| Options of Pydantic | Description |
| Knowledge validation | Verifies that enter knowledge conforms to anticipated sorts (e.g., int, str, checklist) and may apply customized guidelines. |
| Automated sort conversion | Converts suitable knowledge sorts robotically, akin to turning “2023-10-17” right into a datetime.date. |
| Knowledge serialization | Fashions can serialize knowledge into codecs like JSON, simplifying interactions with APIs. |
| Default values | Fields will be optionally available or have default values, offering flexibility in dealing with inputs. |
Instance of Pydantic Fashions
Let’s make a UserModel that inherits from BaseModel from Pydantic and the instantiated class ought to have an “id” as int, identify as “str” and e-mail as “e-mail”
from pydantic import BaseModel
class UserModel(BaseModel):
id: int
identify: str
e-mail: str
# Legitimate enter
valid_user = UserModel(id=1, identify="Vidhya", e-mail="[email protected]")
print(valid_user)id=1 identify="Vidhya" e-mail="[email protected]"
# Invalid enter (raises a validation error)
attempt:
invalid_user = UserModel(id="one", identify="Vidhya", e-mail="[email protected]")
besides ValueError as e:
print(f"Validation Error: {e}")Validation Error: 1 validation error for UserModel
id
Enter must be a sound integer, unable to parse string as an integer
[type=int_parsing, input_value="one", input_type=str]
For additional info go to https://errors.pydantic.dev/2.9/v/int_parsing
We obtained a validation error after we tried passing a string in id and Pydantic additionally returns the hyperlink to documentation that helps us perceive the error.
Now, let’s take a look at different options of Pydantic, akin to optionally available, date, and default worth.
from pydantic import BaseModel
from typing import Elective
from datetime import date
class EventModel(BaseModel):
event_name: Elective[str] = None # It is Elective to supply a price
event_loc: str = "India" # Default worth
event_date: date
# Automated conversion of a string to a `date` object
occasion = EventModel(event_date="2024-10-17")
print(occasion)event_name=None event_loc="India" event_date=datetime.date(2024, 10, 17)
We have now handed solely the event_date, but it’s not throwing any error as a result of we set event_name to Elective and event_loc to “India” by default. Additionally, discover that the date is being typecast from string to datetime.
Installations
We shall be utilizing CrewAI all through this information, so be sure you set up CrewAI.
!pip set up crewaiStructuring Inputs in Agentic Techniques
The inputs will be formatted whereas defining the Agent and Activity utilizing curly braces { } with a variable identify inside. Setting the human_input=True will immediate the consumer for suggestions for the output. Let’s outline an agent and activity to reply questions associated to physics:
Be aware: I’ll be utilizing the ‘gpt-4o-mini-2024-07-18’ from openai all through the information.
from crewai import Agent, Activity, Crew
import os
os.environ['OPENAI_API_KEY']=''
llm_config={'mannequin':'gpt-4o-mini-2024-07-18'}
teacher_agent = Agent(
position="Physics Professor",
purpose="You've robust foundational data on physics "
"You give one of the best reply to your college students",
backstory=(
"You're employed as a university professor"
"You train physics and make clear doubts {doubts}"
"You reply the query solely whether it is associated to physics"
"You just remember to present one of the best solutions!"
),
allow_delegation=False,
verbose=True)
doubt_clarification_task = Activity(
description=(
"Tackle the physics doubts raised by the coed {doubts}. "
"You do not reply the query if the query just isn't associated to physics"
"Present clear, detailed explanations, examples, and any needed formulation to assist the coed perceive the ideas."
