SFT finetuning with Unsloth

Finetune an LLM with LoRA on a single GPU using Unsloth's optimized training.

Run

openmodal run examples/sft_finetune.py

With custom settings:

openmodal run examples/sft_finetune.py --model-name unsloth/Qwen3-4B --max-steps 1000 --lora-r 32

What it does

1. Provisions a GPU (H100 spot by default)
2. Downloads Qwen3-4B (4-bit quantized) and the FineTome-100k dataset
3. Applies LoRA adapters and trains with Unsloth's optimized kernels
4. Saves checkpoints and final model to persistent volumes
5. Auto-retries if the spot instance is preempted

The code

import openmodal

app = openmodal.App("sft-finetune")

train_image = (
    openmodal.Image.debian_slim()
    .uv_pip_install(
        "accelerate", "datasets", "peft",
        "transformers", "trl",
        "unsloth[cu128-torch270]",
    )
    .env({"HF_HOME": "/model_cache"})
)

model_cache = openmodal.Volume.from_name("sft-model-cache", create_if_missing=True)
checkpoints = openmodal.Volume.from_name("sft-checkpoints", create_if_missing=True)

@app.function(
    image=train_image,
    gpu="H100",
    volumes={"/model_cache": model_cache, "/checkpoints": checkpoints},
    timeout=6 * 60 * 60,
    retries=3,
)
def finetune(config):
    from unsloth import FastLanguageModel
    from trl import SFTTrainer

    model, tokenizer = FastLanguageModel.from_pretrained(
        model_name=config.model_name,
        load_in_4bit=True,
    )
    # ... LoRA setup, dataset loading, training ...
    trainer.train()
    model.save_pretrained("/checkpoints/final_model")

@app.local_entrypoint()
def main(model_name="unsloth/Qwen3-4B", max_steps=5):
    finetune.remote(TrainingConfig(model_name=model_name, max_steps=max_steps))

Features used

Feature	How it's used
gpu="H100"	Single GPU for training
Volume.from_name(create_if_missing=True)	Persistent storage for model weights and checkpoints
retries=3	Auto-retry on spot preemption
timeout=6*60*60	6 hour max training time
finetune.remote(config)	Runs training on the GPU
CLI args (--max-steps, --lora-r)	Tweak hyperparameters from command line