Hugo Podworski’s Post

New model! GPT-4o!! Fully multimodal!! Did I integrate it immediately into my API wrapper? Of course I did. "Cheaper and faster" is too good to pass up. ("Cheaper" sure is nice, but honestly I'm looking forward to "faster". The inference speed of GPT-4 Turbo was ... not awesome.)

Classic example of GPT4 giving you a wrong answer while sounding so sure of it! The right answer to my question below is YES. Streamlit will add the key associated with the button to the session_state.

Quick update on industry-specific models: DeepSeek 2.5 is now the top model for code autocomplete, beating out Claude Sonnet and completely leaving GPT-4o-mini in the dust. Bigger models weren’t included because their latency is just too high. Just another reminder that while AGI sounds amazing, task-specific models and fine-tuning are still making a real difference right now.

How fast can you produce a working game using #genAI? It took me less than a minute to produce minesweeper. Everything is on this page. Time taken = [write prompt] + [wait for GPT output] + [check output] + [copy & paste code] + [run code] = 28 + 11 + 15 + 3 + 2 = 59 seconds 🤯

🚀 Curious about what's under the hood of a GPT? @moebio did a clever visualization of the inner workings of an LLM. Check it out here: https://hubs.ly/Q02qp-tv0 🔍🧠

granite-20b-code outperforming gpt-4 on bird benchmark: https://lnkd.in/dYNPbU7u

GPT API users, it's so easy to switch to GPT-4o, and it gives some impressive benefits. A 16 second video talking about them! https://lnkd.in/e_m9JNsh

Ok, GPT 4o was the im-also-a-good-gpt2-chatbot hah! I've been lowkey staying on LMSys just to keep using this version of gpt because the output was really good. Link to tweet: https://lnkd.in/e-bzVGKN Link to openai: https://lnkd.in/eqZQEf4j

OpenAI quietly released GPT-4o mini last week. Besides the press release hype, here were my takeaways: 📌 GPT-4o mini is the new “free-tier” It’s a small, cheap model that is 60% cheaper than GPT-3.5-turbo. Offered to all users (including free users). 📌 Great for low latency and real-time use-cases Voice calling, function calling, swarm of agents…this model is for you. You get a model that is faster and cheaper for high-volume, low reasoning tasks. 📌 Multi-Modality is the same price The cost to use image capabilities of GPT-4o mini is the SAME price as GPT-4o. You might as well use the bigger model! Video goes into more detail. Comment if you have any questions!

Harnessing the Power of GPT-4 Omni for Multimodal Processing: A Comprehensive Guide https://buff.ly/3wsfx58 A step by step tutorial on how to get started with the GPT-4o API