chrisvoncsefalvay/funsloth

Streamlines GPU cloud computing by automating RunPod instance management, deployment, and machine learning workflow orchestration.

Validates and analyzes datasets for Unsloth fine-tuning, checking format, tokens, Chinchilla optimality, and training readiness.

Manages local GPU training for Unsloth, validating CUDA, selecting GPUs, monitoring progress, and handling checkpoints with ease.

Trains machine learning models for sloth behavior prediction using advanced neural network architectures and behavioral datasets.

Launches fine-tuning jobs on Hugging Face cloud GPUs with hardware selection, cost estimation, and optional WandB monitoring.

Uploads fine-tuned models to Hugging Face Hub with comprehensive model cards, export formats, and optional GGUF quantization.

Complete workflow for fine-tuning LLMs with Unsloth: dataset validation, notebook generation, training management, and model deployment

Claude Code skills marketplace for fine-tuning language models with Unsloth