Tinker
A model fine-tuning platform offering a managed LoRA training API, supporting rapid iteration on adapters in the cloud without copying the entire model.
Timeline (last 90 days)
The project uses .NET Minimal API to wrap the fine-tuned model as a service, including API key authentication, rate limiting, exception handling, integration with the Tinker inference API, and a browser-side playground.
The project uses Tinker to perform supervised fine-tuning of Qwen3-8B.
The author used Tinker's LoRA training API to fine-tune Qwen3.5-4B on 5,755 Open Food Facts entries; the training set contains 5,000 ordinary products, 550 refusal samples, and 205 control samples.
Thinking Machines released the Dimuon optimizer, a Muon variant designed specifically for LoRA fine-tuning, and it has been released in Tinker.
The custom role Tinker can perform 'pack a punch' upgrades on items held by other players
All Facts (5)
The project uses .NET Minimal API to wrap the fine-tuned model as a service, including API key authentication, rate limiting, exception handling, integration with the Tinker inference API, and a browser-side playground.
50%UnverifiedThe project uses Tinker to perform supervised fine-tuning of Qwen3-8B.
50%UnverifiedThe author used Tinker's LoRA training API to fine-tune Qwen3.5-4B on 5,755 Open Food Facts entries; the training set contains 5,000 ordinary products, 550 refusal samples, and 205 control samples.
50%UnverifiedThe custom role Tinker can perform 'pack a punch' upgrades on items held by other players
50%UnverifiedThinking Machines released the Dimuon optimizer, a Muon variant designed specifically for LoRA fine-tuning, and it has been released in Tinker.
50%