Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
login
throwaway74432
on April 4, 2024
|
parent
|
context
|
favorite
| on:
Improvements to the fine-tuning API and expanding ...
Number of tokens is still a useful metric, as their endpoints have Tokens Per Minute quotas. Decreasing number of tokens used means increasing throughput, up until the Request Per Minute quota.
afro88
on April 4, 2024
[–]
But the sentence was "Indeed was able to improve cost..."
mechagodzilla
on April 4, 2024
|
parent
|
next
[–]
From OpenAI’s perspective the cost improved!
throwaway74432
on April 4, 2024
|
parent
|
prev
[–]
"...and latency" Higher throughput = lower latency*
*Under certain conditions
Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search:
