Forefront

Open AI starts here. Run & fine-tune open-source models on your data.

[Forefront @ForefrontAI]

Today we’re launching Forefront chat—a better ChatGPT experience—in free alpha. Sign up to get free access to GPT-4, image generation, custom personas, shareable chats, and much more: https://t.co/lqsY9bkvl8

[Vik Paruchuri @VikParuchuri]

Announcing Surya OCR 2! It uses a new architecture and improves on v1 in every way:

- OCR with automatic language detection for 93 languages (no more specifying languages!)
- More accurate on old/noisy documents
- 20% faster
- Basic English handwriting support

[Nico Albanese @nicoalbanese10]

very cool open-source project built with the ai sdk, nextjs, and shadcn

great use case and love that it supports multiple models and providers, configurable with just a dropdown!

[Forefront @ForefrontAI]

Have an idea for a feature? Contribute a pull request on Github, leave a comment, or send us a DM. We'd love your feedback!

Enjoy

[Forefront @ForefrontAI]

🚀 Today we're launching a free, open-source tool for AI dataset management:

🔍 Quickly view, filter, and edit datasets with keyboard shortcuts
🧠 Generate or rewrite samples using AI, powered by @vercel AI SDK
💾 Download datasets for fine-tuning anywhere
🔒 Everything is on-device to preserve privacy
🌐 Fully open-source project

⬇️ Check out the links in the thread below

[Forefront @ForefrontAI]

You can select from the list of AI providers or bring your own endpoint as long as it's compatible with OpenAI schema.

The current list of providers includes @OpenAI , @MistralAI , @GroqInc , @openrouter , and @FireworksAI_HQ

We NEVER store your data or API keys, all activity happens locally on your machine

[Mike @lisanalgaib]

Tokenizers for OpenAI models are everywhere, not the case for Mistral.

I always have to dig up a python notebook when I need to count tokens.

No longer the case...

Mistral tokenizer is now live at https://t.co/qTKrBtvVUb

[Forefront @ForefrontAI]

Workflows are an incredibly flexible tool to automate virtually any kind of work, and we're excited to see what you can automate. Join our Discord for product updates: https://t.co/07CQrChX55

[Forefront @ForefrontAI]

Today we're releasing Workflows. Workflows are reusable chat templates to automate many different types of tasks.

Watch the following demo to see how to automate content generation, sales outreach, and financial news briefings.

Get started for free: https://t.co/tNBUeeNRAv

[Forefront @ForefrontAI]

Soon we'll add a variety of integrations which will enable workflows to run based on events in external applications as well as take actions in external applications.

[Forefront @ForefrontAI]

This latest update along with others coming over the next few days are building the foundation for a product-defining release planned in the coming weeks—stay tuned and join our Discord: https://t.co/07CQrChX55

[Forefront @ForefrontAI]

Today we’re releasing chat with your data, URL browsing, improved assistants & files, and nicknames.

Now you can filter, visualize, and analyze your data without writing a single line of code. Explore data to identify trends and outliers, visualize data with charts and graphs, create complex filters, and much more—without writing a single line of code.

Get started for free: https://t.co/wn3JHpdmhs

[Forefront @ForefrontAI]

Personalize your chat experience by providing a nickname in Settings -> Profile. Models and assistants will then refer to you as your nickname.

[Forefront @ForefrontAI]

Shoutout to the team at StabilityAI: @emostaque, @robrombach, and others.

Start using Stable Diffusion XL for free in a unified chat experience with powerful language models, Internet browsing, chat with files, and much more: https://t.co/S0mcacjMcd

[Forefront @ForefrontAI]

Stable Diffusion XL is now live on https://t.co/lqsY9bkvl8 for free.

Tell GPT-4 to browse the Internet for prompt ideas, remix the prompts with a persona of your favorite artist, and generate stunning images.

The image quality is insane 👇

[Forefront @ForefrontAI]

More mind-blowing images from SDXL: