Package index • dragonfarm

Setup

dragon_check(): Check the Python environment and hardware
dragon_python_requirements(): Python requirements used by dragonfarm
dragon_presets(): Recommended small models

Data

dragon_dataset(): Create a dataset for fine-tuning
dragon_map(): Map dataset columns to prompt, response, and system text
dragon_preview(): Preview mapped rows as chat turns
dragon_split(): Hold out rows for evaluation
dragon_example_data(): Path to the bundled example dataset

Settings

dragon_lora(): LoRA settings
dragon_train_args(): Training settings
dragon_hardware(): Hardware settings

Training

dragon_train(): Fine-tune a model with LoRA
dragon_resume(): Resume a run from its latest checkpoint
dragon_wait(): Wait for a run to finish
dragon_cancel(): Cancel a run

Runs

dragon_run(): Reopen an existing run
dragon_runs(): List runs
dragon_runs_dir(): Directory where runs are stored
dragon_status() dragon_progress() dragon_logs(): Inspect a run
dragon_code(): R code that reproduces a run

After training

dragon_evaluate(): Evaluate a finished run
dragon_generate(): Generate replies from a fine-tuned model
dragon_merge(): Merge the adapter into the base model
dragon_export_gguf(): Export a merged model to GGUF

Cloud GPUs

dragon_bundle(): Package a run for a cloud GPU
dragon_remote(): Open a cloud GPU provider for a bundled run
dragon_import(): Import results trained on another machine
dragon_remote_providers(): Cloud GPU providers

App

dragon_app(): Launch the dragon-farm app