8000 GitHub - altfund/pinkhaus · GitHub
[go: up one dir, main page]

Skip to content
altfundPublic

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

Pre-Requisites

AI

Local AI w/ OpenWebUI and Ollama

RAG

on FASTRAG:

  • pc: https://news.ycombinator.com/item?id=42174829 yes I believe they're transformers. and yes to the training idea. so this argument for fast graph rag made me feel really good. this ground is fairly well trod. putting everything in a database and hoping its useful works but caps out. and so they're pro graphs and relationships. more research revealed there are plenty of approaches we can take, to improve a model's effectiveness with the data it does have. so I think we sprint towards the 0.1 you're imagining then start making it more sophisticated.

  • eric:

from fast_graphrag import GraphRAG

DOMAIN = "Evaluate these market odds, match results, and podcast transcripts for soccer and develop a model for what players have an impact on teams performances, which teams beat / underperform their odds, how many goals they score, and what other patterns exist."

EXAMPLE_QUERIES = [
"How many goals are Arsenal likely to score against Nottingham Forest",
"What are the odds that Asenal beat Nottingham Forest?",
"What Arsenal and Nottingham players have an impact on games and how are they likely to react against each other"
]

ENTITY_TYPES = ["Team", "Player", "League", "Manager", "Owner", "Match", "Stadium"]

grag = GraphRAG(
working_dir="./soccer",
domain=DOMAIN,
example_queries="\n".join(EXAMPLE_QUERIES),
entity_types=ENTITY_TYPES
)

with open("./odds_results_and_transcripts.txt") as f:
grag.insert(f.read())

print(grag.query("which team will win [next game]?").response)

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages

0