Getting Started

This guide walks you through account creation, dashboard orientation, and launching your first training job.

1. Create Your Account

Creating an account takes less than a minute and does not require a credit card. Every new account receives a $10 welcome credit to start training right away.

1. Navigate to /register and fill in your Full Name, Email Address, and a Password (minimum 8 characters). Confirm your password in the second field.
2. Click Create Account. A verification email is sent to your inbox immediately.
3. Open the verification email and click the activation link. This confirms your email and activates your account.
4. You are redirected to the login page. Enter your email and password to sign in.
5. On first login, the system automatically provisions a personal organization, a default workspace, and a wallet with $10 welcome credit. You land on your dashboard ready to start.

What if I do not receive the verification email? Check your spam folder first. If it is still missing, click Resend Verification on the login page or use thePOST /api/auth/resend-verification endpoint.

2. Dashboard Overview

The dashboard is your central control panel for managing all aspects of the USF BIOS platform. When you log in, you land on the Dashboard Overview page, which gives you an at-a-glance summary of your activity. The top section shows four key metric cards: yourWallet Balance, Active Jobs count, Datasets count, and Total Jobs completed. Below the metrics, you will see a table of yourRecent Training Jobs with their status, and a Quick Actions panel for common tasks.

The left sidebar provides navigation to every section of the platform:

• Dashboard: summary of recent jobs, dataset count, and wallet balance.
• Datasets: upload, preview, and manage training data.
• Training: create, monitor, and manage fine-tuning jobs.
• Models: browse 250+ supported base models.
• Deployments: serve fine-tuned or open models behind an OpenAI-compatible API.
• Billing: wallet balance, add credits, transaction history.
• Members: manage team access and roles.
• Settings: workspace and organization settings.

3. Your First Training Job

Here is a complete walkthrough for launching your first fine-tuning job on USF BIOS. The entire process takes about 5 minutes to set up.

1. Upload a dataset. Navigate to Datasets in the sidebar and click Upload Dataset. Select a JSONL, Parquet, or CSV file from your computer (max 200MB). The platform automatically validates the file format, detects columns, and generates a preview of the first rows. If your column names differ from the expected names (e.g., "question" instead of "instruction"), use the column mapping feature to align them.
2. Ensure you have credits. Navigate to Billing and check your wallet balance. New accounts start with $10 in welcome credits. If you need more, click Add Credits and enter the amount. You will be redirected to Stripe Checkout to complete the payment. Credits appear in your wallet immediately after payment.
3. Create the training job. Go to Training and click New Training Job. The 7-step wizard walks you through the entire configuration:
   • Step 1: Model. Search from 250+ base models. For your first job, try Llama 3.1 8B Instruct or Mistral 7B.
   • Step 2: Method. Select SFT (Supervised Fine-Tuning) for instruction-following tasks.
   • Step 3: Adapter. Choose LoRA for cost-effective training (trains less than 1% of parameters).
   • Step 4: Dataset. Select the dataset you uploaded earlier.
   • Step 5: GPU. The platform recommends an appropriate GPU tier based on your model size.
   • Step 6: Config. Use the default hyperparameters for your first run (learning rate, epochs, batch size).
   • Step 7: Review. Confirm settings and estimated cost, then click Launch.
4. Monitor progress. Once launched, the training detail page shows live metrics including training loss curves, evaluation metrics, streaming logs, and checkpoint snapshots. Jobs are billed per second of GPU compute, so you only pay for the time your model actually trains.

4. Next Steps

• Learn about dataset formats and column mapping in the Datasets section.
• Connect your HuggingFace account under Integrations to import private datasets.
• Set up auto top-up so your jobs never run out of credits.
• Explore the API Reference to integrate USF BIOS into your workflow.