Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
Phi-3 14B model with 128k context
(
huggingface.co
)
2 points
by
Handy-Man
on May 21, 2024
|
hide
|
past
|
favorite
|
1 comment
causal
on May 21, 2024
[–]
Really impressive for the size. The 4.2B vision model apparently outperforms GPT-4V-Turbo on 5 out of 7 vision benchmarks.
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
