Generative AI with Python: Build Apps in OpenAI & LangChain

OpenAI's developer documentation shows how to start with API, sign up and get a $5 credit, while exploring chart completion API, audio, embeddings, and vector-based data to train language models.

Install the OpenAI Python library, configure your API key, and explore GPT language models to draft documents, write code, and translate texts while learning tokenization and token usage.

Create a new generative AI project with Python by setting up a virtual environment, installing dependencies, configuring an OpenAI API key, and running a two-option menu to ask questions.

Configure a Python project to use the OpenAI API by loading environment variables, setting the API key, installing the OpenAI library, selecting GPT 3.5 turbo, and defining a test prompt.

Define a prompt and call the OpenAI API to generate completions, test responses, adjust token limits, and loop for continuous user queries with token usage insights.

Learn to build AI powered chatbots in Python by setting up a project and virtual environment, installing packages, and using the chat completion API with GPT-3.5 Turbo and OpenAI.

Create a basic chatbot using the chat completion api with GPT-3.5 turbo, configure a system persona and message roles, and set up a virtual environment and api key.

Define an API key environment variable, revoke and create a secret key, save it, and update the dot env to run a Python chatbot with a system message.

Define how a system message sets the chatbot's role and behavior, including tone, via a message object with role and contents. Explain how prompts guide chat completions from start.

Craft chat completions by sending user messages to a language model, managing system messages, and printing the assistant's response via the Chat Completions API.

Explore the moderation endpoint to add a moderation layer that classifies user text against harmful content categories, filters out hate and threats, and safeguards a safe, respectful online environment.

Integrate a moderation layer in a Python app with OpenAI and LangChain, define moderation requests, obtain classification and category scores, and handle flagged content.

Learn to generate original images from text using Dall-E and OpenAI APIs, then install packages like Pillow, manage API keys, and build a gallery that saves outputs to media.

Build an ai image generation gallery by sending text prompts to OpenAI, generating images with DALL-E, and displaying results in a Streamlit gallery with a media download option.

Generate and save images from prompts like enchanted forest and old oak tree with sunlight, then read from image folder and display them in a two-column streamlit gallery with captions.

Discover whisper, an open source multilingual speech recognition model by OpenAI, enabling transcription and translation. Install whisper, process mp3 and wav files, and explore language identification with the OpenAI API.

Demonstrate audio transcription with Whisper in a Streamlit app that uploads a file, runs transcription on submit, displays the transcript, and explores translating the result into English.

Transcribe and translate audio with Whisper Audio API to produce original and English translated transcripts, saved as paired files for French and German samples.

Leverage OpenAI's function calling to build AI-powered agents for natural interactions and text processing tasks (e.g., generating images, transcribing audio) that fetch data via public or backend APIs, including weather.

Understand function calling to let language models fetch current data and call APIs, producing weather forecasts by defining functions and parameters for location and units.

Explore function calling execution by teaching the model to detect tool calls and execute defined functions, including handling the tool calls object and integrating external APIs like weather.

Explore a retrieval augmented technique for chatbots using LangChain and OpenAI, with embeddings stored in Chroma DB, prompts, chain expression language, and a Streamlit interface.

Discover how LangChain uses prompt templates and the Link Chain expression language to connect language models. Build chain-based apps by invoking chat or string models and parsing outputs.

Create chain objects by combining a prompt template with a model. A pipe operator links components, enabling outputs to flow into the language model and into a vector store.

Load documents as a data source, split them into chunks, and build an AI assistant trained on custom data using chroma as a vector store and embeddings.

Train the custom AI assistant to retrieve knowledge by loading document chunks, creating embeddings in a chroma vector store, enabling search by similarity, and using a system prompt with context.

Build a Streamlit-based user interface for the AI assistant in app.py, wiring document chunking, embeddings in a vector store, a retriever context, and a runnable passthrough for queries.

Generative ai expands across business and customer service, reshaping how we work with technology. Explore multilingual nlp, computer vision, emotional intelligence, iot integrations as the ai ecosystem evolves.