Large Language Models (LLMs) - Notes¶
Overview¶
A large language model (LLM) is a language model consisting of a neural network with many parameters (typically billions of weights or more), trained on large quantities of unlabeled text using self-supervised learning or semi-supervised learning + NLP techniques.
Applications¶
- Chatbots (ChatGPT, Bard ...)
- Search Engines (Being AI, Google ...)
- Multi-language Translator
- Image & videos generation (AI art...)
Tools & Frameworks¶
- LangChain
- Cohere
- GPT3.5 / GPT-4 (OpenAI)
- LLaMA (Meta)
- ...
Hello World!¶
LLMs Hello World example using LangChain bash process to perform simple filesystem commands.
from langchain_experimental.llm_bash.base import LLMBashChain
from langchain.llms import OpenAI
llm = OpenAI(temperature=0)
text = "Please write a bash script that prints 'Hello World' to the console."
bash_chain = LLMBashChain.from_llm(llm, verbose=True)
bash_chain.run(text)
Prompt sequences:
> Entering new LLMBashChain chain...
Please write a bash script that prints 'Hello World' to the console.
```bash
echo "Hello World"
```
Code: ['echo "Hello World"']
Answer: Hello World
> Finished chain.
'Hello World\n'
For more details, check the entire notebook here.
Vision LLMS vs Vision Language Models¶
- OpenAI DALLE 2
- Flamingo (DeepMind)
- GPT-3 (OpenAI)
- LaMDA (Google)
- PaLM (Google)
-
Llama2 (Meta AI)
-
Src: Most Powerful 7 Language (LLM) and Vision Language Models (VLM) Transforming AI in 2023
References¶
Lectures & Tutorials: