Operating system#

• Ubuntu 22.04: Ensure your system is running Ubuntu version 22.04.

Hardware#

• AMD Instinct™ GPUs: This tutorial was tested on an AMD Instinct MI300X GPU. Ensure you are using an AMD Instinct GPU or compatible hardware with ROCm support and that your system meets the official requirements.

Software#

• ROCm 6.3: Install and verify ROCm by following the ROCm install guide. After installation, confirm your setup using:

  rocm-smi
  

  This command lists your AMD GPUs with relevant details, similar to the image below.

  

• Docker: Ensure Docker is installed and configured correctly. Follow the Docker installation guide for your operating system.

  Note: Ensure the Docker permissions are correctly configured. To configure permissions to allow non-root access, run the following commands:

  sudo usermod -aG docker $USER
  newgrp docker
  

  Verify Docker is working correctly:

  docker run hello-world
  

Hugging Face API access#

• Obtain an API token from Hugging Face for downloading models.

• Ensure the Hugging Face API token has the necessary permissions and approval to access the Meta Llama checkpoints.

Data preparation#

• This tutorial uses a sample dataset from Hugging Face, which is prepared during the setup steps.

1. Pull the Docker image#

Ensure your system meets the system requirements.

Pull the Docker image required for this tutorial:

docker pull rocm/pytorch-training:latest

2. Launch the Docker container#

Launch the Docker container and map the necessary directories. Replace /path/to/notebooks with the full path to the directory on your host machine where these notebooks are stored.

docker run -it --rm \
  --network=host \
  --device=/dev/kfd \
  --device=/dev/dri \
  --group-add=video \
  --ipc=host \
  --cap-add=SYS_PTRACE \
  --security-opt seccomp=unconfined \
  --shm-size 8G \
  -v $(pwd):/workspace \
  -w /workspace/notebooks \
  rocm/pytorch-training:latest

Note: This command mounts the current directory to the /workspace directory in the container. Ensure the notebook file is either copied to this directory before running the Docker command or uploaded into the Jupyter Notebook environment after it starts. Save the token or URL provided in the terminal output to access the notebook from your web browser. You can download this notebook from the AI Developer Hub GitHub repository.

3. Install and launch Jupyter#

Inside the Docker container, install Jupyter using the following command:

pip install jupyter

Start the Jupyter server:

jupyter-lab --ip=0.0.0.0 --port=8888 --no-browser --allow-root

Note: Ensure port 8888 is not already in use on your system before running the above command. If it is, you can specify a different port by replacing --port=8888 with another port number, for example, --port=8890.

4. Install the required libraries#

Install the libraries required for this tutorial. Run the following commands inside the Jupyter notebook running within the Docker container:

# Install PyTorch, torchvision, torchao nightlies
!pip install --pre --upgrade torch torchvision torchao --index-url https://download.pytorch.org/whl/nightly/rocm6.3/
!pip install --pre --upgrade torchtune --extra-index-url https://download.pytorch.org/whl/nightly/rocm6.3/ 
# This note book is verified under torch==2.7.0.dev20250302+rocm6.3, torchao==0.10.0.dev20250303+rocm6.3,  torchvision==0.22.0.dev20250302+rocm6.3, torchtune==0.6.0.dev20250302+rocm6.3

Verify the installation:

# Verify the installation and version of the required libraries
!pip list | grep torch

Here is the expected output:

pytorch-triton-rocm     3.2.0+git4b3bb1f8
torch                   2.7.0.dev20250302+rocm6.3
torchdata               0.11.0
torchtune               0.6.0.dev20250302+rocm6.3

5. Verify torchtune for ROCm 6.3#

To confirm that the package is installed correctly, run the following command:

!tune --help

You should see the following output:

usage: tune [-h] {download,ls,cp,run,validate,cat} ...

Welcome to the torchtune CLI!

options:
  -h, --help            show this help message and exit

subcommands:
  {download,ls,cp,run,validate,cat}
    download            Download a model from the Hugging Face Hub or Kaggle
                        Model Hub.
    ls                  List all built-in recipes and configs
    cp                  Copy a built-in recipe or config to a local path.
    run                 Run a recipe. For distributed recipes, this supports
                        all torchrun arguments.
    validate            Validate a config and ensure that it is well-formed.
    cat                 Pretty print a config, making it easy to know which
                        parameters you can override with `tune run`.

⚠️ Important: ensure the correct kernel is selected

If the verification process fails, ensure the correct Jupyter kernel is selected for your notebook. To change the kernel, follow these steps:

1. Go to the Kernel menu.

2. Select Change Kernel.

3. Select Python 3 (ipykernel) from the list.

Important: Failure to select the correct kernel can lead to unexpected issues when running the notebook.

6. Provide your Hugging Face token#

You’ll require a Hugging Face API token to access Llama-3.1. Generate your token at Hugging Face Tokens and request access for Llama-3.1 8B. Tokens typically start with “hf_”.

Run the following interactive block in your Jupyter notebook to set up the token:

Note: Uncheck the “Add token as Git credential” option.

from huggingface_hub import notebook_login, HfApi

# Prompt the user to log in
notebook_login()

Verify that your token was accepted correctly:

from huggingface_hub import HfApi

try:
    api = HfApi()
    user_info = api.whoami()
    print(f"Token validated successfully! Logged in as: {user_info['name']}")
except Exception as e:
    print(f"Token validation failed. Error: {e}")

Set and verify the GPU availability#

Begin by specifying the GPUs available for fine-tuning. Verify that they are properly detected by PyTorch.

import os
import torch
gpus= [0, 1] # Rank 0 is for MI300x single device finetune, and Rank 0/1 for full 
os.environ.setdefault("CUDA_VISIBLE_DEVICES", ','.join(map(str, gpus)))
# Ensure PyTorch detects the GPUs correctly
print(f"PyTorch detected number of available devices: {torch.cuda.device_count()}") 

Download the Llama model#

Run the following command to download the weights to your local machine. This also downloads the tokenizer model and a responsible use guide.

To download Llama-3.1, use the following command:

!tune download meta-llama/Meta-Llama-3.1-8B-Instruct \
    --output-dir /tmp/Meta-Llama-3.1-8B-Instruct \
    --ignore-patterns "original/consolidated.00.pth"

Run the fine-tuning recipes#

Use these fine-tuning recipes for single GPU or distributed training.

!tune run lora_finetune_single_device --config llama3_1/8B_lora_single_device

!tune run --nproc_per_node 2 full_finetune_distributed --config llama3_1/8B_full

Customize your recipes for Llama#

There are two ways to modify configurations.

!tune run lora_finetune_single_device \
    --config llama3_1/8B_lora_single_device \
    batch_size=16 \
    enable_activation_checkpointing=False 

!tune cp llama3_1/8B_lora_single_device ./my_custom_config_llama3_1_8B_lora_single_device.yaml
# Copied to ./my_custom_config_llama3_1_8B_lora_single_device.yaml

!tune run lora_finetune_single_device --config ./my_custom_config_llama3_1_8B_lora_single_device.yaml

Custom datasets#

torchtune supports fine-tuning on a variety of different datasets, including instruct-style, chat-style, preference datasets, and more. To learn more about how to apply these components to fine-tune on your own custom dataset, see the provided links along with the torchtune API docs.

Monitoring GPU memory#

To monitor GPU memory during training, run the following command in a terminal:

!rocm-smi

This command displays memory usage and other GPU metrics to ensure your hardware resources are being optimally used.