I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend. It's basically nanochat LLM training core stripped down to a single-GPU, one file version of ~630 lines of code, then:
- the human iterates on the prompt (.md)
- the AI agent iterates on the training code (.py)
The goal is to engineer your agents to make the fastest research progress indefinitely and without any of your own involvement. In the image, every dot is a complete LLM training run that lasts exactly 5 minutes. The agent works in an autonomous loop on a git feature branch and accumulates git commits to the training script as it finds better settings (of lower validation loss by the end) of the neural network architecture, the optimizer, all the hyperparameters, etc. You can imagine comparing the research progress of different prompts, different agents, etc.
https://t.co/YCvOwwjOzF
Part code, part sci-fi, and a pinch of psychosis :)
@rene8047@f1multiviewer I'm watching one stream on my laptop, and the other 11 on my desktop. I just have a modern PC (updated within the last few years) with a hardwired network connection. I'm not sure what's required. @f1multiviewer would know.
@brrrake I rather enjoy your complex regurgitation. As a newbie to the sport I've appreciated both your social presence and the style of your videos. They've got some cognitive dissonance shut down due to the nature of the conversation.
Remember yesterday’s news about TikTok releasing a go live platform? Turns out it’s a fork of @OBSProject
Shoutout to @HunterAP23 for pointing this out
STOP STEALING FROM OBS JESUS
@NoMBeMusic While the catalysts are tragic, it’s uplifting to see that SO MANY people see beyond bias and understand that we’re all humans on this earth together; just trying to live.