Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
aidenn0
on June 13, 2023
|
parent
|
context
|
favorite
| on:
Llama.cpp: Full CUDA GPU Acceleration
AFAIK it's not fully, but you can use cuBLAS/clBlast for a pretty good speedup.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: