← Înapoi la Urmăritori

Cel mai recent articol

Ultima verificare

Urmăritori 0 urmăritori
Articole: 200
  1. How to Get and Interpret GPU Memory Profiling

  2. o3-mini Level Hard Problem

  3. DeepSeek's Recipe Evolution

  4. Can You Spot Error Proving 1 = -1?

  5. Phi-4 Now On HuggingFace with MIT Licence!

  6. Phi-4: The Most Powerful smol Model!

  7. Announcement: PhD Internship for AI for Reasoning

  8. Some Hand Wavy Estimates on FLOPs and Computational Reducibility

  9. Erdős Open Problems as ASI Benchmark

  10. Tokenization and Decoding

  11. Scaling Laws and Data Wall

  12. OpenCoder

  13. The WSD Learning Rate Schedule

  14. Healthier Flossing with Silk!

  15. Tricky Validation Plot

  16. DeepMind's Chess Model

  17. Solving Chicken-Egg Dilemma

  18. New Optimizer: Muon

  19. Vocab Scaling

  20. Using Cursor

  21. Terence Tao Grades o1

  22. Things on Big Bang Theory Writer's Table

  23. Inference Scaling

  24. IMO 2024 Test

  25. LLM Judges Prefers Fluff?

  26. Announcement: Phi-3 14B Rivaling Llama3 70B

  27. Power of Inference Scaling

  28. Human Hours to FLOPs

  29. Sample Efficiency

  30. Phi-3: "ChatGPT" Natively on Your Phone

  31. Inference Memory

  32. Llama3 Leaps Ahead with Data Alone

  33. Pretraining has Gotten 10x Faster in Past Year!

  34. Great Papers

  35. Marvels of ENIAC

  36. Intelligence and FLOPs Efficiency

  37. CIFAR-10 Training to 94% Accuracy in 3.29s

  38. Matmul on CPUs

  39. Using rusage

  40. When Uniform Isn't Uniform: A Bug Hunt Tale

  41. GPT-4 Training Time

  42. Scarce to Plentiful

  43. A Lifetime in 1T Tokens

  44. Power of Two Choices

  45. PyTorch Distributed Debugging Tips

  46. EMA Tuning

  47. Phixtral Fusion

  48. Phi-2: The Most Powerful 2.7B Model Yet

  49. Amazing Robotics Demo with Cheap Hardware