Harbor + Tinker: RL fine-tune a 30B model for a few hundred dollars
simonguozirui · x · 2026-11-11
Developer @GrdrVictor highlights a combo: Harbor framework turns one year old and Tinker's inkling 2 just launched — together they already let you run a small RL fine-tune on a 30B model for a few hundred dollars of compute. RL fine-tuning of large models is quickly reaching individual-developer budgets.
More from coding & agent
- Stanford's new CS329Z AI Agents course releases all materials publicly, lectures going live daily — ChengleiSi · 2026-11-11
- Google ships Gemini 4 Carbon with /swarm preview: one prompt, agents split work and check each other — OnlineInference · 2026-11-11
- Mitra adds MCP support so ChatGPT and Claude can make calls from your phone number — testingcatalog · 2026-11-11
- LangChain: evidence-graph GraphRAG lifts multi-source agent accuracy from 10% to 52% — JeremyCMorgan · 2026-11-11
- AI agents are finding critical CVEs — design your SaaS assuming the server is lost — haltakov · 2026-11-11
- Workday agent prototyping cut from 4-5 months to under a day with LangGraph — LangChain · 2026-11-11