diff --git a/website/src/components/BenchmarkDefinitions.astro b/website/src/components/BenchmarkDefinitions.astro index 26bbfd0..b6e98e1 100644 --- a/website/src/components/BenchmarkDefinitions.astro +++ b/website/src/components/BenchmarkDefinitions.astro @@ -7,10 +7,10 @@ import { ModelDefinition, TTFTDefinition, TPSDefinition, TotalTimeDefinition } f
Definitions
Methodology
-Distributed Footprint → We run our tools daily in multiple data centers using Fly.io. Currently we run in cdg, iad, and sea.
-Short Requests → Input requests are a short, single message (~20 tokens) and ask for a brief output response.
-Max Output → Max output tokens is set to 100 to avoid distortion of TPS values from long outputs.
-Warmup for Consistent Setup → A warmup connection is made to remove any connection setup latency.
-TTFT Roundtrip → The TTFT clock starts when the HTTP request is made and stops when the first token result is received in the response stream.
-Try 3, Keep 1 → For each provider, three separate inferences are done, and the best result is kept (to remove any outliers due to queuing etc).
+Distributed Footprint: We run our tools daily in multiple data centers using Fly.io. Currently we run in cdg, iad, and sea.
+Short Requests: Input requests are a short, single message (~20 tokens) and ask for a brief output response.
+Max Output: Max output tokens is set to 100 to avoid distortion of TPS values from long outputs.
+Warmup for Consistent Setup: A warmup connection is made to remove any connection setup latency.
+TTFT Roundtrip: The TTFT clock starts when the HTTP request is made and stops when the first token result is received in the response stream.
+Try 3, Keep 1: For each provider, three separate inferences are done, and the best result is kept (to remove any outliers due to queuing etc).
Source
-Raw Data → All data is in this public GCS bucket.
-Benchmarking Tools → The full test suite is available in the ai-benchmarks repo.
-Website → Full source code for this site is on GitHub.
+Raw Data: All data is in this public GCS bucket.
+Benchmarking Tools: The full test suite is available in the ai-benchmarks repo.
+Website: Full source code for this site is on GitHub.