1
2
3
4
5
Drag to adjust the number of frozen columns
Replicate - Performance Metrics
GPT- Neo (1.3B)
GPT- Neo (125M)
Roberta Large
With Cold Start
249 secs
126.63 secs
182.23 secs
Inference time
4.66 secs
3.8 secs
4.6 secs
Latency
253.66 secs
130.43 secs
186.83 secs
Variability
Stable
Stable
Stable
Autoscaling (c=5)
Holds Latency
Holds Latency
Holds Latency
5 records

Alert

Lorem ipsum
Okay
View larger version