Training

  1. Start of training
  2. Pre-training, 8KKimi K3not stated8KNTPtext and visionMTPallMTP layer
  3. Pre-training, 64KKimi K3not stated64KNTPtext and visionMTPall
  4. Cooldown, 256KKimi K3not stated256KNTPall
  5. Cooldown, 1MKimi K3not stated1MNTPall
  6. Off this path: Draft model1 stage
  7. The frozen target and the time: not stated. Speculative decoding is used in RL rollouts
  8. Draft-model fine-tuningKimi K3not statednot statedLKdraftMTP → draft
  9. Draft model
  10. SFTKimi K3not statednot statedSFTnot statedMXFP4 experts
  11. Cold-start model
  12. RL: nine expertsKimi K3not stated1MRLKimi K2.5’s, near GRPOnot stated
  13. Nine experts
  14. Where the student starts: not stated
  15. Multi-teacher on-policy distillationKimi K3not statednot statedMOPD9 teachersnot stated
  16. Kimi K3
  • tokens, log scale
  • derived: computed from stated numbers
  • context length
  • all parameters train
  • a part trains
  • all but a part trains
  • NTPdashed: concluded, no source names it
  • not stated by the sources
  • the architecture changes
  • checkpoint
  • checkpoint, weights published
  • a join the sources leave unexplained

How Kimi K3 was made: the stages on its path, top to bottom, and what changes from one to the next.

Of the 8 stages (7 on the path, 1 off it), the sources give or let us compute tokens for 0, and state or let us conclude the context length for 5, the objective for 8 and what is trained for 5; what is concluded is drawn dashed. The rest is hatched: nothing is filled in.

Select a stage for its numbers, its reason and the model as that stage trains it.