Deepseek R1 and Open R1:
With Deepseek R1 and Open R1, we can soon finetune this free and open reasoning model to fit our use cases (quantum
computing and personalized visualization).
Deep Research from OpenAI:
I think it's revolutionary for data analysis. It doesn't quite get to data visualization or scientific visualization
yet, since it is now mostly gathering and reasoning on information based on the definite needs from users but in
SciVis or DataVis, users initially may not know what they are looking for. However, I can see it will quickly (in less
than a year) be capable of doing that, giving insights and visualization based on raw data and vague questions.
Train your own R1 reasoning model with Unsloth:
By unsloth.ai, this magically reduces VRAM requirement for post-train a model into a reasoning model from 160GB to
7-40GB. This is very helpful for turning a model with domain-specific datasets, e.g., quantum computing and/or
visualization