@reitzen.bsky.social
For LLMs? My 4090 is collecting dust.
At first it's magical to run 4o mini level models locally, but then you realize any non chat use is bursty and spending 20c calling cloud inference in parallel saves ten minutes of waiting around.
Not sure the equation would shift even with double the VRAM.
At first it's magical to run 4o mini level models locally, but then you realize any non chat use is bursty and spending 20c calling cloud inference in parallel saves ten minutes of waiting around.
Not sure the equation would shift even with double the VRAM.
May 23, 2025 at 3:01 AM
For LLMs? My 4090 is collecting dust.
At first it's magical to run 4o mini level models locally, but then you realize any non chat use is bursty and spending 20c calling cloud inference in parallel saves ten minutes of waiting around.
Not sure the equation would shift even with double the VRAM.
At first it's magical to run 4o mini level models locally, but then you realize any non chat use is bursty and spending 20c calling cloud inference in parallel saves ten minutes of waiting around.
Not sure the equation would shift even with double the VRAM.
What's the game?
April 21, 2025 at 3:12 AM
What's the game?
Hi Trey, it's not my list - I just used an importer tool to grab a bunch. Sorry!
April 13, 2025 at 3:46 AM
Hi Trey, it's not my list - I just used an importer tool to grab a bunch. Sorry!