Update data with fresh R2 pull (5,743 rows) and add retention table to dataset card

README.md

@@ -37,13 +37,13 @@ LLM inference benchmark dataset measuring serving performance (TTFT, TPOT, ITL,
 
 This dataset provides two benchmark configurations reflecting distinct data collection methodologies:
 
-### trace_replay
+### trace_replay (3,147 rows)
 
 Requests replay exact ISL/OSL sequences from recorded agent sessions (SWE-Bench, TerminalBench, OSWorld, ShareGPT). Input distributions are empirically grounded in real tool-use patterns, capturing realistic burstiness and turn-depth correlations.
 
 Profiles: `chat-medium`, `chat-multiturn-long`, `chat-multiturn-medium`, `chat-multiturn-short`, `chat-short`, `chat-singleturn`, `coding-singleturn`, `decode-heavy`, `osworld-multiturn-long`, `osworld-multiturn-medium`, `osworld-multiturn-short`, `prefill-heavy`, `random-1k`, `swebench-multiturn-medium`, `swebench-multiturn-short`, `terminalbench-multiturn-medium`, `terminalbench-multiturn-short`
 
-### distributional
+### distributional (245 rows)
 
 Requests sample ISL/OSL from parameterized statistical distributions (e.g., lognormal) fitted to real workload statistics. Shorter to run than full trace replays, enabling faster characterization of profile-level serving behavior across the hardware matrix. MSE validation confirms distributional runs reproduce the latency and throughput characteristics of their trace-replay counterparts.
 
@@ -55,7 +55,14 @@ Profiles: `chat-multiturn`, `chat-singleturn`, `coding-singleturn`, `osworld-mul
 
 ### Concurrency filtering
 
-Rows where declared concurrency exceeds the session pool size have been excluded. This affects trace_replay data at concurrency > 100 (session pool was 100) and distributional/current data at concurrency > 10 (session pool was 10). Distributional data collected after the fix has no such limitation.
+The benchmark harness capped actual concurrent connections at the session pool size. Rows where declared concurrency exceeds the pool size never achieved the intended load and have been excluded.
+
+| Scope | Raw rows | After filter | Removed | Filter rule |
+|-------|----------|-------------|---------|-------------|
+| trace_replay (archive) | 4,814 | 3,147 | 1,667 | C > 100 (session pool = 100) |
+| distributional (current) | 843 | 159 | 684 | C > 10 (session pool = 10) |
+| distributional (fixed) | 86 | 86 | 0 | No cap — bug fixed |
+| **Total** | **5,743** | **3,392** | **2,351** | **Retention: 59.1%** |
 
 ## Coverage
 

croissant.json

@@ -75,7 +75,7 @@
       "name": "distributional/summary.parquet",
       "contentUrl": "https://huggingface.co/datasets/lynae-1219/AgentPerfBench/resolve/main/distributional/summary.parquet",
       "encodingFormat": "application/x-parquet",
-      "sha256": "b55a7f80bb7c23d68ce30d9cc8b0ddb5c412479581fdb18755d394303329138d"
+      "sha256": "bb9a48351e0b612b3125c5d0a8e200638d790269f821e45aae342a4042bfd951"
     },
     {
       "@type": "cr:FileObject",

distributional/summary.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b55a7f80bb7c23d68ce30d9cc8b0ddb5c412479581fdb18755d394303329138d
-size 54449
+oid sha256:bb9a48351e0b612b3125c5d0a8e200638d790269f821e45aae342a4042bfd951
+size 70836