quick summary of @eugeneyan
's epic post on "Patterns for Building LLM-based Systems"
LLMS
-
Summary of LLM-based Systems Building Patterns
By
–
-
LLM Patterns: Essential Design Patterns for Large Language Models
By
–
https://
eugeneyan.com/writing/llm-pa
tterns/
… -
Running Deep Learning Models on M1 Mac mini
By
–
I wonder if it'll still work if you start running deep learning models or an LLM on this M1 Mac mini.
-
Less Data Better Results Finetuning LLMs with Trimmed Datasets
By
–
Next to LIMA, this is another interesting paper highlighting that more data is not always better when finetuning LLMs: https://
arxiv.org/abs/2307.08701 Trimming the orig 52k Alpaca dataset to 9k can improve the performance when finetuning 7B and 13B parameter LLMs. -
Novel LLM Project Combines Language Models with External Data
By
–
This was a really cool project! Very impressive and novel way to combine LLMs with external data
-
Brief CNN testing experience, focus shifted to LLM work
By
–
Phew, it's been a while. I think I only briefly tried it on some basic CNNs, and it worked fine out of the box. But afaik it's mainly for computer vision applications (have been super focused on LLM work recently)
-
Execute Python Code Directly from LLM Output
By
–
Python REPL You can pipe the output of an LLM call into a Python REPL to run that code
-
OpenAI Function Calling: Structuring Model Output with Bind
By
–
Function Calling @OpenAI Functions are extremely popular for structuring output – you can easily attach them to a model call using `.bind`
-
Transform Chat Model Outputs with Prompt Templates and Parsers
By
–
Prompt Template + LLM + Output Parser Some types of models (chat models) return chat messages It's often desirable to transform that into a string so you can easily use it in downstream tasks – for that we can add in a simple output parser
-
Prompt Templates and LLM Integration Basics
By
–
Prompt Template + LLM Starting with the most basic, a simple combination of a prompt template + a call to an LLM `prompt | llm`