DOCSEARCHbyAlgolia
Submit Docs

T4mako Knowledge Base

website·Indexed 29 days ago

https://t4mako.github.io

A technical documentation site covering a wide range of programming and AI topics, including Machine Learning (Reinforcement Learning), AI Agents (OpenAI Agents SDK, multi-agent systems), Java (JDK, Maven, JWT), Vue.js (Vue 2, Vuex, Vue Router, VuePress), Docker, and LeetCode algorithm solutions.

Tokens
361.2K
Snippets
1.5K
Records
2K
Agent score
75%

What's inside T4mako

  2. Overview of Supervised Fine-Tuning (SFT)

    Supervised Fine-Tuning (SFT) is the first step in post-training a Large Language Model (LLM) to achieve its final goals. It uses high-quality, human-written or synthetic 'instruction-response' pairs to teach the model how to understand and follow human instruction formats. The primary goal of SFT is to learn style and instruction following rather than injecting new world knowledge.

  3. Overview of MyBatis Features

    MyBatis is a semi-automatic Object Relational Mapping (ORM) framework that allows developers to customize SQL, stored procedures, and advanced mappings. It eliminates most JDBC boilerplate code, such as manual parameter setting and result set retrieval, by using XML or annotations to map Java POJOs to database records.

  4. Overview of JSON Web Token (JWT) Standard

    JSON Web Token (JWT) is an open standard (RFC 7519) that defines a compact and self-contained way for securely transmitting information between parties as a JSON object. Because the information is digitally signed, it can be verified and trusted. JWTs can be signed using the HMAC algorithm or using a public/private key pair with RSA or ECDSA.

  5. Overview of native Ajax and XMLHttpRequest

    Ajax (Asynchronous JavaScript and XML) is a technique that allows web pages to be updated asynchronously by exchanging data with a server in the background. This enables the retrieval of data and updating of partial page content without requiring a full page refresh. The core object used to implement this in native JavaScript is the XMLHttpRequest (XHR) object.

  6. Overview of Pandas Data Structures

    Pandas is an open-source Python library for structured data analysis and processing. It introduces two primary data structures:

    • Series: A one-dimensional array-like structure consisting of data and associated labels (index). It can be viewed as a single column of a DataFrame.
    • DataFrame: A two-dimensional tabular structure with both row and column indices. It can be viewed as a collection of Series objects sharing the same index.

  7. Overview of the Skills on-demand loading system

    The Skills system implements 'progressive loading' to prevent context window bloat and attention dilution. Rather than including all specialized instructions in the system prompt, the agent is provided with a lightweight list of available skills. When a specific task requires specialized knowledge (e.g., PDF manipulation or database operations), the agent calls the load_skill tool to inject the full instruction set into the current conversation history.

  8. Overview of common torch.nn basic layers

    Commonly used basic layers in torch.nn include:

    • nn.Linear: Fully connected layer.
    • nn.Embedding: Maps discrete features (like words) to dense vectors.
    • nn.Flatten: Flattens a multi-dimensional tensor into a 1D tensor.
    • nn.Dropout / nn.Dropout2d / nn.Dropout3d: Randomly drops elements for regularization.
    • nn.Threshold: Clips input values outside a specified range.
    • nn.ConstantPad2d / nn.ReplicationPad1d / nn.ZeroPad2d: Various padding strategies to extend tensor dimensions.

  9. Overview of the Hook System

    The Hook system allows the AI Agent's main loop to trigger external shell commands at specific execution milestones. This decouples core logic from additive behaviors like auditing, security checks, and notifications. The main loop exposes 'timing' events, and the HookManager executes registered commands, aggregating their results to decide whether to proceed, modify parameters, or inject messages into the conversation.

  10. Overview of Vuex State Management

    Vuex is a plugin for Vue.js designed to implement a centralized state management pattern. It provides a consistent way to manage shared state across multiple components, serving as a communication mechanism applicable to any components regardless of their relationship. It is primarily used when multiple components depend on the same state or when behaviors in different components need to change the same state.

  11. Overview of Elasticsearch and the ELK Stack

    Elasticsearch is an open-source distributed search engine based on Apache Lucene. It is designed for fast retrieval of content from massive datasets and is commonly used for search, log statistics, analysis, and system monitoring. It serves as the core of the Elastic Stack (ELK), which consists of Elasticsearch (storage, search, and analysis), Logstash, Kibana, and Beats.

  12. Overview of Spring Framework Core Concepts

    Spring is a lightweight, open-source JavaEE framework designed to reduce the complexity of enterprise application development. Its two core components are:

    1. IOC (Inversion of Control): Transfers the responsibility of object creation and lifecycle management to the Spring container to achieve loose coupling.
    2. AOP (Aspect-Oriented Programming): Allows adding functionality to existing code without modifying the original source code.