"GPT-4o (\"o\" for \"omni\") is designed to handle a combination of text, audio, and video inputs, and can generate outputs in text, audio, and image formats.\n",
"\n",
"### Background\n",
"Before GPT-4o, users could interact with ChatGPT using Voice Mode, which operated with three separate models. GPT-4o will integrate these capabilities into a single model that's trained across text, vision, and audio. This unified approach ensures that all inputs—whether text, visual, or auditory—are processed cohesively by the same neural network.\n",
"\n",
"### Current API Capabilities\n",
"Currently, the API supports `{text, image}` inputs only, with `{text}` outputs, the same modalities as `gpt-4-turbo`. Additional modalities, including audio, will be introduced soon. This guide will help you get started with using GPT-4o for text, image, and video understanding.\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Getting Started"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Install OpenAI SDK for Python\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%pip install --upgrade openai --quiet"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Configure the OpenAI client and submit a test request\n",
"To setup the client for our use, we need to create an API key to use with our request. Skip these steps if you already have an API key for usage. \n",
"\n",
"You can get an API key by following these steps:\n",
"1. [Create a new project](https://help.openai.com/en/articles/9186755-managing-your-work-in-the-api-platform-with-projects)\n",
"2. [Generate an API key in your project](https://platform.openai.com/api-keys)\n",
"3. (RECOMMENDED, BUT NOT REQUIRED) [Setup your API key for all projects as an env var](https://platform.openai.com/docs/quickstart/step-2-set-up-your-api-key)\n",
"\n",
"Once we have this setup, let's start with a simple {text} input to the model for our first request. We'll use both `system` and `user` messages for our first request, and we'll receive a response from the `assistant` role."
" {\"role\": \"system\", \"content\": \"You are a helpful assistant. Help me with my math homework!\"}, # <-- This is the system message that provides context to the model\n",
" {\"role\": \"user\", \"content\": \"Hello! Could you solve 2+2?\"} # <-- This is the user message for which the model will generate a response\n",
"To find the area of the triangle, we can use Heron's formula. Heron's formula states that the area of a triangle with sides of length \\(a\\), \\(b\\), and \\(c\\) is:\n",
"While it's not possible to directly send a video to the API, GPT-4o can understand videos if you sample frames and then provide them as images. It performs better at this task than GPT-4 Turbo.\n",
"\n",
"Since GPT-4o in the API does not yet support audio-in (as of May 2024), we'll use a combination of GPT-4o and Whisper to process both the audio and visual for a provided video, and showcase two usecases:\n",
"1. Summarization\n",
"2. Question and Answering\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Setup for Video Processing\n",
"We'll use two python packages for video processing - opencv-python and moviepy. \n",
"\n",
"These require [ffmpeg](https://ffmpeg.org/about.html), so make sure to install this beforehand. Depending on your OS, you may need to run `brew install ffmpeg` or `sudo apt install ffmpeg`"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%pip install opencv-python --quiet\n",
"%pip install moviepy --quiet"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Process the video into two components: frames and audio"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [],
"source": [
"import cv2\n",
"from moviepy.editor import VideoFileClip\n",
"import time\n",
"import base64\n",
"\n",
"# We'll be using the OpenAI DevDay Keynote Recap video. You can review the video here: https://www.youtube.com/watch?v=h02ti0Bl6zk\n",
"VIDEO_PATH = \"data/keynote_recap.mp4\""
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"MoviePy - Writing audio in data/keynote_recap.mp3\n"
"Now that we have both the video frames and the audio, let's run a few different tests to generate a video summary to compare the results of using the models with different modalities. We should expect to see that the summary generated with context from both visual and audio inputs will be the most accurate, as the model is able to use the entire context from the video.\n",
"\n",
"1. Visual Summary\n",
"2. Audio Summary\n",
"3. Visual + Audio Summary\n",
"\n",
"#### Visual Summary\n",
"The visual summary is generated by sending the model only the frames from the video. With just the frames, the model is likely to capture the visual aspects, but will miss any details discussed by the speaker."
