Quickstart

Initial Setup

1. Install the Gretel CLI and Gretel Trainer either on your system or in your Notebook.

   # Command line installation
   pip install -U gretel-client gretel-trainer
   
   # Notebook installation
   !pip install -Uqq gretel-client gretel-trainer
   

2. Add your Gretel API key via the Gretel CLI.

   Use the Gretel client to store your API key to disk. This step is optional, the trainer will prompt for an API key in the next step.

   gretel configure
   

Train Synthetic Data

1. Train or fine-tune a model using the Gretel API.

   from gretel_trainer import trainer
   
   dataset = "https://gretel-public-website.s3-us-west-2.amazonaws.com/datasets/USAdultIncome5k.csv"
   
   model = trainer.Trainer()
   model.train(dataset)
   

2. Generate synthetic data!

   df = model.generate()
   

Conditional Data Generation

1. Load and preview the dataset, and set seed fields.

   # Load and preview the patient dataset
   import pandas as pd
   from gretel_trainer import trainer
   
   DATASET_PATH = 'https://gretel-public-website.s3.amazonaws.com/datasets/mitre-synthea-health.csv'
   SEED_FIELDS = ["RACE", "ETHNICITY", "GENDER"]
   
   print("\nPreviewing real world dataset\n")
   pd.read_csv(DATASET_PATH)
   

2. Train the model.

   # Train model
   model = trainer.Trainer()
   model.train(DATASET_PATH, seed_fields=SEED_FIELDS)
   

3. Conditionally generate data.

   # Conditionally generate data
   seed_df = pd.DataFrame(data=[
       ["black", "african", "F"],
       ["black", "african", "F"],
       ["black", "african", "F"],
       ["black", "african", "F"],
       ["asian", "chinese", "F"],
       ["asian", "chinese", "F"],
       ["asian", "chinese", "F"],
       ["asian", "chinese", "F"],
       ["asian", "chinese", "F"]
   ], columns=SEED_FIELDS)
   
   model.generate(seed_df=seed_df)