Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
fredliu
on Sept 14, 2023
|
parent
|
context
|
favorite
| on:
Efficient Memory Management for Large Language Mod...
I might be wrong, but looks like this could help with speculative decoding which can already vastly improves the inference speed?
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
