This site is not available on Mobile. Please return on a desktop browser.
Visit our main site at guardrailsai.com
This site is not available on Mobile. Please return on a desktop browser.
Visit our main site at guardrailsai.com
| Developed by | Guardrails AI |
| Date of development | Feb 15, 2024 |
| Validator type | Format |
| Blog | |
| License | Apache 2 |
| Input/Output | Output |
This validator can perform the following checks:
$ guardrails hub install hub://guardrails/valid_length
In this example, we verify that an LLM generated response contains anywhere from 3-6 characters.
# Import Guard and Validator
from guardrails import Guard
from guardrails.hub import ValidLength
# Setup Guard
guard = Guard().use(
ValidLength, min=3, max=6, on_fail="exception"
)
response = guard.validate("hello") # Validator passes
try:
response = guard.validate("hello world!") # Validator fails
except Exception as e:
print(e)
Output:
Validation failed for field with errors: Value has length greater than 6. Please return a shorter output, that is shorter than 6 characters.
This example applies the validator to a list of a JSON object, and ensures that the length of the list is within an expected range.
# Import Guard and Validator
from pydantic import BaseModel, Field
from guardrails.hub import ValidLength
from guardrails import Guard
val = ValidLength(min=1, max=2, on_fail="exception")
# Create Pydantic BaseModels
class ProductInfo(BaseModel):
"""Information about a single product."""
product_name: str = Field(description="Name of the product")
product_summary: str = Field(description="A summary of the product")
class ProductCategory(BaseModel):
"""List of products."""
category_name: str = Field(description="Name of product category")
products: list[ProductInfo] = Field(
description="List of products", validators=[val]
)
# Create a Guard to check for valid Pydantic output
guard = Guard.from_pydantic(output_class=ProductCategory)
# Run LLM output generating JSON through guard
guard.parse(
"""
{
"category_name": "Hair care",
"products": [
{
"product_name": "Hair spray",
"product_summary": "Helps your styled hair stay in place."
},
{
"product_name": "Shampoo",
"product_summary": "Helps clean your hair."
}
]
"""
)
try:
# Run LLM output generating JSON through guard
guard.parse(
"""
{
"category_name": "Hair care",
"products": [
{
"product_name": "Hair spray",
"product_summary": "Helps your styled hair stay in place."
},
{
"product_name": "Shampoo",
"product_summary": "Helps clean your hair."
},
{
"product_name": "Conditioner",
"product_summary": "Helps condition your hair."
}
]
}
"""
)
except Exception as e:
print(e)
Output:
Validation failed for field with errors: Value has length greater than 2. Please return a shorter output, that is shorter than 2 characters.
__init__(self, min=None, max=None, on_fail="noop")
Initializes a new instance of the Validator class.
Parameters
min (int): Min expected length of the object (str, list).max (int): Max expected length of the object (str, list).on_fail (str, Callable): The policy to enact when a validator fails. If str, must be one of reask, fix, filter, refrain, noop, exception or fix_reask. Otherwise, must be a function that is called when the validator fails.__call__(self, value, metadata={}) -> ValidationResult
Validates the given value using the rules defined in this validator, relying on the metadata provided to customize the validation process. This method is automatically invoked by guard.parse(...), ensuring the validation logic is applied to the input data.
Note:
guard.parse(...) where this method will be called internally for each associated Validator.guard.parse(...), ensure to pass the appropriate metadata dictionary that includes keys and values required by this validator. If guard is associated with multiple validators, combine all necessary metadata into a single dictionary.Parameters
value (Any): The input value to validate.metadata (dict): A dictionary containing metadata required for validation. No additional metadata keys are needed for this validator.The validator playground is available to authenticated users. Please log in to use it.