Sublime
An inspiration engine for ideas
Happy to share 💭 Mixture of Thoughts 💭
A curated, general reasoning dataset that trims down over 1M samples from public datasets to ~350k through an extensive set of ablations 🧑🍳
Models trained on this mix match or exceed the performance of DeepSeek's distilled models -- not just on mat... See more
For example, the questions below seem like they are difficult or might require a lot of research to answer: How many prospective customers landed on my Web site? What are the demographics of my best and worst customers? Are customers who interact with my support team more likely to stay customers longer? However, they
Gabriel Weinberg • Traction: How Any Startup Can Achieve Explosive Customer Growth
So I might have been wrong putting my cards on Qwen 3 Coder. GLM 4.5 seems to be better. Also, looking at trending models, the world agrees. https://t.co/qx1QuxZT1E