palashshah

totally. it got to a point where most of the text generated in our images was incorrect, and so it wasn't a great look showing that to our clients.

we're actually working on some form of what you described where we take images generated from LLMs + add consistent logos discretely rather than generatively.

totally. the way i think about it (purely based on intuition) is that asking an LLM to do understanding + image generation is too complex for it to be effective. if we separate out the tasks into discrete steps, the evaluation becomes better, and the generation simply becomes instruction following.

this is incredible to hear! i plan to keep writing on a weekly basis, and will be posting them on twitter.

hey! we're working with an initial set of customers, and plan to launch full capabilities soon. stay tuned :)

we're currently in the process of doing this. i think something that could potentially work is to iterate upon the initial image composition / structure using cheaper models, and then upscale at the end. this way you're saving on that iteration cost, but eventually land on a higher-scale image.

appreciate the compliment! yep, it's definitely necessary and is the bare minimum for building image generation systems in production.

totally agreed here. i think my goal primarily with the mask generation was to test out how effective openai's capabilities were.

we're currently working on pipelines that limit the the involvement of AI to various tasks. for example, when generating an ad there's usually logo, some banner text, and background image.

we can use gpt-image-1 to generate the background image, another LLM to identify the coordinates of where we place the logo, and just add the logo onto the image. this is just one example!

Excited to announce Nylon: an interactive grammar of machine learning. Modify what you want, and let us handle the rest. Inspired by Vega-Lite.

Hi everyone!

Just launched on Product Hunt today for public use. Would love some support!

Before building Paraglide, I asked myself the question: why aren't more people automating machine learning, even-though the tools exist? After talking to hundreds of users, the answer was obvious. The tools that exist right now are all the same. They force you to work completely hands off, and give their platform full control! Paraglide is different. You can customize exactly what you want, and let us handle the rest! We've built it in layers, the natural language grammar + custom modules, to help users of all expertise levels customize the AI pipeline.

At this stage, I'm primarily just looking for feedback. That's why we've made the service so cheap compared to its competitors. Would love to have you on the API trying it out!

I'm the founder of Paraglide, and I'd love for anybody to sign up on the waitlist! We're building a tool that lets you create, collaborate on, and customize automated AI workflows in minutes.

my team and I have been working on this for quite a long time. proud to say it's finally gaining traction. our goal is to reach 1k stars by the end of the month. please help us out hackernews :)

My team and I have created an API that lets you build and train models in just one line of code. It's been used by top executives like Steve Nouri and Isaac Faber. Please let me know what you think and how I can improve.

Would really appreciate stars on the GitHub, this has been a 6 month effort!