Understand & Visualise Microsoft GraphRAG
Вставка
- Опубліковано 16 вер 2024
- -
Code: github.com/gra...
#datascience #machinelearning #deeplearning #datanalytics #predictiveanalytics #artificialintelligence #generativeai #largelanguagemodels #computervision #naturallanguageprocessing #agents #transformers #embedding #graphml #graphdatascience #datavisualization #businessintelligence #optimization #montecarlosimulation #simulation #LLMs #python #aws #azure #gcp
If you found this content useful, pleases consider sharing it with others who might benefit. Your support is greatly appreciated :)
loving this graph RAG series. please continue and would love examples with financial statements etc
Thanks for your support :)
great video, really great
Thanks for your support :)
Really such a great video, thanks for sharing. This is the video that i was looking for to visualize the graphs. Please share the notebook too
Thanks for your support :) The link is now added in the description....
Awesome 🎉
no its not
Thanks for your support :)
can u share notebook
sorry, the link now added in the description...
very helpful thanks. its better than LLMGraph but the GraphQAChain doesnt work here in case of GraphRAG, is there any other retriever or resource which can help to build a custom retriever any help would be greatly appreciated.
Here are different retrievers, including how to create custom retriever, pls check - python.langchain.com/v0.1/docs/modules/data_connection/retrievers/
thanks a lot , one doubt , sorry for my dumb questions , what is the use of this one ? what is the objective ?
knowledge graphs have their own use cases, the visualisation is only for some intuitive understanding but we can use cypher to query KGs similar to databases using sql.
Hi, how can we summarize a txt file containing different stories in it in which each story starts as story :name of the story, how can we get thè summary of each story as i tried it ,it is mixing all the stories and not providing me summary of each story as i want,i tried using both local and global search
You have to pre-process your doc, extract the story using a function and then feed only one story at a time to llm.
@@SridharKumarKannam is there any other way like making some changes or provide with some parameters which overrides my default settings and help me out directly
@@034_mayankparashar9 unfortunately not, AFAIK
Whats that create index in the 2nd cell
sorry, I don't see "create index" in 2nd cell. Can you pls give me the timestamp where the code is shown? thanks,
Make video on hipporag
Thanks for the suggestion, sure, I'll try...
why no I just use langchain with knowledge grapghs rather than this expensive graphRAG
yes graphRAG is expensive and we don't have to use it. Microsoft proposed it first in its current farm and then other frame works have implemented it.
If you found this content helpful, please consider sharing it with others who might benefit. Your support is greatly appreciated :)
can you share the code please?
The link is now added in the description....
@@SridharKumarKannam thanks!