roborovskis

4 ms·

1.
by roborovskis 7mo ago
What would you define as 'distillation' versus 'learning'? How do you know that what a LLM is doing is 'distillation' vs a process closer to a human reading a book? From my perspective, pretraining is pretty cl
2.
Tinker: General Availability and Vision Input(thinkingmachines.ai)
5 points by roborovskis 10mo ago|0 comments
3.
by roborovskis 2y ago
Where are you seeing this? On https://github.com/deepseek-ai/DeepSeek-R1/tree/main?tab=rea... I only see the paper and related figures.
4.
Reviewing Post-Training Techniques: DeepSeek, Qwen-2, and Phi-4(brianfitzgerald.xyz)
2 points by roborovskis 2y ago|0 comments
5.
by roborovskis 2y ago
https://stable-baselines3.readthedocs.io/en/master/ is a great resource for hacking on implementations for RL - many good RL courses out there but https://www.youtube.com/playlist?list=PLwRJQ4m4UJj
6.
Tricks for DPO Tuning a Code LLM, Part 1: Logit Curriculum Learning(brianfitzgerald.xyz)
2 points by roborovskis 2y ago|0 comments
7.
ChatGPT for macOS can now work with apps on your desktop(twitter.com)
1 points by roborovskis 2y ago|0 comments
8.
Implementation of a Diffusion Transformer and Rectified Flow Sampling in Jax(github.com)
1 points by roborovskis 2y ago|0 comments
9.
Stable Diffusion 3 API Now Available(stability.ai)
246 points by roborovskis 2y ago|61 comments
10.
by roborovskis 3y ago
You could definitely use this for upsampling negative prompts, though I haven't tested that much. In theory, future T2I models shouldn't need to be negatively prompted as much; I find it's better to focus on really high quali
11.
by roborovskis 3y ago
Yup, the model will still forget details sometimes. This is a common issue with prompt upsampling methods, but I'm hoping to improve this with the next version.
12.
by roborovskis 3y ago
Thanks for the kind words! I started with the 780M param flan-t5-large model, and kept trying smaller and smaller base models - I was shocked at how good the output was at 77M. As you go smaller, though, it's much easier to accidentall
13.
by roborovskis 3y ago
As Invoke is open-source and already has transformers as a dependency, it should be pretty easy to add.
14.
by roborovskis 3y ago
I haven't tested extensively with non SDXL based checkpoints but there's nothing really SDXL specific about the model; if you're using a fine-tune that's trained on booru-style tags, it will probably not work as well - b
15.
by roborovskis 3y ago
will fix these, thanks for the heads up!
16.
SuperPrompt: Better Text to Image Prompts in 77M Parameters(brianfitzgerald.xyz)
150 points by roborovskis 3y ago|31 comments
17.
StableLM Zephyr 3B(stability.ai)
123 points by roborovskis 3y ago|38 comments
18.
Stable Video Diffusion(stability.ai)
1330 points by roborovskis 3y ago|302 comments
19.
by roborovskis 3y ago
https://dreamstudio.ai/
20.
Comma Three Devkit(comma.ai)
95 points by roborovskis 5y ago|60 comments
21.
Elon Musk Called a Tesla Critic’s Boss to Complain About Him(jalopnik.com)
3 points by roborovskis 8y ago|1 comments
22.
by roborovskis 13y ago
The fact that they can see what plugins I have makes uTorrent wanting a browser plugin make total sense. Just thinking.