Upload 9 files

data.txt

@@ -0,0 +1,7 @@
+# data.txt
+
+19 Dec - Go to the movie
+20 Dec - Have a dinner with family
+21 Dec - Go to the birthday party
+22 Dec - Go to the dentist
+23 Dec - Finish writing a draft for blog post

data_loader.py

@@ -0,0 +1,7 @@
+from llama_index.core import SimpleDirectoryReader, VectorStoreIndex
+
+# Load data from a specified directory and file
+document = SimpleDirectoryReader(input_files=['/path/to/data.txt']).load_data()
+
+# Process data (chunking, embedding, indexing) and store them
+index = VectorStoreIndex.from_documents(document)

embed initializer.py

@@ -0,0 +1,10 @@
+from llama_index.core.indices.prompt_helper import PromptHelper
+from llama_index.core.node_parser import SentenceSplitter
+from llama_index.core import Settings
+# Initialize PromptHelper and SentenceSplitter
+prompt_helper = PromptHelper(context_window=2048)
+node_parser = SentenceSplitter(chunk_size=300, chunk_overlap=20)
+
+# Update Settings with helpers
+Settings.prompt_helper = prompt_helper
+Settings.node_parser = node_parser

full run.py

@@ -0,0 +1,57 @@
+import os
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain_community.llms import Ollama
+from llama_index.core import Settings, SimpleDirectoryReader, VectorStoreIndex
+from llama_index.core.indices.prompt_helper import PromptHelper
+from llama_index.core.node_parser import SentenceSplitter
+
+
+def setup_environment():
+    # Set environment variables
+    os.environ['llm'] = 'tinyllama'
+
+
+def initialize_settings():
+    # Initialize and configure the LLM and embedding model
+    llm_model = os.getenv('llm')
+    llm = Ollama(model=llm_model)
+    embed_model = HuggingFaceEmbeddings(model_name="BAAI/bge-small-en-v1.5")
+
+    # Initialize prompt helper and sentence splitter
+    prompt_helper = PromptHelper(context_window=2048)
+    node_parser = SentenceSplitter(chunk_size=300, chunk_overlap=20)
+
+    # Configure global settings for application
+    Settings.llm = llm
+    Settings.embed_model = embed_model
+    Settings.prompt_helper = prompt_helper
+    Settings.node_parser = node_parser
+
+
+def load_and_index_data(file_path):
+    # Load data from the specified file path
+    document = SimpleDirectoryReader(input_files=[file_path]).load_data()
+
+    # Process data (chunking, embedding, indexing) and store them in a vector store index
+    return VectorStoreIndex.from_documents(document)
+
+
+def query_data(query_engine, query):
+    # Query the indexed data and return the response
+    return query_engine.query(query)
+
+
+if __name__ == "__main__":
+    setup_environment()
+    initialize_settings()
+
+    # Assuming the file is stored at a specified path
+    file_path = 'data.txt'
+    index = load_and_index_data(file_path)
+
+    # Build a query engine from the index
+    query_engine = index.as_query_engine()
+
+    # Example query
+    response = query_data(query_engine, 'show me my calander dates.')
+    print(response)

initializer2.py

@@ -0,0 +1,7 @@
+from llama_index.core import Settings
+
+# Configure a global setting for our app so that the VectorStoreIndex and the QueryEngine can use these components by default.
+Settings.llm = llm
+Settings.embed_model = embed_model
+Settings.prompt_helper = prompt_helper
+Settings.node_parser = node_parser

run_ollama.sh

@@ -0,0 +1,9 @@
+#!/bin/bash
+
+# This script will first install ollama and then run it with the 'mistral' command.
+
+# Download and execute the installation script
+curl -fsSL https://ollama.com/install.sh | sh
+
+# Run ollama with the 'mistral' command
+ollama run mistral

setup.py

@@ -0,0 +1,14 @@
+from langchain_community.llms import Ollama
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from llama_index.core import Settings
+
+# Create an Ollama instance with the model configuration
+llm = Ollama(model='tinyllama')
+
+# Use the llm instance directly where needed
+Settings.llm = llm
+
+# Configure the embedding model for your settings
+Settings.embed_model = HuggingFaceEmbeddings(
+    model_name="BAAI/bge-small-en-v1.5"
+)

tester.py

@@ -0,0 +1,6 @@
+# Build a query engine from the index
+query_engine = index.as_query_engine()
+
+# Execute a query and print the response
+response = query_engine.query('Give me my calendar.')
+print(response)