Thuta Learning
ရှာဖွေရန်
IntermediateAIbeginner

Introduction to Large Language Models

စိတ်လျှော့ပါ။ ဒီခန်းကို စာအုပ်လိုမဟုတ်ဘဲ စကားပြောသလိုပဲ၊ နားလည်လွယ်အောင် ရှင်းပါမယ်။

ဒီခန်းပြီးရင် ဘာတတ်သွားမလဲ

  • Introduction to Large Language Models ကို ကြောက်စရာမလိုအောင် နားလည်မယ်
  • ကိုယ်တိုင် စမ်းသုံးတတ်အောင် လုပ်မယ်
  • မှားလွယ်တဲ့နေရာကို ပြုံးပြီး ရှောင်တတ်မယ်

LLM က စာသားကို ဘယ်လိုလက်ခံ၊ ခန့်မှန်း၊ ပြန်ထုတ်သလဲနဲ့ model ကို database လို့မယူဆသင့်တဲ့အကြောင်း လေ့လာမည်။

ခဏလေး ဒီလိုပဲ စဉ်းစားကြည့်

Large Language Model ဆိုတာ စာအုပ်ကြီးတစ်အုပ်ထဲက အဖြေကို copy ပြန်ရှာပေးတဲ့ database မဟုတ်ပါဘူး။ Text pattern အများကြီးကနေ နောက် token ရဲ့ ဖြစ်နိုင်ခြေကို သင်ယူထားတဲ့ model ဖြစ်ပါတယ်။ ဒီကြောင့် စာရေးပုံကောင်းပေမယ့် fact မှားနိုင်ပါတယ်။

Introduction to Large Language Models lesson illustration
Prompting and LLMs — Introduction to Large Language Models

နေ့စဉ်ဘဝနဲ့ ချိတ်ကြည့်မယ်

LLM ကို reasoning assistant၊ summarizer၊ classifier နဲ့ content drafting tool အဖြစ် သုံးနိုင်ပါတယ်။ လတ်တလော သို့မဟုတ် ကိုယ့်ကုမ္ပဏီအချက်အလက်လို model မသိနိုင်တာကို context၊ retrieval သို့မဟုတ် tool ကနေ ပေးရပါတယ်။

အတူတူ လက်တွေ့စမ်းမယ်

text
User message + Instructions + Context

          Language Model

       Probable next tokens

              Response
You should see
LLM ရဲ့ အားသာချက်နဲ့ ကန့်သတ်ချက်ကို လူတစ်ယောက်ကို ရှင်းပြနိုင်မည်။

၅ မိနစ် စမ်းကြည့်

LLM ကို အသုံးချမယ့် feature တစ်ခုရွေးပြီး model သိနိုင်တာ၊ မသိနိုင်တာနဲ့ ပေးရမည့် context ကို သုံးကော်လံရေးပါ။

သတိလေးတစ်ချက်

AI output ကို အပြီးသတ်အမှန်လို့မယူပါနှင့်။ အရေးကြီးတဲ့အချက်၊ code နဲ့ user data ကို လူကပြန်စစ်ပြီးမှ အသုံးပြုပါ။

OpenAI — Text generationOpenAI

ဒီနေရာမှာ လူအများမှားတတ်တယ်

  • LLM ကို အမြဲမှန်တဲ့ search engine လို့ယူဆခြင်း
  • Model ကြီးလာတာနဲ့ domain data မလိုတော့ဘူးလို့ယူဆခြင်း

လေ့ကျင့်ခန်း

LLM ကို အသုံးချမယ့် feature တစ်ခုရွေးပြီး model သိနိုင်တာ၊ မသိနိုင်တာနဲ့ ပေးရမည့် context ကို သုံးကော်လံရေးပါ။

You'll know it worked when: LLM ရဲ့ အားသာချက်နဲ့ ကန့်သတ်ချက်ကို လူတစ်ယောက်ကို ရှင်းပြနိုင်မည်။

Introduction to Large Language Models | Thuta Learning