huggingface
A small language model built to train and run Korean character chatbots on personal hardware.
SITE CAPTUREBased on information published on the service's website. Selection and review policy
WHAT IT DOES
What does it do?
The model is smaller and focused on Korean so it can be trained and run on personal hardware. Fine-tuning with example conversations shapes a character’s speaking style and personality; the base model has no predefined character. Instructions are provided for running it with llama.cpp, vLLM, Ollama, and other tools.
KEY FEATURES
Key features
- Run a model for Korean character conversations on personal hardware
- Fine-tune a character’s personality and speaking style with example data
- Revise data and check results repeatedly in short training cycles
- Run the model with llama.cpp, vLLM, Ollama, and other tools