Introducing vision to the fine-tuning API
Developers can customize the model to have stronger image understanding capabilities which enables applications like enhanced visual search functionality, improved object detection for autonomous vehicles or smart cities, and more accurate medical image analysis.
Since we first introduced fine-tuning on GPT-4o, hundreds of thousands of developers have customized our models using text-only datasets to improve performance on specific tasks. However, for many cases, fine-tuning models on text alone doesn’t provide the performance boost expected.
How it works
Vision fine-tuning follows a similar process to fine-tuning with text—developers can prepare their image datasets to follow the proper format(opens in a new window) and then upload that dataset to our platform. They can improve the performance of GPT-4o for vision tasks with as few as 100 images, and drive even higher performance with larger volumes of text and image data.
Developers can customize the model to have stronger image understanding capabilities which enables applications like enhanced visual search functionality, improved object detection for autonomous vehicles or smart cities, and more accurate medical image analysis.
Since we first introduced fine-tuning on GPT-4o, hundreds of thousands of developers have customized our models using text-only datasets to improve performance on specific tasks. However, for many cases, fine-tuning models on text alone doesn’t provide the performance boost expected.
How it works
Vision fine-tuning follows a similar process to fine-tuning with text—developers can prepare their image datasets to follow the proper format(opens in a new window) and then upload that dataset to our platform. They can improve the performance of GPT-4o for vision tasks with as few as 100 images, and drive even higher performance with larger volumes of text and image data.
👍1
Google’s Gemini AI Chatbot tells a user to die
Gemini AI told a user to "die" in response to their test question. Gemini generated three drafts in response, and while the second and third drafts were actual answers, the first draft's text told the user to "Please die."
Looks like Gemini's tired of solving questions, lol.
Gemini AI told a user to "die" in response to their test question. Gemini generated three drafts in response, and while the second and third drafts were actual answers, the first draft's text told the user to "Please die."
Looks like Gemini's tired of solving questions, lol.
😁3👍2❤1
Simplifying, stabilizing, and scaling continuous-time consistency models
Diffusion models have revolutionized generative AI, enabling remarkable advances in generating realistic images, 3D models, audio, and video. However, despite their impressive results, these models are slow at sampling.
We are sharing a new approach, called sCM, which simplifies the theoretical formulation of continuous-time consistency models, allowing us to stabilize and scale their training for large scale datasets. This approach achieves comparable sample quality to leading diffusion models, while using only two sampling steps. We are also sharing our research paper(opens in a new window) to support further progress in this field.
Diffusion models have revolutionized generative AI, enabling remarkable advances in generating realistic images, 3D models, audio, and video. However, despite their impressive results, these models are slow at sampling.
We are sharing a new approach, called sCM, which simplifies the theoretical formulation of continuous-time consistency models, allowing us to stabilize and scale their training for large scale datasets. This approach achieves comparable sample quality to leading diffusion models, while using only two sampling steps. We are also sharing our research paper(opens in a new window) to support further progress in this field.
👍1
Sora (video generation model made by OpenAI, same company that made ChatGPT) has been released!!!
Sign ups are currently at a high so they are unavailable that the moment, but word on the street is its not too far off from what we are already seeing with other generative video models that are currently out.
Everything is happening fast. There are some models from China that have been really showcasing great quality.
The future is about to get even weirder my friends 🤓
Sign ups are currently at a high so they are unavailable that the moment, but word on the street is its not too far off from what we are already seeing with other generative video models that are currently out.
Everything is happening fast. There are some models from China that have been really showcasing great quality.
The future is about to get even weirder my friends 🤓
👍1👏1
Introducing canvas
We’re introducing canvas, a new interface for working with ChatGPT on writing and coding projects that go beyond simple chat. Canvas opens in a separate window, allowing you and ChatGPT to collaborate on a project. This early beta introduces a new way of working together—not just through conversation, but by creating and refining ideas side by side.
Canvas was built with GPT-4o and can be manually selected in the model picker while in beta. Starting today we’re rolling out canvas to ChatGPT Plus and Team users globally. Enterprise and Edu users will get access next week. We also plan to make canvas available to all ChatGPT Free users when it’s out of beta.
We’re introducing canvas, a new interface for working with ChatGPT on writing and coding projects that go beyond simple chat. Canvas opens in a separate window, allowing you and ChatGPT to collaborate on a project. This early beta introduces a new way of working together—not just through conversation, but by creating and refining ideas side by side.
Canvas was built with GPT-4o and can be manually selected in the model picker while in beta. Starting today we’re rolling out canvas to ChatGPT Plus and Team users globally. Enterprise and Edu users will get access next week. We also plan to make canvas available to all ChatGPT Free users when it’s out of beta.