),
expected_output=(
"A transparent response to the coed's doubts, "
"together with explanations and examples if wanted. "
),
agent=teacher_agent,
human_input=True
)
crew = Crew(
brokers=[teacher_agent],
duties=[doubt_clarification_task],
verbose=True,
reminiscence=False
)The inputs will be handed to the “inputs “ parameter in crew.kickoff()
inputs = {
"doubts": "What's thermodynamics",
}
end result = crew.kickoff(inputs=inputs)Output
# Agent: Physics Professor## Activity: Tackle the physics doubts raised by the coed What's
thermodynamics. You do not reply the query if the query just isn't associated
to physicsProvide clear, detailed explanations, examples, and any needed
formulation to assist the coed perceive the ideas.# Agent: Physics Professor
## Last Reply:
Thermodynamics is the department of physics that offers with the relationships
between warmth, work, temperature, and power. It supplies a framework for
understanding how power is transferred and remodeled in programs, whether or not
they're mechanical engines, fridges, or organic organisms.The examine of thermodynamics relies on 4 basic legal guidelines that describe
how power strikes and adjustments type:1. **Zeroth Legislation of Thermodynamics**: This legislation establishes the idea of
temperature. It states that if two programs are in thermal equilibrium with a
third system, then they're additionally in thermal equilibrium with one another. For
instance, if System A is in equilibrium with System C, and System B can also be
in equilibrium with System C, then System A and System B are in equilibrium
with one another. This enables us to outline temperature as a measurable entity.2. **First Legislation of Thermodynamics**: Typically said because the legislation of power
conservation, this legislation asserts that power can't be created or destroyed,
solely remodeled. Mathematically, it may be expressed as:[
Delta U = Q - W
]
the place ( Delta U ) is the change in inner power of the system, ( Q
) is the warmth added to the system, and ( W ) is the work performed by the
system. An instance of the primary legislation in follow is a steam engine, the place warmth
power is transformed into work.3. **Second Legislation of Thermodynamics**: This legislation introduces the idea of
entropy, stating that in any power switch, the entire entropy of an
remoted system can by no means lower over time. It means that power
transformations will not be 100% environment friendly and that programs naturally progress
towards a state of dysfunction. A typical instance is the concept that warmth flows
spontaneously from a warmer physique to a colder physique, not the reverse.4. **Third Legislation of Thermodynamics**: This legislation states that because the temperature
of a system approaches absolute zero, the entropy of an ideal crystal
approaches zero. This means that it's unimaginable to succeed in absolute zero
in a finite variety of steps, an idea that underscores the elemental
limits of thermodynamic processes.In sensible functions, thermodynamics is utilized in quite a few fields akin to
engineering (designing engines, fridges), chemistry (response
spontaneity), and environmental science (understanding local weather programs).In abstract, thermodynamics is important for analyzing how power is utilized
and transferred, offering insights into bettering effectivity and
understanding pure processes in varied bodily programs.## Last End result: Thermodynamics is the department of physics that offers with the
relationships between warmth, work, temperature, and power. It supplies a
framework for understanding how power is transferred and remodeled in
programs, whether or not they're mechanical engines, fridges, or organic
organisms.The examine of thermodynamics relies on 4 basic legal guidelines that describe
how power strikes and adjustments type:1. **Zeroth Legislation of Thermodynamics**: This legislation establishes the idea of
temperature. It states that if two programs are in thermal equilibrium with a
third system, then they're additionally in thermal equilibrium with one another. For
instance, if System A is in equilibrium with System C, and System B can also be
in equilibrium with System C, then System A and System B are in equilibrium
with one another. This enables us to outline temperature as a measurable
entity.2. **First Legislation of Thermodynamics**: Typically said because the legislation of power
conservation, this legislation asserts that power can't be created or destroyed,
solely remodeled. Mathematically, it may be expressed as:[
Delta U = Q - W
]
the place ( Delta U ) is the change in inner power of the system, ( Q
) is the warmth added to the system, and ( W ) is the work performed by the
system. An instance of the primary legislation in follow is a steam engine, the place
warmth power is transformed into work.3. **Second Legislation of Thermodynamics**: This legislation introduces the idea of
entropy, stating that in any power switch, the entire entropy of an
remoted system can by no means lower over time. It means that power
transformations will not be 100% environment friendly and that programs naturally progress
towards a state of dysfunction. A typical instance is the concept that warmth flows
spontaneously from a warmer physique to a colder physique, not the reverse.4. **Third Legislation of Thermodynamics**: This legislation states that because the temperature
of a system approaches absolute zero, the entropy of an ideal crystal
approaches zero. This means that it's unimaginable to succeed in absolute zero
in a finite variety of steps, an idea that underscores the elemental
limits of thermodynamic processes.In sensible functions, thermodynamics is utilized in quite a few fields akin to
engineering (designing engines, fridges), chemistry (response
spontaneity), and environmental science (understanding local weather programs).In abstract, thermodynamics is important for analyzing how power is utilized
and transferred, offering insights into bettering effectivity and
understanding pure processes in varied bodily programs.=====
## Please present suggestions on the Last End result and the Agent's actions:
Give the reply in 3 traces
# Agent: Physics Professor
## Last Reply:
Thermodynamics is the department of physics that offers with the relationships
between warmth, work, temperature, and power. It's ruled by 4
basic legal guidelines that describe how power is transferred and transformed in
bodily programs. For instance, the primary legislation of thermodynamics states that
power can't be created or destroyed, solely remodeled, which is expressed
mathematically as ΔU = Q - W, the place ΔU is the change in inner power, Q
is the warmth added to the system, and W is the work performed by the system.
