TypoX Community

We’re excited to announce that Sprint Round 2 is now live!

🚀 Don’t miss out on the chance to compete and win amazing rewards. Join us and showcase your skills!
👉 Participate here: https://zealy.io/cw/typoxai/questboard

🤗 If you have any questions, feel free to DM us in our Telegram group: https://t.me/TypoGraphyAI

And big congratulations to the winners of TypoX AI Zealy Sprint Round 1!

A huge thank you to everyone who participated. We have distributed the rewards on BNB Chain based on the rankings as follows, please check your wallet on BNB Chain Network:

🥇 1st Place: $50
🥈 2nd Place: $30
🥉 3rd Place: $10
🎖️ 4th-25th Place: $5 each

c57c364c-1635-40d5-832f-4312892cd807-image.png

dd4d30c3-93b4-4b75-b382-506a6b8dd3bf-image.png

About TypoX AI

Typo move to TON
https://community.tonup.io/t/typox-ai-move-to-ton/885

About TPX IDO on TONUP.io

All you need to know about TONUP.io IDO
https://community.tonup.io/t/typox-ai-move-to-ton/885

Date & Time

Whitelist Round

Date: May 29th Time: 00:00 UTC - 12:00 UTC Note: Whitelist Required

Public Round

Date: May 29th 12:00 UTC - May 31st 12:00 UTC Note: First Come, First Served

Token Distribution

Date: May 31st Time: 13:00 UTC IDO Details

IDO Target

Amount: 10,000 TON

IDO Max Amount

Whitelist Round: 15,000 TON Public Round: 50,000 TON

IDO Price

1 $TPX: 0.0786 $TON (~0.5 USD)

Vesting Details

Whitelist Round & Public Round: 100% fully unlocked at Token Generation Event (TGE)

Participation Limitation

Whitelist Round: May 29th, from 10 TON up to 300 TON per whitelist Public Round: May 29th, from 10 TON up to 200 TON per address Tokenomics

b1e79292-d3c1-4efe-8745-349b64035383-image.png

FAQ

When can I get the IDO TPX token?
As there is no lock-up period. You will receive the full amount of TPX allocated to you right after the IDO.

How many circulating supply at the moment of initial launch and after?
Since this is the initial launch, no other portion of the token has been unlocked prior to this IDO. The detailed lock-up terms and vesting schedule are fully outlined in the tokenomics section, and they will be executed accordingly.

What is the IDO raising target?
The initial TONUP.io 4 IDO is target to 10,000 TON value equivalent TPX and 65,000 TON hard cap. With target launch price of $0.5 USD (0.0786 $TON).

Is there any VCs backed TypoX AI?
TypoX AI is fully incubated by KNN3 Network, benefiting from the support of many of KNN3 Network’s investors. However, as TypoX AI was funded and developed from the ground up by KNN3 for almost a year, it skipped the seed round fundraising, (meaning there are no independent TypoX AI seed investors), this provides TypoX AI with the flexibility to distribute TPX with the community initial launch without the pressure of prior rounds.

Additionally, there is an ongoing private sales round that recently commenced. This round primarily aims to provide TPX with extra liquidity operation space, facilitate top exchange listings, and elevate brand awareness to propel the retailing AI ecosystem to the next level.

When will $TPX be listed, and is there any liquidity plan?
$TPX will be listed on decentralized exchanges (DEX) within less than a week after the initial distribution event (IDO).

The majority of the funds raised during the IDO will be allocated for liquidity provision and market value management purposes, allowing ample space for market price discovery.

More questions

If you have more question, feel free to reach out us in TPX IDO TG group: https://t.me/TypoXAI_IDO, or simply leave you question under this thread, thanks!

RLHF’s Optimization Assumption of LLM under Web3 Incentive Mechanism Abstract

This article explores the potential and strategies of utilizing Web3 incentive mechanisms to optimize large language models (LLMs). By conducting an in-depth analysis of the concept, principles, and application of Reinforcement Learning from Human Feedback (RLHF) in LLMs, this study aims to address how incentive mechanisms can improve the quality of human feedback, thereby enhancing the performance and accuracy of language models. The article begins by outlining the basics of RLHF, including its relationship with traditional reinforcement learning, advantages, and existing limitations. Subsequently, the paper proposes a series of ideas for employing Web3 incentive mechanisms, aimed at encouraging users to provide higher quality and more honest feedback through economic rewards. These proposals include methods for determining the sincerity of suggested replies and rating feed-back, reward calculation mechanisms, and strategies for enhancing rating judgments using implicit
feedback.

Access Paper:

View PDF

TPX Verified Sources.png

Following the successful IDO of TPX token within the TON ecosystem, we have compiled a list of verified websites.

These platforms offer reliable, current information on $TPX token metrics, prices, and more, ensuring our community has access to trusted resources for all TPX token needs.

✔ Verified Websites List:

GeckoTerminal: TPX/TON DYOR Ninja: TPX/TON STONfi: TPX/TON

Exploring Multi-Agent Framework CrewAI for Efficient GameFi Data Extraction

Introduction
The rapid evolution of AI technologies has led to the development of sophisticated tools designed to automate and optimize web scraping tasks. Among these innovations is CrewAI, a multi-agent framework that offers powerful capabilities for extracting data from websites. This study explores the use of CrewAI’s ScrapeWebsiteTool to gather GameFi information from Playtoearn, demonstrating its efficiency and potential challenges.

Methodology
The experiment involved utilizing CrewAI’s ScrapeWebsiteTool to scrape data from Playtoearn, a prominent platform for GameFi information. By inputting a URL into ScrapeWebsiteTool, CrewAI’s agent autonomously extracts webpage content, showcasing high efficiency and automation capabilities. This process is akin to delegating the task of page parsing and content extraction to AI, including handling automatic pagination seamlessly.

However, the challenge arises when attempting to navigate to specific subpages within the website. For instance, extracting detailed game information from a list page requires a more complex approach than what a single agent can accomplish. The solution involves deploying two agents: LinkAgent and DetailAgent.

a1.png
Figure 1. The Initial Scrape Page.
a2.png
Figure 2. The target pages from which comprehensive information needs to be scraped.

Agent Definition and Implementation

LinkAgent: This agent’s primary task is to retrieve links to detailed game pages from the list page. It scans the main page, identifies the relevant links, and prepares them for further extraction.

DetailAgent: Using the links provided by LinkAgent, DetailAgent navigates to each detailed game page and scrapes specific information. The extracted data is formatted and structured into JSON, containing fields such as “name”, “url”, “genres”, “blockchain”, “platforms”, “status”, “content”, “NFT_support”, “free_to_play”, “play_to_earn”, “socialscore_number”, and “socialscore_change”.

a3.png
Figure 3. The content extracted for each game.

Discussion
The use of CrewAI’s multi-agent framework for web scraping in the GameFi sector demonstrates notable improvements in efficiency and automation. While minor stability issues persist, the ability to log and address errors ensures that the benefits far outweigh the drawbacks. Future work could focus on enhancing the robustness of AI parsing and exploring additional applications of multi-agent frameworks in other domains.

Reference:
https://www.crewai.com/
https://docs.crewai.com/tools/ScrapeWebsiteTool/

Introduction to CrewAI

CrewAI is a multi-agent framework with a more humanized development concept. You act as a project manager, define various employees (agents), assign them different tasks, and finally define a team (crew). This team consists of various agents and subtasks, and the final task of the team is to fulfill the client's requirements.
To quickly understand CrewAI, you only need to grasp its three basic elements: Agent, Task, and Crew.

I. Agent

An agent is the basic operational unit in CrewAI, similar to a virtual role with specific duties and capabilities. In CrewAI, an agent is defined by the following four basic elements, which are described in natural language in the code.

1. Role
The role is the identity and duty description of the agent. It defines the agent's position and function within the system, determining the agent's behavior and interaction with other agents.

2. Goal
The goal is the specific task or result that the agent needs to achieve. Each agent has a clear goal that guides its behavior and decision-making.

3. Backstory
The backstory provides context and background for the agent, helping to define its behavior and decision-making style. It may include the agent's hypothetical experiences, knowledge areas, and working style.

4. Tools
Tools are the specific resources or methods used by the agent to perform tasks. These can include external APIs, data processing libraries, search engines, etc. Agents can work without tools, but tools are not mandatory.