❤1👍1
Forwarded from AI Revolution
Google Gemini just introduced Gemini v2.0 with real time API.
Mckay Wrigley on his Twitter account is showing how powerful it is.
He turns it into a live code tutor just by sharing his screen and talking to it 🤯
Crazy!
https://x.com/mckaywrigley/status/1866930933842186427
Mckay Wrigley on his Twitter account is showing how powerful it is.
He turns it into a live code tutor just by sharing his screen and talking to it 🤯
Crazy!
https://x.com/mckaywrigley/status/1866930933842186427
Introducing the Realtime API
Today, we're introducing a public beta of the Realtime API, enabling all paid developers to build low-latency, multimodal experiences in their apps. Similar to ChatGPT’s Advanced Voice Mode, the Realtime API supports natural speech-to-speech conversations using the six preset voices(opens in a new window) already supported in the API.
We’re also introducing audio input and output in the Chat Completions API(opens in a new window) to support use cases that don’t require the low-latency benefits of the Realtime API. With this update, developers can pass any text or audio inputs into GPT-4o and have the model respond with their choice of text, audio, or both.
From language apps and educational software to customer support experiences, developers have already been leveraging voice experiences to connect with their users. Now with Realtime API and soon with audio in the Chat Completions API, developers no longer have to stitch together multiple models to power these experiences.
Today, we're introducing a public beta of the Realtime API, enabling all paid developers to build low-latency, multimodal experiences in their apps. Similar to ChatGPT’s Advanced Voice Mode, the Realtime API supports natural speech-to-speech conversations using the six preset voices(opens in a new window) already supported in the API.
We’re also introducing audio input and output in the Chat Completions API(opens in a new window) to support use cases that don’t require the low-latency benefits of the Realtime API. With this update, developers can pass any text or audio inputs into GPT-4o and have the model respond with their choice of text, audio, or both.
From language apps and educational software to customer support experiences, developers have already been leveraging voice experiences to connect with their users. Now with Realtime API and soon with audio in the Chat Completions API, developers no longer have to stitch together multiple models to power these experiences.
❤1👍1
Prompt Caching in the API
Many developers use the same context repeatedly across multiple API calls when building AI applications, like when making edits to a codebase or having long, multi-turn conversations with a chatbot. Today, we’re introducing Prompt Caching, allowing developers to reduce costs and latency. By reusing recently seen input tokens, developers can get a 50% discount and faster prompt processing times.
Prompt Caching Availability & Pricing
Starting today, Prompt Caching is automatically applied on the latest versions of GPT-4o, GPT-4o mini, o1-preview and o1-mini, as well as fine-tuned versions of those models. Cached prompts are offered at a discount compared to uncached prompts.
Many developers use the same context repeatedly across multiple API calls when building AI applications, like when making edits to a codebase or having long, multi-turn conversations with a chatbot. Today, we’re introducing Prompt Caching, allowing developers to reduce costs and latency. By reusing recently seen input tokens, developers can get a 50% discount and faster prompt processing times.
Prompt Caching Availability & Pricing
Starting today, Prompt Caching is automatically applied on the latest versions of GPT-4o, GPT-4o mini, o1-preview and o1-mini, as well as fine-tuned versions of those models. Cached prompts are offered at a discount compared to uncached prompts.
👍2❤1👏1
Awesome chatgpt prompts
This repo includes ChatGPT prompt curation to use ChatGPT better.
Creator: Fatih Kadir Akın
Stars ⭐️: 114k
Forked By: 15.6k
GithubRepo: https://github.com/f/awesome-chatgpt-prompts
#chatgpt
➖➖➖➖➖➖➖➖➖➖➖➖➖➖
Join @python_bds for more cool repositories.
*This channel belongs to @bigdataspecialist group
This repo includes ChatGPT prompt curation to use ChatGPT better.
Creator: Fatih Kadir Akın
Stars ⭐️: 114k
Forked By: 15.6k
GithubRepo: https://github.com/f/awesome-chatgpt-prompts
#chatgpt
➖➖➖➖➖➖➖➖➖➖➖➖➖➖
Join @python_bds for more cool repositories.
*This channel belongs to @bigdataspecialist group
GitHub
GitHub - f/awesome-chatgpt-prompts: This repo includes ChatGPT prompt curation to use ChatGPT and other LLM tools better.
This repo includes ChatGPT prompt curation to use ChatGPT and other LLM tools better. - f/awesome-chatgpt-prompts
ChatGPT Tasks
ChatGPT has recently launched a new feature - Tasks, that signals a future where AI is far more capable than today.
Tasks let users schedule actions to be carried out in the future or at regular intervals. Scheduled Tasks focuses on automated, time-based interactions.
What Are ChatGPT Scheduled Tasks?
