Recent submissions

Comment · 3 months ago
Cervantes?
Comment · 4 months ago
As a public proof of P34, this is bullshit. The headline claim rests on missing code, hidden notebook state, inconsistent published numbers, and no reproducible environment.
Comment · 9 months ago
seems a fork of storyboarder without attribution
Comment · 1 year ago
No signup or payment, inmediate payment flow on landing page
Comment · 1 year ago
69 dollars per month? just use airalo
Comment · 1 year ago
what is wallapop? never heard about it
Comment · 2 years ago
Hype
Comment · 2 years ago
A chinese huggingface clone
Comment · 3 years ago
No add documents or memory context?
Comment · 3 years ago
We're thrilled to announce the launch of AST-1, an innovative language model that uses an advanced attention mechanism to produce coherent and contextually relevant text while efficiently handling computational and memory requirements.

The key to AST-1's success is its novel RNN attention mechanism. Unlike traditional methods, RNN attention lets the model dynamically focus on different parts of the input sequence, resulting in more natural-sounding text and the ability to handle long input sequences.

AST-1's architecture is carefully designed for optimal performance and the model parameters are fine-tuned for efficiency and accuracy. Plus, its training process enables it to handle a range of natural language tasks, from sentiment analysis to machine translation.

We believe that AST-1 represents a significant milestone in language model development and invite you to try it out for yourself. Discover the power of AST-1 today and let us know your thoughts!

Comment · 3 years ago
Don't
Comment · 3 years ago
Probably you forgot the most important part. Have rich parents.
Comment · 4 years ago
Yes
Comment · 4 years ago
I'm working exactly on this

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton