(api_key=os.getenv('PINECONE_API_KEY'))
 
+pc = Pinecone(api_key=os.getenv('PINECONE_API_KEY2'))
+pc_256 = pc.Index('subset1-voyage-large-2-instruct-cs256')
 
 from llama_index.vector_stores.pinecone import PineconeVectorStore
 from llama_index.core import VectorStoreIndex
-from llama_index.core.response.pprint_utils import pprint_source_node
-from llama_index.llms.octoai import OctoAI
 
-octoai = OctoAI(
-    token=os.getenv('OCTOML_KEY'),
-    model="meta-llama-3-70b-instruct",
+vector_store = PineconeVectorStore(pinecone_index=pc_256)
+vindex = VectorStoreIndex.from_vector_store(vector_store, embed_model=embed_model)
+
+from llama_index.llms.fireworks import Fireworks
+
+fireworks_model = 'accounts/fireworks/models/llama-v3-70b-instruct'
+llm = Fireworks(
+    api_key=os.getenv('FIREWORKS_API_KEY'),
+    model=fireworks_model,
     max_tokens=512,
     temperature=0.1,
 )
 
-
-from llama_index.core.memory import ChatMemoryBuffer
-
 import gradio as gr
-from io import StringIO
-
-def get_credit_dist(history):
-    _out = StringIO()
-    print("Disabled momentarily...", file=_out)
-
-    return _out.getvalue()
+from llama_index.core.memory import ChatMemoryBuffer
 
+def get_chat_engine():
+    memory = ChatMemoryBuffer.from_defaults(token_limit=5000)
+    return vindex.as_chat_engine(
+        chat_mode="context",
+        llm=llm,
+        memory=memory,
+        system_prompt="You are a chatbot, able to have normal interactions, as well as talk about news events.",
+    )
 
 with gr.Blocks() as demo:
-    chatbot = gr.Chatbot(height=800)
+    chatbot = gr.Chatbot(height="80vh")
     msg = gr.Textbox()
     clear = gr.Button("Clear")
 
-    credit_box = gr.Textbox(label="Credit distribution", lines=20, autoscroll=False)
-    credit_btn = gr.Button("Credit response")
-
-    def get_chat_engine():
-        vector_store = PineconeVectorStore(pinecone_index=pc.Index('prorata-postman-ds-256'))
-        vindex = VectorStoreIndex.from_vector_store(vector_store)
-        
-        memory = ChatMemoryBuffer.from_defaults(token_limit=5000)
-        return vindex.as_chat_engine(
-            chat_mode="context",
-            llm=octoai,
-            memory=memory,
-            system_prompt="You are a chatbot, able to have normal interactions, as well as talk about news events provided in the context of the conversation.",
-        )
-
     chat_engine_var = gr.State(get_chat_engine)
 
     def user(user_message, history):
@@ -67,8 +56,6 @@ with gr.Blocks() as demo:
     msg.submit(user, [msg, chatbot], [msg, chatbot], queue=False).then(bot, [chatbot, chat_engine_var], chatbot)
     clear.click(lambda x: x.reset(), chat_engine_var, chatbot, queue=False)
 
-    credit_btn.click(get_credit_dist, chatbot, credit_box)
-    
 if __name__ == "__main__":
     demo.queue()
-    demo.launch()
+    demo.launch()