The concept is pretty straightforward – Tasks lets us schedule prompts to be run in the future or at regular intervals, such as once a day, then wait for the results to be relayed back to us by notification or email.
This might not sound like much, but possibilities with tasks are unlimited.
You can practice new language every day, get your exercise plan, recommendation what to make for lunch, reminders for birthdays..anything you want.
Useful prompt:
This is one step closer to AI agents.
Another huge step to it is OpenAI Operator, we will talk about it in our next post 😉
ChatGPT has recently launched a new feature - Tasks, that signals a future where AI is far more capable than today.
Tasks let users schedule actions to be carried out in the future or at regular intervals. Scheduled Tasks focuses on automated, time-based interactions.
What Are ChatGPT Scheduled Tasks?
The concept is pretty straightforward – Tasks lets us schedule prompts to be run in the future or at regular intervals, such as once a day, then wait for the results to be relayed back to us by notification or email.
This might not sound like much, but possibilities with tasks are unlimited.
You can practice new language every day, get your exercise plan, recommendation what to make for lunch, reminders for birthdays..anything you want.
Useful prompt:
Hey, give me Every morning at 7 am, send me the latest news and content on the subject of ____
This is one step closer to AI agents.
Another huge step to it is OpenAI Operator, we will talk about it in our next post 😉
OpenAI has released o3-mini, its fastest, smartest, and most cost-efficient AI yet.
Designed for STEM, coding, and problem-solving, o3-mini now supports function calling, web access, and structured outputs. It also delivers 39% fewer major errors and 24% faster responses compared to o1-mini.
While o3-mini outperforms all previous OpenAI models, o1-pro remains the most powerful, but it's reserved for $200/month subscribers.
Designed for STEM, coding, and problem-solving, o3-mini now supports function calling, web access, and structured outputs. It also delivers 39% fewer major errors and 24% faster responses compared to o1-mini.
While o3-mini outperforms all previous OpenAI models, o1-pro remains the most powerful, but it's reserved for $200/month subscribers.
🔥1
ChatGPT can now generate images.
I tried generating 2 types of images, one is a simple girl and second is complex visual representation of data partitioning vs. sharding.
It's clear that it's not able to grasp this complex visual representations yet, but for girl it did a decent job. I just don't understand why it added this text next to her 😅
PS. Data Partitioning vs Data Sharding might seem as an odd choice, but it is part of newest post I am creating for our @bigdataspecialist Instagram page 😊
I tried generating 2 types of images, one is a simple girl and second is complex visual representation of data partitioning vs. sharding.
It's clear that it's not able to grasp this complex visual representations yet, but for girl it did a decent job. I just don't understand why it added this text next to her 😅
PS. Data Partitioning vs Data Sharding might seem as an odd choice, but it is part of newest post I am creating for our @bigdataspecialist Instagram page 😊
ChatGPT and Deepseek brawl 🤣🤣
Hilarious video of Jimmy Falon talking with both ChatGPT and DeepSeek at the same time in his The Tonight Show
https://youtu.be/Zs__y8N06TY?si=sKWA-pcOIlO6CTfw&t=179
Hilarious video of Jimmy Falon talking with both ChatGPT and DeepSeek at the same time in his The Tonight Show
https://youtu.be/Zs__y8N06TY?si=sKWA-pcOIlO6CTfw&t=179
This media is not supported in your browser
VIEW IN TELEGRAM
Image Creation With ChatGPT.
This is example of what ChatGPT can do when it comes to image creation 🔥
While I still belive there are much better options for generating images, ChatGPT is still great free option (limited of course)
What do you think?
This is example of what ChatGPT can do when it comes to image creation 🔥
While I still belive there are much better options for generating images, ChatGPT is still great free option (limited of course)
What do you think?
Microsoft study finds relying on AI kills your critical thinking skills
A new study from Microsoft and Carnegie Mellon University shows that using generative AI at work could impact our cognitive abilities. Surveying 319 workers, the research found that people who trusted AI outputs tended to think less critically, while those skeptical of AI were more likely to verify results and think analytically.
The study suggests AI may shift professionals from problem-solvers to overseers of AI results, potentially reducing independent thinking and creativity. This raises concerns about AI’s long-term effect on cognitive skills and highlights the need to balance technology with critical thinking development.
A new study from Microsoft and Carnegie Mellon University shows that using generative AI at work could impact our cognitive abilities. Surveying 319 workers, the research found that people who trusted AI outputs tended to think less critically, while those skeptical of AI were more likely to verify results and think analytically.
The study suggests AI may shift professionals from problem-solvers to overseers of AI results, potentially reducing independent thinking and creativity. This raises concerns about AI’s long-term effect on cognitive skills and highlights the need to balance technology with critical thinking development.
👍2