2/ Ghostwriter Chat: We get it – coding on mobile or tablet can be tough. That’s why we brought Ghostwriter directly into the app. You type less and build more. Use Ghostwriter to help you generate, complete, and transform code so you launch faster.
@replit
-
Replit Mobile App Reaches Million Downloads Milestone
By
–
1/ Ever tried coding on your mobile or tablet? People said we were crazy when we launched the Replit app 8 months ago. But almost a million downloads later, here we are…
-
Extension Logs Developer Tools Debugging Improvements
By
–
5/ Extension Logs We added a new section to the Developer Tools pane for Extensions developers. Debugging is now a lot easier during the development process.
-
Replit Secret Presets Automates Environment Variables Setup
By
–
2/ Secret Presets Digging through existing code to find out how to set the correct environment variables is a time-consuming process, especially for building full stack apps After forking a Repl, the key names of existing Environment Variables will now be set in Replit Secrets
-
Fast Search and Replace in Large Repl Projects
By
–
3/ Search & Replace Navigating large projects requires fast and responsive search. Using ripgrep and crosis, we built the functionality to search your Repls. We then used a variety of streaming techniques to make it fast.
-
Replit Changelog: 914 Pull Requests Merged, Six Features Released
By
–
The Replit Changelog In the past three weeks, we merged 914 Pull Requests. Here are six features that just landed
-
Training Unconventional LLMs: Expert Panel on AI Innovation
By
–
“People should be willing to do weird stuff” when training LLMs. Instead of building another LLaMA quality model, why not train something off the wall?
— Replit ⠕ (@Replit) 4 juillet 2023
Watch the full AI panel with @MosaicML’s Chief Scientist @jefrankle in conversation with Replit’s @amasad and @pirroh pic.twitter.com/vfKbdqH34D“People should be willing to do weird stuff” when training LLMs. Instead of building another LLaMA quality model, why not train something off the wall? Watch the full AI panel with @MosaicML
’s Chief Scientist @jefrankle in conversation with Replit’s @amasad and @pirroh -
Quadratic AI Training Costs: How Engineers Build Bigger Models Affordably
By
–
Did you know that the cost of training an AI model increases approximately quadratically as you increase the number of tokens and the model size? ML engineers are still building bigger models without breaking the bank. pic.twitter.com/tB7EiintQl
— Replit ⠕ (@Replit) 3 juillet 2023Did you know that the cost of training an AI model increases approximately quadratically as you increase the number of tokens and the model size? ML engineers are still building bigger models without breaking the bank.
-
GPU Advances Enable Bigger, Wider LLM Model Architectures
By
–
With newer GPU technologies, the shape of our LLM models will begin to change. Bigger, wider, models are now a possibility.
— Replit ⠕ (@Replit) 3 juillet 2023
To learn more watch the full AI panel with @MosaicML’s Chief Scientist @jefrankle in conversation with Replit’s @amasad and @pirroh pic.twitter.com/A4g56a5ZtjWith newer GPU technologies, the shape of our LLM models will begin to change. Bigger, wider, models are now a possibility. To learn more watch the full AI panel with @MosaicML
’s Chief Scientist @jefrankle in conversation with Replit’s @amasad and @pirroh -
Replit Prioritizes Product Quality Over AI Benchmarks
By
–
At Replit we’re not interested in hitting benchmarks and training LLMS to look good, we’re interested in building good products.
— Replit ⠕ (@Replit) 1 juillet 2023
To learn more watch the full AI panel with @MosaicML’s Chief Scientist @jefrankle in conversation with Replit’s @amasad and @pirroh pic.twitter.com/OdFeWLdUdkAt Replit we’re not interested in hitting benchmarks and training LLMS to look good, we’re interested in building good products. To learn more watch the full AI panel with @MosaicML
’s Chief Scientist @jefrankle in conversation with Replit’s @amasad and @pirroh