view article Article There is no such thing as a tokenizer-free lunch By catherinearnett • Sep 25 • 84
Unsloth 4-bit Dynamic Quants Collection Unsloths Dynamic 4bit Quants selectively skips quantizing certain parameters; greatly improving accuracy while only using <10% more VRAM than BnB 4bit • 28 items • Updated 26 days ago • 87
🧠Reasoning datasets Collection Datasets with reasoning traces for math and code released by the community • 24 items • Updated May 19 • 172
Scaling Laws for Downstream Task Performance of Large Language Models Paper • 2402.04177 • Published Feb 6, 2024 • 19
OpenMoE: An Early Effort on Open Mixture-of-Experts Language Models Paper • 2402.01739 • Published Jan 29, 2024 • 28
DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models Paper • 2402.03300 • Published Feb 5, 2024 • 129
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning Paper • 2501.12948 • Published Jan 22 • 420
The Deepseek AI Collection Collection Papers and Models by Deepseek AI • 7 items • Updated Apr 4 • 1
Aya Dataset: An Open-Access Collection for Multilingual Instruction Tuning Paper • 2402.06619 • Published Feb 9, 2024 • 56