Generate text using LLMs
Generate text from a large language model like GPT.
This key transformation is responsible for interacting with LLMs like GPT and Claude. It features a prompt box which allows you to configure a prompt and inject variables using the {{ }} syntax. It also allows you to select which model to use and configure its parameters such as temperature.
You must provide an API key
To use a model from a provider, you must have added an API key for that provider. If you do not, you'll get an error when running the prompt.
Screenshot of the LLM transformation in the Notebook
Common errors
- Missing API key
- Prompt length exceeds maximum supports by model
- Rate-limited by the model provider
Schema
Inputs
| Name | Description | Type | Required |
|---|---|---|---|
| Prompt text | The prompt that will be sent to the LLM for completion | String | ✅ |
| Model | The large language model to use for completion | Enum | ✅ |
| Temperature | The temperature to specify to the LLM | Number | ❌ |
| History | An array of objects to pass in as history for chat-based completions. Syntax is { role: 'user', message: 'hello world' } | Array | ❌ |
Outputs
| Name | Description | Type |
|---|---|---|
| Prompt | The prompt that was sent to the LLM with variable substitution | String |
| Answe | The completion returned by the LLM | String |
Updated 4 months ago