Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Question Answering on NarrativeQA (Prefill TP, Cache Hit, F1)
Loading...
24,686.84
Prefill Throughput (tok/s)
ContextPilot w/o Annotations
8,048.0256
12,367.7178
16,687.41
21,007.1022
Nov 5, 2025
Prefill Throughput (tok/s)
Cache Hit Rate
F1 Score
Updated 5mo ago
Evaluation Results
Method
Method
Links
Prefill Throughput (tok/s)
Cache Hit Rate
F1 Score
ContextPilot w/o Annotations
Hardware=32×H20
2025.11
24,686.84
35.71
40.38
ContextPilot
Hardware=32×H20, Annot...
2025.11
24,686.84
35.71
41.08
Vanilla
Hardware=32×H20
2025.11
16,247.32
5.14
40.2
ContextPilot w/o Annotations
Hardware=16×H20
2025.11
13,201.52
38.24
40.38
ContextPilot
Hardware=16×H20, Annot...
2025.11
13,201.52
38.24
41.08
Vanilla
Hardware=16×H20
2025.11
8,687.98
6.08
40.2
Feedback
Search any
task
Search any
task