OpenAI Assistants with OpenAI Python API

Set up your OpenAI account and generate a secret API key to connect Python to the OpenAI API, enabling chat completions with models like GPT-3.5 Turbo.

Explore messages and parameters for building chat bots with the OpenAI Python API, using one-shot and few-shot prompts, system roles, and temperature controls to sustain conversations.

Learn to use the chat completion API with a while-loop exercise notebook to build an astronomy explainer bot that answers questions in simple, five-year-old friendly terms.

Build a space tutor chat bot with the OpenAI Python API, using a chat completion loop of system prompts, user questions, and assistant replies for a five-year-old audience.

Combine a large language model with tools and files, including a code interpreter and custom functions, to create capable assistants and show how threads and runs coordinate responses.

Discover how a base large language model uses a tool to convert math problems into Python code and evaluate it with a code interpreter to ensure deterministic results.

Learn to create and manage OpenAI assistants with the beta API, build threads and messages, and run Python code using the code interpreter in a math tutor example.

Learn to create and manage runs that execute a thread with an assistant, monitor status from queued to completed, retrieve thread messages and run steps, and delete or list assistants.

learn to implement an OpenAI Python API assistant workflow that uses a code interpreter to calculate mortgage payments, manage threads and messages, run tasks, and display results.

Build an OpenAI Python API based assistant with a code interpreter to parse a natural language restaurant bill and fairly split costs with an 18% tip, via a code-along workflow.

Code along to build an OpenAI Python API assistant: create a client and assistant, set instructions, run a thread, exchange messages, and refine bill splitting by order.

Learn how knowledge retrieval uses uploaded files and file IDs, or an always-on file-attached assistant, with embeddings and vector search to answer questions like vacation policy.

Learn to add a single file for an OpenAI Python API assistant by uploading a PDF, retrieving its file ID, and using a retrieval tool in a message.

Upload a file, reference it in a message with its file ID, and let the code interpreter run Python to analyze the file or produce outputs like plots or images.

Upload multiple pdf files, associate the assistant with their file IDs, and use retrieval-powered conversations via the OpenAI Python API to answer with the correct policy documents.

Walk through the exercise to load the exercise_sales_data.csv, connect to OpenAI, create an assistant, generate a bar plot of total dollars per year, and display the image.

Follow along to solve the assistant knowledge retrieval exercise by uploading a sales data CSV, configuring a retrieval-enabled OpenAI assistant, and generating a bar plot of annual purchase expenditure.

Learn to use function calling with OpenAI assistants to run local functions, convert natural language to json outputs, and handle the run status requiring action.

Learn to convert a Python function into the json structure required for OpenAI function calling, defining the function name, json types, parameter descriptions, and required fields with practical examples.

Learn to implement function calling with an assistant by converting a Python function to JSON, creating the assistant, and handling runs, required actions, and tool calls.

Explore the assistant function calling exercise by converting a quiz function to json, creating an assistant, running a thread, and evaluating run results for a president-state quiz.

Demonstrates a code-along for function calling with the OpenAI Python API: convert a Python function to json, build a quiz assistant, and run and test threaded interactions.