Great "info-dump" idea. GTFIH

[sellatine]

I have had this idea for quite some time and my experience in CS has made brain storming this easier. I was thinking of an idea where either the founders of this site (highly unlikely), or I could get the info from this site . I have already made a simple scraper which scrapes all the info from the "Looksmaxing" section and compiles it. Using ML + Rules-Based Filtering, we could make a theoretical "full" guide encompassing the best strategies burrowed down on this site for ages. Here's my plan once I get all the data (as long as I can do so without getting banned. Mods pls tell me):

• Use a chunking strategy to break down the dataset into smaller pieces
• Apply modeling (like BERTopic or LDA) to get sort of main ideas and summaries
• Use GPT, LLaMa, or Claude (which has the highest token limit) to summarize content within each topic
• Create a searchable knowledge base from the results

If any mods see this and dont condone this just tell me and ill stop like a good boy.

 

[goyimslayer]

I have had this idea for quite some time and my experience in CS has made brain storming this easier. I was thinking of an idea where either the founders of this site (highly unlikely), or I could get the info from this site . I have already made a simple scraper which scrapes all the info from the "Looksmaxing" section and compiles it. Using ML + Rules-Based Filtering, we could make a theoretical "full" guide encompassing the best strategies burrowed down on this site for ages. Here's my plan once I get all the data (as long as I can do so without getting banned. Mods pls tell me):

• Use a chunking strategy to break down the dataset into smaller pieces
• Apply modeling (like BERTopic or LDA) to get sort of main ideas and summaries
• Use GPT, LLaMa, or Claude (which has the highest token limit) to summarize content within each topic
• Create a searchable knowledge base from the results

If any mods see this and dont condone this just tell me and ill stop like a good boy.

try it bro

 

[sellatine]

try it bro

i will i need sum help with the LLM shit tho