1.1 Example

Suppose we create a Research Specialist Agent.

1. Role
Research Specialist

2. Goal
Collect and organize detailed information about conference participants, including their background, company, and recent activities.

3. Backstory
An experienced researcher with extensive data analysis and information gathering experience, capable of efficiently mining and organizing relevant information to provide strong support for conference preparation.

4.Tools
Web scraping tool (ScrapeWebsiteTool), PDF reading tool (PDFSearchTool)

II. Task

A task is the specific work that an agent needs to complete. CrewAI defines a task through the following four elements:

1. Description
A detailed description of the task, clearly stating its purpose and content.

2. Expected Output
The specific results or products required upon task completion. This sets a clear standard for task completion, ensuring the agent knows what kind of result to deliver.

3. Tools
The specific resources or methods used by the agent in task execution. These tools are the same as those defined for the agent and can be temporarily relied upon for the task.

4. Agent
The specific role that will execute the task.

2.1 Example

Suppose we create a "Collect Conference Participants Information" task.

1. Description
Collect detailed background information on all conference participants, including their educational background, career experience, current job position, and responsibilities.

2. Expected Output
An Excel sheet containing detailed information for each participant, with columns for name, educational background, career experience, current position, and current company.

3. Tools
ExcelTool, a tool for creating Excel sheets.

4. Agent
Research Specialist Agent

III. Crew

A crew is a team composed of multiple agents working together to complete more complex tasks. Defining a crew mainly involves defining a team of agents, a list of tasks, and a process, which can be understood as the project execution process.

1. Agents
Define which members (agents) are in the crew.

2. Tasks
Define the specific tasks the crew needs to complete.

3. Process
The key to defining a crew is defining the process. Generally, this is a sequential structure where the next task starts after the previous one is completed, and the final task represents the crew's ultimate output. For advanced operations such as defining parallel and sequential tasks, CrewAI has a dedicated Process class, which is beyond the scope of this introductory chapter.

3.1 Example

Suppose we define a crew to send product quotations.

1. Agents

list itemProductInfoCollectorAgent: Responsible for collecting product information. CustomerInfoCollectorAgent: Responsible for collecting customer information. PricingCalculatorAgent: Responsible for calculating product prices. QuoteDocumentGeneratorAgent: Responsible for generating the quotation document. EmailSenderAgent: Responsible for sending the quotation via email.

2. Tasks

CollectProductInformationTask: Collect detailed product information. CollectCustomerInformationTask: Collect detailed customer information. CalculatePricingTask: Calculate prices based on product and customer information. GenerateQuoteDocumentTask: Generate the product quotation document. SendQuoteEmailTask: Send the quotation to the customer via email.

3. Process

Sequential structure.

Today, I attempted to use CrewAI's ScrapeWebsiteTool to establish a RAG system. It can scrape website content based on a given URL and directly use the content to respond without storing it.
Using the GameFi consulting section as an example, here are the steps to achieve this with two agents:

GameLinkSearchAgent: Uses ScrapeWebsiteTool to provide links to game listing pages and find links to detailed game pages. GameDetailAgent: Uses ScrapeWebsiteTool to search for answers within the specified webpage based on user queries and organize the response.

Comparison with a regular RAG system:

Regular RAG system: Offline work: Scrape content and store it in Vector DB. Online service: User query > Vector search > Document + AI analysis > Response. Online crawler RAG system: Offline work: None. Online service: User query > GameLinkSearchAgent > Link + GameDetailAgent > Response.

It is evident that the online crawler RAG system is simpler and directly eliminates the offline work of building a VectorDB. However, it has a significant drawback: the response time is too slow.
The following two images show the time comparison between the online crawler RAG system I tried and the regular RAG system I previously built. The average response time for the regular system is about 4 seconds, while the online crawler RAG system takes over 30 seconds.

This is easy to understand why, as live scraping takes time. In contrast, Vector Search in a Vector DB is a very fast search method. It only requires parallel computation of the similarity between documents and the query, followed by quick sorting. Compared to finding content on a webpage, this method significantly saves time. Additionally, the documents in the vector database are already parsed content, eliminating the time needed for online parsing.
Although completely eliminating the offline crawling work is still unrealistic in actual development, it does not mean that CrewAI is useless in this context. CrewAI can be used in two scenarios:

Offline crawling + storing in vector database: Use CrewAI for offline crawling, parse the scraped content, and store it in a vector database. This approach improves the efficiency of crawling work development and does not affect the speed of resource retrieval for subsequent online services. Combining online crawler with RAG system: While a vector database can store a large number of documents, its capacity is limited and cannot cover all information, especially the latest news. Combining the online crawler setup with the RAG system can handle the latest content that has not yet been stored in the vector database. This method leverages the real-time nature of the online crawler to compensate for the limitations of the vector database.

In the Crew system, a Crew can be understood as a team project, with Agents and Tasks being the members and tasks within this team project. The order or structure in which these tasks are carried out requires the use of Crew AI's Process mechanism. Currently, there are two developed Processes.

Sequential Processing
This is very simple, as tasks are executed in sequence. In the code, you only need to define a Tasks list when defining the Crew, and it will default to sequential processing. You can also explicitly set the process to be sequential.

Hierarchical Processing
This is a bit more complex but also flexible enough. Typically, you need to define a manager_llm, which can be understood as a manager that automatically decides the execution order and subset relationships of each task. Of course, you can customize this manager through a prompt.
In practice, if the tasks are too complex and numerous, this manager_llm will not always be 100% accurate. If you want to manually define the task execution order while supporting parallel and serial structures, you can define the callback method when defining tasks to trigger the next task.

Typo Curator Mini App Beta Launch & $TPX Giveaway: 3000 $TON Buyback Event

Welcome to Typo Curator:

🎮 What's Typo Curator?
Dive into our cutting-edge mini app, designed to turn AI labeling into a game! Typo Curator invites you to solve puzzles and help build essential datasets, all while earning $TPX for your contributions.

🚀 Powering Growth with TON:
By fully utilizing the TON ecosystem's native capabilities for acquiring, spending, and staking $TPX, Typo Curator is set to expand quickly, enhancing app adoption and promoting $TPX usage.

Solving.png

Event Details

⏰ Typo Curator Launch Timing:

The big day is June 13 at 11:00 AM (UTC+8) ! Join our countdown and be part of the excitement as this isn't just an app launch; it’s your gateway to contributing to the TON ecosystem while earning rewards.

💡 Participate & Rewards:

Test to Earn: Participate in the Typo Curator beta testing to earn $TPX rewards!
Reward pool is limited, first-come, first-serve. Fast Reward Distribution: Beta testers rejoice! You’ll receive your $TPX within 24 hours of participation.

TypoCurator.png

💸 WHY IS THIS LAUNCH CAMPAIGN CALLED THE 3000 TON BUYBACK?

3000 TON Buy Back: Why this name? It’s simple—we're using 3000 TON to buy back $TPX from the current DEX pool for rewards. This approach ensures a focused distribution based on the current market value of $TPX, without increasing the circulating supply.

Transparency
Considering the transparency and authenticity of the entire process, we will disclose a dedicated treasury address after the start of the 3000 TON buyback, allowing the community to monitor the entire buyback process on-chain and ensuring that all repurchased $TPX is used for Beta testing user incentives.

Additionally, to prevent wash trading or data manipulation, a dedicated website will be available when Beta testing incentives go live, allowing real-time tracking of the test user address queue for the community to easily check progress.

🔔 Get Ready to Jump In: Full participation details will be released as we launch. Stay tuned to our official channels to get all the information you need to get started.

➡ Join Beta Testing Telegram: HERE
➡ Check buyback progress: HERE
🈯 查看中文版活動詳情

🎉 Congratulations to the winners of TypoX AI Zealy Sprint 2! 🏆

A huge thank you to everyone who participated. We have distributed the rewards on BNB Chain based on the rankings as follows, please check your wallet on BNB Chain Network:

🥇 1st Place: $50
🥈 2nd Place: $30
🥉 3rd Place: $10
🎖️ 4th-25th Place: $5 each

If you have any questions, feel free to DM us in our Telegram group: https://t.me/TypoGraphyAI.

Please note that future rewards will be integrated with the TON ecosystem and TPX Token market activities.

Stay tuned for more updates! 🚀

20b6bb8f-96c5-41aa-bd2e-9b29b06d7920-image.png

Custom tools are a powerful and versatile way to extend the functionality of CrewAI agents. They are easy to create, requiring only the inheritance of CrewAI's baseTool class. Once inherited, the tool's name and description can be defined as follows:

from crewai_tools import BaseTool class MyCustomTool(BaseTool): name: str = "Name of my tool" description: str = "What this tool does. It's vital for effective utilization." def _run(self, argument: str) -> str: # Your tool's logic here return "Tool's result"

The _run method can accept multiple arguments, which CrewAI will automatically parse from the user's query. This behavior is similar to OpenAI's tools and can be further customized using an args schema for more precise argument parsing.
In addition to inheriting from BaseTool, a simpler approach involves using the @tool decorator. For instance:

from crewai_tools import tool @tool("Tool Name") def my_simple_tool(question: str) -> str: """Tool description for clarity.""" # Tool logic here return "Tool output"

The method's docstring serves as the tool's description, making this approach more concise.
Leveraging custom tools empowers CrewAI to connect with various databases, third-party APIs, and more.

Background
TypoX AI is dedicated to enabling a broader range of users to better connect with Web3 application scenarios and innovative businesses. As a blockchain deeply integrated with Telegram and retail user scenarios, TON will bring nearly one billion users into Web3.

TypoX AI aims to become the most important AI protocol within the TON ecosystem, accelerating Web3 mass adoption. Through TypoX x TON, more people will be empowered to contribute to the transformation of AI technology.

Here is the revised version with improved grammar and clarity, in both Chinese and English:

English Version
Operating Mechanism
As an exploratory attempt, TPX Treasury will allocate a budget of 3,000 $TPX (for first month) to recruit 5-8 community members as the first batch of community ambassadors.

During their tenure, ambassadors will assist the community management team with various operational tasks, such as distributing important community information, answering community questions, compiling content for the community forum, and establishing and maintaining community rules. Task assignments will be arranged based on members’ preferences and actual circumstances through consultation.

The detailed initial action guide will be announced later and maintained and iterated through future community governance meetings and activities.

How to Participate
Fill out the form and submit your intention. The community administrator will contact you shortly.
🔗 Submit here: https://forms.gle/Gq6j9atuUVGYyMCf8

Chinese Version
運行機制
作為探索性嘗試，TPX Treasury將準備3000 TPX的預算招募5-8名社區成員，成為第一批社區大使。

運作期間，大使將協助社區管理組處理各項運營事務，如社群重要信息分發、社區問題答疑、社區論壇內容編撰、社區規則搭建和維護等。任務分工將根據成員喜好和實際情況協商安排。

具體的初版行動指南將在稍後公佈，並通過未來的社區治理會議和活動進行維護和迭代。

如何參與
填寫表單，提交你的意向，社區管理員稍後將會與你取得聯繫。
🔗 提交鏈接：https://forms.gle/Gq6j9atuUVGYyMCf8

Typo Curator迷你應用Beta版上線及$TPX贈送.png

歡迎探索 Typo Curator： 🎮Typo Curator 是什麼？
深入我們的前沿迷你應用程式，將 AI 標籤化變成遊戲！Typo Curator 邀請您解謎並幫助建立必要的數據集，同時為您的貢獻賺取 $TPX。 🚀通過 TON 助力成長：
通過充分利用 TON 生態系統的原生能力來獲得、支出和抵押 $TPX，Typo Curator 即將迅速擴展，增強應用採納並推廣 $TPX 的使用。
Solving.png 活動詳情

⏰ Typo Curator 啟動上線時間：

就在6 月 13 日上午 11:00 (UTC+8) ！ 加入我們的倒計時，成為這激動人心的一部分，這不僅僅是一個應用程序的發布；這是您進入 TON 生態系統並賺取獎勵的入口。

💡參與 & 獎勵：

測試以賺取：參與 Typo Curator 的 beta 測試以賺取 $TPX 獎勵！ 獎勵池有限，先到先得。 快速獎勵發放：Beta 測試者歡呼！您將在參與後 24 小時內收到您的 $TPX。

TypoCurator.png

💸 為什麼這次發起活動被稱為 3000 TON 回購？

為什麼這個名字？這很簡單——我們使用 3000 TON 從當前的 DEX 池回購 $TPX 作為獎勵。這種方法確保了基於當前市場價值的 $TPX 的集中分配，而不會增加流通供應。

回購透明公開
鑑於對整個流程的透明性和真實性的考量，我們將在3000 TON回購開始後，公布專用的treasury地址，以便社區在鏈上監控整個回購過程，並確保所有回購的$TPX用於Beta測試用戶的激勵。

此外，為防止刷單或數據操縱，Beta測試激勵上線時，將提供一個專門的網站，實時追踪測試用戶的地址隊列，方便社區查詢進度。

準備加入活動： 完整的參與細節將隨著我們的發布而公布。 請關注我們的官方渠道以獲得開始所需的所有信息。

➡ 加入 Beta 測試 Telegram：點擊加入
➡ 檢查回購進度：點擊查詢
🆎 View the event details in English

Запуск бета-версии мини-приложения Typo Curator.png

Добро пожаловать в Typo Curator: 🎮 Что такое Typo Curator?
Погрузитесь в наше передовое мини-приложение, созданное для превращения маркировки ИИ в игру! Typo Curator приглашает вас решать головоломки и помогать создавать необходимые наборы данных, зарабатывая при этом $TPX за ваши вклады. 🚀 Рост мощностей с помощью TON:
Используя все возможности экосистемы TON для приобретения, расходования и стейкинга $TPX, Typo Curator быстро развивается, повышая привлекательность приложения и способствуя использованию $TPX.

TypoCurator Features.png

Детали мероприятия

⏰ Время запуска Typo Curator:

Большой день – 13 июня, 06:00 по московскому времени! Присоединяйтесь к нашему обратному отсчету и станьте частью волнения, ведь это не просто запуск приложения; это ваш шанс внести свой вклад в экосистему TON и заработать награды.

💡Участвуйте и получайте награды:

Тестируй и зарабатывай: принимайте участие в бета-тестировании Typo Curator и зарабатывайте награды $TPX! Фонд наград ограничен, награды распределяются по принципу кто первый пришел, тот первый обслужен. Быстрая выдача наград: участники бета-тестирования, радуйтесь! Вы получите свои $TPX в течение 24 часов после участия.

TypoCurator.png

💸 ПОЧЕМУ ЭТУ КАМПАНИЮ НАЗЫВАЮТ ВЫКУПОМ 3000 $TON?

3000 $TON выкуп: почему такое название? Всё просто - мы используем 3000 $TON для выкупа $TPX из текущего пула DEX в качестве наград. Этот подход обеспечивает сосредоточенное распределение на основе текущей рыночной стоимости $TPX, не увеличивая обращающееся предложение.

Прозрачность
Учитывая прозрачность и подлинность всего процесса, мы раскроем адрес выделенного казначейства после начала выкупа 3000 TON, что позволит сообществу контролировать весь процесс выкупа в блокчейне и гарантировать, что все выкупленные $TPX будут использоваться для стимулирования пользователей бета-тестирования.

Кроме того, для предотвращения манипуляций с данными или фиктивной торговли будет доступен специальный веб-сайт, когда начнутся стимулы бета-тестирования, что позволит в реальном времени отслеживать очередь адресов пользователей теста, чтобы сообщество могло легко проверять прогресс.

Готовьтесь к участию: Полные детали участия будут опубликованы во время запуска. Следите за нашими официальными каналами, чтобы получить всю необходимую информацию для начала.

➡ Присоединяйтесь к бета-тестированию в Telegram: ЗДЕСЬ
➡ Проверьте прогресс выкупа: ЗДЕСЬ
🆎 View the event details in English

Our company has recently replaced GPT-3.5 with Gemini-Flash entirely. Gemini is Google's LLM, and the Gemini-Pro version has been available for over a year. While Gemini-Pro is comparable to OpenAI's GPT-4, its performance is noticeably inferior to GPT-4, not to mention OpenAI's latest GPT-4o.

However, the recently launched Gemini-Flash-1.5 which targets GPT-3.5, surpasses GPT-3.5 significantly. It not only has a higher context token limit of 2.8 million (compared to GPT-3.5's maximum of 16K) but is also much cheaper. The table below compares the data of GPT-3.5-turbo-0125 and Gemini-Flash-1.5.

a3ce5146-d449-4503-baaf-eb84dca4a300-1717636088365.jpg

However, the Gemini ecosystem is still not as developed as GPT. For example, CrewAI currently does not have a ready-made method to connect with Gemini. According to the CrewAI Assistant created by the official CrewAI team, you can connect to Gemini using a custom tool. However, a custom tool cannot fully replace the need for GPT because the tool is only used when entering an Agent. Before entering an Agent, the default LLM is still required to handle certain tasks.
The good news is Langchain now supports connecting to Gemini. It’s only need to install langchain-google-genai:

pip install langchain-google-genai

Therefore, I will next attempt to see if CrewAI can fully replace the default LLM with Gemini by leveraging Langchain.

Crew AI is a powerful and easy-to-use multi-agent framework designed to simplify the AI development process and provide users with flexible tools and resources. In this series of articles, I will share insights and techniques I have gained while learning Crew AI, and demonstrate how to apply this knowledge to practical projects. This is not only a detailed Crew AI tutorial but also showcases its practical application through the development of a GameFi consultation system.

There are three key elements in the Crew AI framework: Agent, Task, and Crew. An Agent is an entity that performs specific tasks, a Task is a set of instructions defining the actions of an Agent, and a Crew is the system that manages and coordinates the Agents. The entire development philosophy is highly personified, akin to you playing the role of a project manager, defining various employees (Agents), assigning them different tasks (Tasks), and ultimately creating a team (Crew) to fulfill the project's requirements for your clients.

This chapter will briefly introduce these three elements and create a simple GameFi recommendation system.

Agent
An Agent is the basic operational unit in Crew AI, similar to a virtual character with specific duties and capabilities. In Crew AI, an Agent is defined by the following four basic elements, which are essentially a single natural language sentence in the code. Role: The role describes the identity and responsibilities of the Agent, defining its position and function within the system. The role determines how the Agent behaves and interacts with other Agents. Goal: The goal is the specific task or result the Agent needs to achieve. Each Agent has a clear goal that guides its behavior and decisions. Backstory: The backstory provides context and circumstances for the Agent, helping to define its behavior and decision-making style. The backstory can include the Agent's hypothetical experiences, knowledge domain, and working style. Tools: Tools are the specific resources or methods the Agent uses to perform tasks. Tools can be external APIs, data processing libraries, search engines, etc. Note that an Agent can still function without tools; they are not mandatory. Task
A Task is the specific work an Agent needs to complete, defined in Crew AI by the following four elements. Description: The description is a detailed explanation of the task, clearly stating its purpose and content. Expected Output: The expected output defines the specific result or product required upon task completion. It sets clear standards for task completion, ensuring the Agent knows what kind of result to deliver. Tools: Tools are the specific resources or methods the Agent uses while performing the task. They are the same as the tools defined for an Agent, meaning an employee can temporarily rely on a tool to complete a specific task. Agent: The Agent is the specific role that executes the task. Crew
A Crew is a team of multiple Agents working together to complete more complex tasks. Defining a Crew primarily involves defining a team of Agents and a list of Tasks, as well as a Process, which can be understood as the project execution process. Agents: Define the members of the Crew, i.e., which Agents are included. Tasks: Define the specific Tasks that need to be completed within the Crew. Process: The key to defining a Crew is defining the Process. In simple terms, it is a sequential structure where each task is processed one after the other, with the final task being the Crew's ultimate output. For advanced operations such as defining parallel and sequential tasks, Crew AI has a specialized Process class that will be elaborated on in subsequent chapters. Practical Application
In this practical application, we will temporarily not involve Tools, and the Process will only need to be set as a sequential structure. The aim of this chapter is to understand the basic usage of Agent, Task, and Crew. So first, we import these three classes. from crewai import Agent, Task, Crew

Next, we define two Agents: one for searching GameFi information and the other for recommending GameFi.