We handed a query within the enter, “What’s thermodynamics?” and with the assistance of a human enter, we obtained the specified reply.
Structuring Outputs in Agentic Techniques
Let’s make brokers who might help me fill out particulars like identify, e-mail, cellphone quantity, and job one after the other. Structuring the outputs as Pydantic or json helps outline what output we’re anticipating and the format we anticipate in order that the following brokers get structured knowledge and context.
Be aware: In a case like mine, not structuring the outputs with anticipated fields may result in a lack of info, and the following brokers may begin hallucinating.
from crewai import Agent, Activity, Crew
import os
os.environ['OPENAI_API_KEY']=''
llm_config={'mannequin':'gpt-4o-mini-2024-07-18'}Let’s outline the courses utilizing pedantic with the fields I anticipate within the output. As defined within the Pydantic fashions’ part, you should use Elective or set a price by default if wanted.
from pydantic import BaseModel
from typing import Checklist
# Outline the courses utilizing Pydantic
class nameEmail(BaseModel):
identify: Checklist[str]
e-mail: Checklist[str]
class phoneNo(BaseModel):
identify: str
e-mail: str
cellphone: int
class allDetails(BaseModel):
identify: str
e-mail: Checklist[str]
cellphone: int
job: strLet’s outline the primary agent and activity that fills out the consumer’s identify and e-mail. Use the “output_pydantic” parameter in Activity(). I handed the nameEmail class as I received’t be utilizing identify and e-mail as output right here.
# Activity 1: Agent for filling identify and e-mail utilizing output_pydantic
name_email_agent = Agent(
position="Knowledge Entry Specialist",
purpose="Your activity is to fill out the consumer's identify and e-mail by yourself.",
backstory=(
"You focus on filling out private info like identify and e-mail."
"You fill the identify and e-mail by yourself"
"You utilize uncommon names and use beginning yr within the e-mail"
),
allow_delegation=False,
verbose=True
)
name_email_task = Activity(
description="Fill out the identify and e-mail of the consumer.",
expected_output="Title and e-mail.",
agent=name_email_agent,
output_pydantic=nameEmail,
human_input=False
)Let’s use the opposite technique to construction the output through the use of the “output_json” parameter in Activity() which provides the output in json format.
# Activity 2: Agent for filling cellphone quantity utilizing output_json
phone_number_agent = Agent(
position="Contact Data Supervisor",
purpose="Your activity is to fill out a random 10 digit cellphone quantity by yourself.",
backstory=(
"You're an knowledgeable in managing contact particulars, particularly cellphone numbers."
),
allow_delegation=False,
verbose=True
)
phone_number_task = Activity(
description="Fill out the consumer's cellphone quantity.",
expected_output="A JSON format with the consumer's cellphone quantity.",
agent=phone_number_agent,
output_json=phoneNo,
human_input=False
)Right here, the agent is designed to fill out the job particulars for the consumer. The ultimate output will include all of the identify, e-mail, cellphone quantity, and job position info as pedantic and saved in output.txt utilizing the “output_file” parameter.
# Activity 3: Agent for filling job description utilizing output_file
job_description_agent = Agent(
position="Job Decider",
purpose="Your activity is to randomly determine a job for the consumer",
backstory=(
"You deal with assigning job roles randomly."
),
allow_delegation=False,
verbose=True
)
job_description_task = Activity(
description="Fill out the consumer's job description.",
expected_output="Show identify, e-mail, cellphone no and job description.",
agent=job_description_agent,
output_pydantic=allDetails,
output_file="/content material/output.txt",
human_input=False
)
# Outline the crew to run all duties
crew = Crew(
brokers=[name_email_agent, phone_number_agent, job_description_agent],
duties=[name_email_task, phone_number_task, job_description_task],
verbose=True,
reminiscence=False
)
end result = crew.kickoff()# Agent: Knowledge Entry Specialist## Activity: Fill out the identify and e-mail of the consumer.
# Agent: Knowledge Entry Specialist
## Last Reply:
Title: Elowen Thorne
E mail: [email protected]
# Agent: Contact Data Supervisor
## Activity: Fill out the consumer's cellphone quantity.
# Agent: Contact Data Supervisor
## Last Reply:
{
"identify": "Elowen Thorne",
"e-mail": "[email protected]",
"phone_number": "1234567890"
}
# Agent: Job Decider
## Activity: Fill out the consumer's job position.
# Agent: Job Decider
## Last Reply:
Title: Elowen Thorne
E mail: [email protected]
Telephone No: 1234567890
Job Function: Knowledge Analyst
That is the output file that was created :
You can even take a look at every activity’s output:
job_description_task.outputTaskOutput(description="Fill out the consumer's job position.", identify=None,
expected_output="Show identify, e-mail, cellphone no and job position.", abstract="Fill
out the consumer's job position....", uncooked='Title: Elowen Thorne nEmail:
[email protected] nPhone No: 1234567890 nJob Function: Knowledge
Analyst', pydantic=allDetails(identify="Elowen Thorne", e-mail=
['[email protected]'], cellphone=1234567890, job='Knowledge Analyst'),
json_dict=None, agent="Job Decider", output_format=<OutputFormat.PYDANTIC:
'pydantic'>)
I obtained the anticipated output, and as a result of structuring the intermediate outputs, the small print remained fixed as every activity progressed.
Additionally, to grasp the Agent AI higher, discover: The Agentic AI Pioneer Program.
Conclusion
This text explored the significance of structuring inputs and outputs in multi-agent utilizing instruments like Pydantic fashions and CrewAI. By organizing knowledge successfully and validating inputs, we will considerably improve the efficiency and reliability of multi-agent programs. Structured outputs assist keep consistency and stop errors akin to knowledge loss and hallucinations, guaranteeing that brokers can collaborate effectively. Implementing these methods permits builders to create extra strong agentic programs able to dealing with complicated duties with larger precision and reliability.
Continuously Requested Questions
Ans. Agent-based programs contain a number of brokers working collectively to resolve issues, talk, and collaborate, offering extra strong options than single programs like LLMs.
Ans. CrewAI is a framework that helps agentic programs, permitting brokers to work collectively to resolve duties. On this article, we use CrewAI with pedantic fashions to construction outputs and guarantee correct knowledge administration between brokers.
Ans. One technique to enter a picture in CrewAI is by assigning the URL of the picture to a variable after which passing it to the inputs parameter in crew.kickoff().
Ans. Pydantic fashions are Python objects used for knowledge validation and serialization. They assist make sure that inputs and outputs between brokers are correctly structured, resulting in extra dependable outcomes in agent-based programs.
Ans. Outputs are structured by defining the anticipated fields (like identify, e-mail, and so forth.) utilizing Pydantic fashions, which ensures that the subsequent agent receives legitimate and well-formatted knowledge for processing.
