<?xml version="1.0" encoding="UTF-8" ?>
<rss
    version="2.0"
    xmlns:atom="http://www.w3.org/2005/Atom"
    xmlns:content="http://purl.org/rss/1.0/modules/content/"
    xmlns:webfeeds="http://webfeeds.org/rss/1.0"
    xmlns:media="http://search.yahoo.com/mrss/"
    >
    <channel>
        <title>Nlp-natural-language-processing Tag - Viblo</title>
        <link>https://viblo.asia/rss</link>
        <description><![CDATA[Free service for technical knowledge sharing]]></description>
        <atom:link href="https://viblo.asia/rss/tags/nlp-natural-language-processing.rss" rel="self"></atom:link>
                <copyright>Sun* Inc.</copyright>
                                                <webfeeds:logo>https://viblo.asia/logo_full.svg</webfeeds:logo>
        <image>
            <url>https://viblo.asia/logo_full.svg</url>
            <title>Nlp-natural-language-processing Tag - Viblo</title>
            <link>https://viblo.asia/rss</link>
        </image>
                                <language>vi-vn</language>
        <lastBuildDate>2026-07-21T21:11:25+07:00</lastBuildDate>
                <item>
            <title><![CDATA[Token AI là gì? Đơn vị dữ liệu và cửa sổ ngữ cảnh của AI]]></title>
                        <link>https://viblo.asia/p/token-ai-la-gi-don-vi-du-lieu-va-cua-so-ngu-canh-cua-ai-K9Vy8wKdLQR</link>
            <guid isPermaLink="true">https://viblo.asia/p/token-ai-la-gi-don-vi-du-lieu-va-cua-so-ngu-canh-cua-ai-K9Vy8wKdLQR</guid>
            <description><![CDATA[Mỗi khi bạn gõ một câu hỏi cho ChatGPT hay Claude, mô hình không đọc chữ cái hay từ ngữ theo cách con người vẫn đọc. Nó đọc token.

**Token AI là đơn ...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Trần Văn Tuấn</dc:creator>
            <pubDate>2026-06-30 16:15:23</pubDate>
                                                                                                        </item>
                <item>
            <title><![CDATA[Paraphrase các câu Tiếng Việt trên Python với VietQuill]]></title>
                        <link>https://viblo.asia/p/paraphrase-cac-cau-tieng-viet-tren-python-voi-vietquill-kNLr3E5WVgA</link>
            <guid isPermaLink="true">https://viblo.asia/p/paraphrase-cac-cau-tieng-viet-tren-python-voi-vietquill-kNLr3E5WVgA</guid>
            <description><![CDATA[Paraphrase, hiểu đơn giản, là viết lại một câu sao cho nghĩa vẫn vậy nhưng cách diễn đạt thì khác đi. Ví dụ câu "Hôm nay trời đẹp quá, mình muốn đi dạ...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Sang Nguyen</dc:creator>
            <pubDate>2026-06-21 13:43:43</pubDate>
                                                                                                        </item>
                <item>
            <title><![CDATA[Cách tôi thay thế LLM API đắt đỏ bằng Rule-Based NLP Engine 100% Offline (Độ trễ 0ms)]]></title>
                        <link>https://viblo.asia/p/cach-toi-thay-the-llm-api-dat-do-bang-rule-based-nlp-engine-100-offline-do-tre-0ms-PoL7ejjl4vk</link>
            <guid isPermaLink="true">https://viblo.asia/p/cach-toi-thay-the-llm-api-dat-do-bang-rule-based-nlp-engine-100-offline-do-tre-0ms-PoL7ejjl4vk</guid>
            <description><![CDATA[Vì sao tự build một Domain-Specific Rule-Based Engine hoàn toàn bằng JavaScript lại là quyết định kiến trúc chuẩn xác nhất cho dự án Open Source của m...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">paudang</dc:creator>
            <pubDate>2026-06-04 08:00:00</pubDate>
                                                                                                        </item>
                <item>
            <title><![CDATA[Deep Dive – Phân Tích Chi Tiết LSTM: Giữ Thông Tin Lâu Dài, So Sánh với RNN và Công Thức Toán Học ]]></title>
                        <link>https://viblo.asia/p/deep-dive-phan-tich-chi-tiet-lstm-giu-thong-tin-lau-dai-so-sanh-voi-rnn-va-cong-thuc-toan-hoc-BQyJKvwW4Me</link>
            <guid isPermaLink="true">https://viblo.asia/p/deep-dive-phan-tich-chi-tiet-lstm-giu-thong-tin-lau-dai-so-sanh-voi-rnn-va-cong-thuc-toan-hoc-BQyJKvwW4Me</guid>
            <description><![CDATA[ Chào mừng bạn quay trở lại! Nếu bạn đã theo dõi bài Tìm Hiểu LSTM: Bí Quyết Giữ Thông Tin Lâu Dài Hiệu Quả  trước đó, bạn sẽ biết LSTM vượt trội RNN ...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Long Nguyễn Thành</dc:creator>
            <pubDate>2025-02-11 13:21:59</pubDate>
                                                                                                        </item>
                <item>
            <title><![CDATA[[Papers Notes] NaViT: Thuật toán Transformer với Tầm nhìn Đa phân giải - Cải thiện khả năng nhìn của Transformer trong xử lý hình ảnh]]></title>
                        <link>https://viblo.asia/p/papers-notes-navit-thuat-toan-transformer-voi-tam-nhin-da-phan-giai-cai-thien-kha-nang-nhin-cua-transformer-trong-xu-ly-hinh-anh-GAWVpZ0XJ05</link>
            <guid isPermaLink="true">https://viblo.asia/p/papers-notes-navit-thuat-toan-transformer-voi-tam-nhin-da-phan-giai-cai-thien-kha-nang-nhin-cua-transformer-trong-xu-ly-hinh-anh-GAWVpZ0XJ05</guid>
            <description><![CDATA[Transformer, ban đầu được thiết kế cho xử lý ngôn ngữ tự nhiên (NLP), đã nhanh chóng trở thành một kiến trúc quan trọng trong thị giác máy tính. Các m...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Dương Xuân Bách</dc:creator>
            <pubDate>2025-01-13 00:53:44</pubDate>
                                                                                                        </item>
                <item>
            <title><![CDATA[LLM Agent: Hướng dẫn cơ bản về LLM Agent]]></title>
                        <link>https://viblo.asia/p/llm-agent-huong-dan-co-ban-ve-llm-agent-aAY4qDnyVPw</link>
            <guid isPermaLink="true">https://viblo.asia/p/llm-agent-huong-dan-co-ban-ve-llm-agent-aAY4qDnyVPw</guid>
            <description><![CDATA[TỔNG QUAN
LLM Agent liên quan đến các ứng dụng LLM có thể thực thi các tác vụ phức tạp thông qua việc sử dụng kiến trúc kết hợp với LLM và các module ...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Nguyễn Chí Minh Hiếu</dc:creator>
            <pubDate>2024-12-16 21:02:43</pubDate>
                                                                                                        </item>
                <item>
            <title><![CDATA[Self-Attention và Multi-head Sefl-Attention trong Transformers]]></title>
                        <link>https://viblo.asia/p/self-attention-va-multi-head-sefl-attention-trong-transformers-n1j4lO2aVwl</link>
            <guid isPermaLink="true">https://viblo.asia/p/self-attention-va-multi-head-sefl-attention-trong-transformers-n1j4lO2aVwl</guid>
            <description><![CDATA[Transformers  là gì?  Liệu nó có gì đặc biệt mà trong machine learning người ta nhắc đến nó nhiều như thế?  Trước thời điểm Google giới thiệu bài báo ...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Khanh Tran Van</dc:creator>
            <pubDate>2024-11-05 00:39:11</pubDate>
                                                                                                        </item>
                <item>
            <title><![CDATA[Sự Khác Biệt Giữa LLM và ChatGPT: Khái Niệm và Ứng Dụng]]></title>
                        <link>https://viblo.asia/p/su-khac-biet-giua-llm-va-chatgpt-khai-niem-va-ung-dung-Yym40Yj9J91</link>
            <guid isPermaLink="true">https://viblo.asia/p/su-khac-biet-giua-llm-va-chatgpt-khai-niem-va-ung-dung-Yym40Yj9J91</guid>
            <description><![CDATA[
LLM là gì?

Trong thời gian gần đây, bạn có thể đã nghe nhiều về các mô hình ngôn ngữ lớn (Large Language Models - LLM). LLM là một loại mô hình trí ...]]></description>
                        <dc:creator xmlns:dc="http://purl.org/dc/elements/1.1/">Trương Văn Qui</dc:creator>
            <pubDate>2024-07-12 20:37:56</pubDate>
                                                                                                        </item>
            </channel>
</rss>
