together-fine-tuning

What problem does it solve?

Fine-tune open-source LLMs on Together AI with an end-to-end workflow, enabling customization and adaptation of models to your data and use cases.

Core Features & Use Cases

• LoRA and Full Fine-tuning for a range of model sizes
• DPO (preference) fine-tuning to optimize outputs
• VLM (vision-language) fine-tuning for image+text tasks
• Reasoning fine-tuning to inject chain-of-thought capabilities
• Function-calling fine-tuning to improve tool integration
• BYOM (Bring Your Own Model) for deploying private models
• End-to-end pipeline including data preparation, uploading, training, monitoring, evaluation, and deployment
• Runnable Python scripts in scripts/ demonstrating complete workflows

Quick Start

Run the provided finetune_workflow.py script to execute a complete fine-tuning workflow from data preparation to deployment.