iac-repair-7b-gguf

Fine-tuned Qwen2.5-Coder-7B for Infrastructure-as-Code repair. Use with stackfix.

pass@1: 0.926 on Cloud-Gym benchmark (188 entries, 8 error categories)

Quick Start

pip install cloud-gym[gguf]

python -c "
from huggingface_hub import hf_hub_download
hf_hub_download('Tetsuto/iac-repair-7b-gguf', 'iac-repair-7b-q4.gguf', local_dir='.')
"

stackfix repair broken.tf --backend gguf --model iac-repair-7b-q4.gguf

Requirements

  • Disk: 4.5 GB (Q4) | RAM: ~8 GB | Speed: ~20 tok/s (CPU)
  • Best for: Premium, servers 8GB+
  • Runs on Linux, macOS, Windows. No GPU required.

Source

Downloads last month
183
GGUF
Model size
8B params
Architecture
qwen2
Hardware compatibility
Log In to add your hardware

16-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support