"The results are as expected - the model is able to capture the high level aspects of the video visuals, but misses the details provided in the speech.\n",
"\n",
"#### Audio Summary\n",
"The audio summary is generated by sending the model the audio transcript. With just the audio, the model is likely to bias towards the audio content, and will miss the context provided by the presentations and visuals.\n",
"\n",
"`{audio}` input for GPT-4o isn't currently available but will be coming soon! For now, we use our existing `whisper-1` model to process the audio"
"Welcome to OpenAI's first-ever Dev Day. Key announcements include:\n",
"\n",
"- **GPT-4 Turbo**: A new model supporting up to 128,000 tokens of context, featuring JSON mode for valid JSON responses, improved instruction following, and better knowledge retrieval from external documents or databases. It is also significantly cheaper than GPT-4.\n",
"- **New Features**: \n",
" - **Dolly 3**, **GPT-4 Turbo with Vision**, and a new **Text-to-Speech model** are now available in the API.\n",
" - **Custom Models**: A program where OpenAI researchers help companies create custom models tailored to their specific use cases.\n",
" - **Increased Rate Limits**: Doubling tokens per minute for established GPT-4 customers and allowing requests for further rate limit changes.\n",
"- **GPTs**: Tailored versions of ChatGPT for specific purposes, programmable through conversation, with options for private or public sharing, and a forthcoming GPT Store.\n",
"- **Assistance API**: Includes persistent threads, built-in retrieval, a code interpreter, and improved function calling.\n",
"\n",
"OpenAI is excited about the future of AI integration and looks forward to seeing what users will create with these new tools. The event concludes with an invitation to return next year for more advancements.\n"
" {\"role\": \"system\", \"content\":\"\"\"You are generating a transcript summary. Create a summary of the provided transcription. Respond in Markdown.\"\"\"},\n",
" {\"role\": \"user\", \"content\": [\n",
" {\"type\": \"text\", \"text\": f\"The audio transcription is: {transcription.text}\"}\n",
"The audio summary is biased towards the content discussed during the speech, but comes out with much less structure than the video summary.\n",
"\n",
"#### Audio + Visual Summary\n",
"The Audio + Visual summary is generated by sending the model both the visual and the audio from the video at once. When sending both of these, the model is expected to better summarize since it can perceive the entire video at once."
" {\"role\": \"system\", \"content\":\"\"\"You are generating a video summary. Create a summary of the provided video and its transcript. Respond in Markdown\"\"\"},\n",
"After combining both the video and audio, we're able to get a much more detailed and comprehensive summary for the event which uses information from both the visual and audio elements from the video.\n",
"\n",
"### Example 2: Question and Answering\n",
"For the Q&A, we'll use the same concept as before to ask questions of our processed video while running the same 3 tests to demonstrate the benefit of combining input modalities:\n",
"1. Visual Q&A\n",
"2. Audio Q&A\n",
"3. Visual + Audio Q&A "
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [],
"source": [
"QUESTION = \"Question: Why did Sam Altman have an example about raising windows and turning the radio on?\""
"Sam Altman used the example about raising windows and turning the radio on to demonstrate the function calling capability of GPT-4 Turbo. The example illustrated how the model can interpret and execute multiple commands in a more structured and efficient manner. The \"before\" and \"after\" comparison showed how the model can now directly call functions like `raise_windows()` and `radio_on()` based on natural language instructions, showcasing improved control and functionality.\n"
"The provided transcription does not include any mention of Sam Altman or an example about raising windows and turning the radio on. Therefore, I cannot provide an answer based on the given transcription.\n"
"Sam Altman used the example of raising windows and turning the radio on to demonstrate the improved function calling capabilities of GPT-4 Turbo. The example illustrated how the model can now handle multiple function calls more effectively and follow instructions better. In the \"before\" scenario, the model had to be prompted separately for each action, whereas in the \"after\" scenario, the model could handle both actions in a single prompt, showcasing its enhanced ability to manage and execute multiple tasks simultaneously.\n"
"Comparing the three answers, the most accurate answer is generated by using both the audio and visual from the video. Sam Altman did not discuss the raising windows or radio on during the Keynote, but referenced an improved capability for the model to execute multiple functions in a single request while the examples were shown behind him.\n",
"\n",
"## Conclusion\n",
"Integrating many input modalities such as audio, visual, and textual, significantly enhances the performance of the model on a diverse range of tasks. This multimodal approach allows for more comprehensive understanding and interaction, mirroring more closely how humans perceive and process information. \n",
"\n",
"Currently, GPT-4o in the API supports text and image inputs, with audio capabilities coming soon."
