BigHugger
GH Repository · Kiln-AI

Kiln

Build, Evaluate, and Optimize AI Systems. Includes evals, RAG, agents, fine-tuning, synthetic data generation, dataset management, MCP, and more.

stars
5,071
30-day movement
+124/day
Related entries
60
Connections
2
makeevalevaluationevalsPythonrlhfaichain-of-thoughtcollaborationmacoswindowsmlprompt-engineeringfine-tuningollamamcpevaluation-frameworkmachine-learningpythonsynthetic-datadataset-generationopenaipython/uvprompt

Kiln is a Python toolchain for building, evaluating, and optimizing AI systems, covering evals, RAG, agents, fine-tuning, synthetic data generation, dataset management, and MCP. It runs on macOS and Windows and works with providers and models such as OpenAI and Ollama.

Reach for it when you want one Python tool that ties together evaluation, dataset generation, and fine-tuning workflows for AI systems.

Use it to

  • Run structured evals on prompts and models
  • Generate synthetic datasets for training
  • Manage datasets for fine-tuning
  • Build agents with MCP support
  • Collaborate on prompt engineering

For ML engineers and teams evaluating and fine-tuning AI systems

Role
eval
Language
Python
Licence
custom (licence file present)
Forks
380
Open issues
24
Last push
2026-09-17
Latest release
0.1.0-alpha · 2024-10-25
topicsevalsfine-tuningsynthetic-datadataset-managementmcpprompt-engineering