claw-finetune

What problem does it solve?

Collect fine-tuning requirements, validate datasets and API keys, generate backend-specific LoRA SFT runs for Tinker or HPC-AI SDK, launch the training, and answer status questions from persisted run artifacts.

Core Features & Use Cases

• Supports local and shared/common datasets, with adaptation paths for built-in presets gsm8k, knights-and-knaves, no_robots, and tulu-3-sft-mixture.
• Lets you compare Tinker and HPC-AI LoRA/SFT flows, generates training scripts, and launches runs.
• Automatically validates datasets, API keys, and endpoint configurations to ensure secure, reproducible runs.

Quick Start

Fine-tune a model by providing a dataset and backend choice, and I will generate, launch, and monitor the LoRA/SFT run.