AnalysisAI Models
16 days ago
Mostaque: Kimi post-training may use compute multiple of pretraining
Emad
@emostaqueBuilding first principles, sovereign AI @ii_posts. Founder @StabilityAI. Consistent inference is possible.
Londonii.inc
Emad Mostaque
@EMostaque
It’ll be interesting to see if the post training for this uses a multiple of the compute of pretraining as cursor did when they tuned Kimi as the base model
·
16 days ago