Train an LLM

After generating your LLM Dataset, you're ready to train your custom language model. This guide walks you through the training process.

Setup Process

1. Create a new Canvas
2. Drag the LLM Trainer Element onto your Canvas
3. Open the element settings by clicking the ... button in the corner

Configure Training Settings

Configure the following parameters to prepare your training:

1. Base Model Architecture

   The default model works well for creating a model quickly.

   Want more options? webAI supports a variety of base models for different use cases. Check out our Supported LLM Base Models page to learn more about all available models.
   
   Supported LLM Base Models

2. Dataset Folder Path

   Using the Select Directory button, choose the folder where you saved your LLM dataset during the Dataset Generation step.

   Need a dataset? Try one of our samples:
   
   Sci-Fi Novels
   
   Logistics Warehouse Management

3. Artifact Save Path

   Using the Select Directory button, choose the folder where you want to save your trained adapter.

4. Base Model Assets Path

   Using the Select Directory button, specify the folder where you want to save your base model.

5. Evaluator API Key

   Add a Groq, OpenAI, Claude, or Gemini API key to enable the Faithfulness and Relevancy benchmarks in your training metrics.

   If using Groq, a single API key is sufficient. If using other providers (OpenAI, Claude, or Gemini), you must provide at least two different API keys.
   
   You can get a free Groq key here:
   
   Groq API Key

6. Batch Size

   Recommended setting: 4 for testing

   For all other settings, the default values work well for most use cases.

Start Training

1. Click the Run button to begin the training process

   The first time you run this flow, dependencies will be installed, which may take some time.

2. Training progress will be displayed in the element

Next Steps

Once training completes, you can use your custom LLM in various workflows or with the Document QnA element.