Requirements:

Github Account: you need an active GitHub account to obtain $FML token.

Hardware: We will train multiple data models that demand significant computational resources with powerful GPUs. While leveraging your local system is possible, it's important to recognize that specific hardware and software configurations are required based on your OS. To streamline the process and ensure accessibility for all users, this guideline will focus on utilizing the IO GPU cluster.

Obtain FML

1. Navigate to the Faucet tab

2. Connect your GitHub account

3. FML will send to you address

Stake FML as Training Node

1. Navigate to the Training Node tab on the Stake page

2. Select chainbase task

3. Stake 10 FML

Get API key

Your API key is required for submitting data models.

Get Hugging Face Token:

1. Create a hugging face account on its site

2. Navigate to Hugging Face > Profile > Settings > Access Tokens and create a write token

3. Save the token

   

   Acknowledge Gemma license

   1. Go to Gemma's page on hugging face

   2. Click on Acknowledge license

      

   Deploy IO cluster

   1. Go to IO cloud site

   2. Follow this path and deploy a cluster

      Ray > General > Render > All countries > High speed > RTX 3070 > Deploy Cluster

      

Access cluster

1. Navigate to the cluster's page

2. Click on Visual Studio and use the IDE Password to access the server

Run the commands

Open the Visual Studio terminal and run the below commands

# clone the script repository
git clone https://github.com/FLock-io/testnet-training-node-quickstart.git
cd testnet-training-node-quickstart

Set up the environment (copy & paste these commands one by one.)

conda create -n training-node python==3.10
conda activate training-node
pip install -r requirements.txt

Get information required to run automation script

FLOCK_API_KEY - available at https://train.flock.io/flock_api

HF_TOKEN - your Hugging Face Access Token

HG_USERNAME - your Hugging Face username

Run full automation script

Once you have all the information listed in the previous step, paste the script below into your terminal, update the respective values, and run the command.

TASK_ID=10 FLOCK_API_KEY="your-flock-api-key-stakes-as-node-for-the-task" HF_TOKEN="your-hf-token" CUDA_VISIBLE_DEVICES=0 HF_USERNAME="your-hf-user-name" python full_automation.py

Congratulations you participated in the Flock AI Arena task successfully.

Terminate cluster

you don't need the cluster anymore and you can terminate it.

Claim rewards

The $FML Token distribution is triggered every 24 hours at midnight UTC. The final round of reward distribution is triggered once the task training period is complete.