gamefi_search_agent = Agent( role="GameFi Search Assistant", goal="Find the GameFi game's detailed according to the user's query,which is {query}", backstory="You are a GameFi Search Assistant, your task is to find the detailed page" " of the GameFi game according to the user's query. GameFi is the blockchain game" ) gamefi_recommend_agent = Agent( role="GameFi Recommend Assistant", goal="Recommend the GameFi game according to the user's query,which is {query}", backstory="You are a GameFi Recommend Assistant, your task is to recommend the GameFi game" " according to the user's query. GameFi is the blockchain game" )

Define their tasks: one for searching GameFi and the other for recommendation.

task_search = Task( description="find GameFi games according to {query}", expected_output="games' details", agent=gamefi_search_agent ) task_recommend = Task( description="recommend GameFi games according to {query}", expected_output="games' recommendation", agent=gamefi_recommend_agent )

Finally, define a Crew to coordinate them.

crew = Crew( agents=[gamefi_search_agent, gamefi_recommend_agent], tasks=[task_search, task_recommend], )

After defining the Crew, call the kickoff method and pass in an inputs dictionary, where the keys correspond to the contents defined in {} for Agents and Tasks.

inputs = {"query": "I need a fps gameFi game"} result = crew.kickoff(inputs=inputs)

Finally, to run the program, you need to define the OpenAI API key in the environment variable.

export OPENAI_API_KEY="Your API Key"

Yes, Crew AI uses OpenAI's GPT model as the default LLM. Once the OpenAI API key is defined in the environment variable, the program will automatically obtain this key without additional operations in the code. Subsequent chapters will introduce how to use LLMs other than GPT.

Next, you can run the program and observe the results.

Effectiveness

The overall effect of these two is almost on par, and GLM4 also supports multimodal input and output like GPT4o, but multimodal capability of GLM4 are inferior to GPT4o. GLM4 performs slightly better in a Chinese environment, with GLM4’s performance when asked questions in Chinese being slightly superior to GPT4o’s. When asked questions in English, GLM4’s performance is not as good as GPT4’s, but the difference is not significant. If we define the effectiveness score of GPT4o when asked questions in English as 5, then:

f52741c5-2245-4307-9555-1d243b63ee33-image.png

Political correctness: GLM4 relies on China’s political correctness, while GPT4o relies on Western political correctness.

Price and Token Count:

933891f3-9202-4bc9-93ff-dc9d59c55e7f-image.png

The price is calculated based on the exchange rate of 1 US Dollar to 7.24 Chinese Yuan.
Overall, GLM is slightly more expensive than GPT. Below is the various GLM models in detail.
GLM-4-0520 is the best GLM4 model.
GLM-4-Air has better performance than GPT3.5-turbo but is slightly slower.
GLM-4-Airx has the same effect as GLM-4-Air but is 2.6 times faster, about the same speed as GPT3.5-turbo.
GLM-4-Flash is only suitable for simple tasks.

Ecosystem:
The ecosystem of GLM4 is far inferior to that of GPT. As of (2024-6), langchain only supports up to GLM3, not to mention frameworks like CrewAI for Multi-Agent Systems.

Open Source:
Neither is open-sourced, but GLM3 is open-sourced.

Currently, discussions about training LLMs primarily revolve around fine-tuning or LoRA (Low-Rank Adaptation). Pre-training LLMs is not within the scope of this discussion (as pre-trained LLMs usually do not require manual annotation). The current annotation methods can be classified into three main categories: scoring, ranking, and text annotation.

Scoring:
This involves rating the responses generated by the LLM. The most basic form is to rate the overall satisfaction with the response. However, beyond overall satisfaction, scoring can be detailed into several directions, such as:
Accuracy: Correctness of the answer.
Completeness: Whether the answer includes all necessary information.
Relevance: Whether the answer is directly related to the question.
Language Quality: Clarity and fluency of the language used.
Creativity: Whether the answer offers unique insights or information.

Ranking:
This method involves having the LLM generate several responses and then manually ranking these responses. The ranking criteria can be overall satisfaction or reference the indicators mentioned in the scoring section. Combining scoring and ranking with Reinforcement Learning from Human Feedback (RLHF) annotation systems can be very effective.

Text Annotation:
This involves providing a query with one or several manually written answers or annotating several manually written replies within a context. This annotation method is the most basic approach for fine-tuning LLMs, but it is also labor-intensive and complex, usually employed for injecting knowledge into the LLM by human experts.