Comment on How to run DeepSeek R1 locallyparentComments−DrPhish1y“you can get a dual EPYC server with 768GB RAM - CPU inference only at around 6-8 tokens/sec.”This is what I run at home. I built it just over a year ago and have run every single model that has been released.
Comments
“you can get a dual EPYC server with 768GB RAM - CPU inference only at around 6-8 tokens/sec.”
This is what I run at home. I built it just over a year ago and have run every single model that has been released.