I have a question: entities_and_triples were never used after it got created. All the embedding is based on the original text which is a simple RAG. I thought using triplex is to achieve GraphRAG.
@@amanpatkar7009 Nothing wrong first of all also I guess we can simply store entities and relationship in txt file and use it further that shall be fine
🔥SciPhi Triplex - Best Alternate of GraphRAG - Install Locally ruclips.net/video/jxQkyRY3Qco/видео.htmlsi=qKGtquS7LCxdyebm
Very very useful content! Thank you for providing this! Just the right amount of detail!
Glad you enjoyed it!
You are doing an amazing work breaking down the most significant technological ai advances to users🎉
Can you show us and example how to use the triplex in conjunction with the LLM model? I am curious to know how.
Already did, plz search the channel.
I have a question: entities_and_triples were never used after it got created. All the embedding is based on the original text which is a simple RAG. I thought using triplex is to achieve GraphRAG.
thats what I thought lol
me2, i dont see the triplet info used....
Yes Correct
He just wants to get views. that's all.
@@amanpatkar7009 Nothing wrong first of all also I guess we can simply store entities and relationship in txt file and use it further that shall be fine
Hi can we use Llamma3? Or should we have to use Triplex for the entity matching?
triplex is fine-tuned for this purpose.
so the big q now here is: how do you create the entity types, entities and predicates from a text? what system (LLM) do you use for that?
plz elaborate the question
Thank you, but it is very slow and always makes errors in any book. Is there a browser setting for the first time to deal with a book pdf
curious... why using jupyter notebook instead of just python?
haha. Because, everyone was saying in comments to use notebook instead of python. :)
@@fahdmirza lol ok... in all seriousness though, have you noticed any benefit over "the traditional way"?