Machine Learning’s Post

GPT 4o does not have this kind of Features. Take a look: Coding with it is 10 times faster. It has a kind of AGI vibe. https://lnkd.in/g4t2P_D5

RoboCoder - GPT-4 Turbo in VS Code | Product Hunt: GPT-4 Turbo is a new feature in VS Code, designed to enhance the coding experience for developers. It offers improved code suggestions, faster performance, and better language understanding, making it a valuable addition to the development environment. - Artificial Intelligence topics! #ai #artificialintelligence #intelligenzaartificiale

New on #KaggleModels! 👋 Meet MosaicML's MPT: MosaicML's GPT-style model pre trained on 1T tokens of text & code. Perfect for coding with an 8k token context. Explore it here: https://goo.gle/3RuuP1m

The future of coding is being debated. Some industry leaders are calling it over in 5 years. Reminds me of a friend that liked to code assembler. The abstraction layers will increase and will make it easier for everyone to bring systems to life. I see it as a positive force for change. Check out this sick demo by Mckay Wrigley. https://buff.ly/4321vCT #ai #aiforcode

48 hours of using GPT 4o, few things that I was not expecting : (Caution: I am a average user and do not have access to their desktop app or early beta user feature that you saw in demo) 1) 4o is bad at coding. It fails to remember your constraints provided. I had a better debugging assistance from 4. But don’t be naive in your assumptions and expect a fully functional solution from these models. These models are good at batch size small code blocks. You are the engineer and brain! I am back to GPT4 for this! 2) After the introduction of 4o, both 4 and 4o have become quite slow. It gives you outputs in chunks and often lag. But, this could be because of heavy traffic load and their new multimodal input capabilities. But, it is slow at the moment. I haven’t noticed a significant better output yet! May be it is just my use cases at the moment (largely coding related) but the incremental improvement isn’t noticeable yet! I’ll keep stress testing this for Procurement use cases and share my feedback! Supernegotiate

The GPT 4o has a bigger output context and that’s a huge deal for developers. I always felt the output was a little too short for some more complex, but reasonably well encapsulated functions / components, etc. With this latest model release, I feel it is much less of a problem. It also seems to be less likely to take “shortcuts” truncating code or adding comments instead of writing working code. What are your experiences with Omni so far?

𝐀 𝐧𝐞𝐰 𝐛𝐞𝐧𝐜𝐡𝐦𝐚𝐫𝐤 𝐜𝐨𝐦𝐩𝐚𝐫𝐢𝐬𝐨𝐧 𝐬𝐢𝐭𝐞: 𝐡𝐭𝐭𝐩𝐬://𝐚𝐫𝐭𝐢𝐟𝐢𝐜𝐢𝐚𝐥𝐚𝐧𝐚𝐥𝐲𝐬𝐢𝐬.𝐚𝐢 • by an independent third party • clearly outlines the quality vs throughput tradeoff • breaks out by model and hosting provider Site: https://lnkd.in/gqXxgzu2 #ai #genai #llm #benchmark #aimodels #speed #price #coding #codegeneration #coding #aireasoning #aiknowledge #aiability

GPT-4o struggles with debugging its own code. It often repeats the same incorrect solution without improvement. For coding tasks, consider using other LLMs like Anthropic Claude 3 Opus or Claude 3.5 Sonnet, which perform substantially better. Despite being trained for real-time voice conversations, GPT-4o seems less effective at long-form, multi-turn conversations compared to older GPT models. The function calling with GPT-4o is actually worse than GPT-4 Turbo.

Learn why running large language model locally matters, and how LLMs actually work. We start with an overview of the technology and show how you can use the open-source tool Ollama to download and run a model on your laptop. From there, we customize the model and build AI applications you can use right away using NodeJS and Python. Get a solid foundation in AI skills that you will be able to use as soon as finish this session. Docker, Inc #llmops #chatgpt #genai #neo4j #docker #ollama #langchain

OpenAI dev day was pretty mind boggling.... Their newest model, GPT-4 Turbo has a 128k context window and the ability to analyze images. A 128k context window means this model can remember text inputs equivalent to over 300 pages of information. Alongside performance improvements, it also comes with a reduction in cost. There's API access to DALL·E 3 for generating images and a new text-to-speech model for audio output. (Midjourney still takes the cake in my book) Another new feature that really excites me is the Assistants API. This new feature aims to facilitate the creation of agent-like experiences. These specialized AI Assistants can do things like run Python code, process data, and perform knowledge retrieval. Interested in using these new tools to innovate your business? Send me a message! For a detailed overview, visit: https://lnkd.in/gBtAzy4X