Fireworks AI has announced that Kimi 2.7 is now fully trainable on its inference infrastructure. This development allows users to fine-tune Kimi 2.7 with their own data, enabling the creation of custom models that can compete with frontier models at a reduced cost. The platform supports various training methods including Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Reinforcement Learning (RL), offering extensive customization options. AI
IMPACT Enables more accessible and cost-effective fine-tuning of existing models for specialized tasks.
RANK_REASON This is a platform announcement enabling training for an existing model, not a new model release or core research.
Read on X — Fireworks (inference infra) →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
