Upload folder using huggingface_hub

README.md

@@ -24,6 +24,10 @@ configs:
     data_files:
       - split: summary
         path: synthetic_distributional/summary.parquet
+  - config_name: per_layer_kernel
+    data_files:
+      - split: summary
+        path: per_layer_kernel/summary.parquet
   - config_name: kernels_labeled
     data_files:
       - split: train
@@ -32,10 +36,6 @@ configs:
     data_files:
       - split: summary
         path: mse_validation/summary.parquet
-  - config_name: layer_roofline
-    data_files:
-      - split: summary
-        path: layer_roofline/summary.parquet
 dataset_info:
   - config_name: trace_replay
     features:
@@ -59,10 +59,6 @@ dataset_info:
         dtype: int64
       - name: duration_s
         dtype: float64
-      - name: successful_requests
-        dtype: int64
-      - name: failed_requests
-        dtype: int64
       - name: request_throughput
         dtype: float64
       - name: input_token_throughput
@@ -105,8 +101,8 @@ dataset_info:
         dtype: float64
     splits:
       - name: summary
-        num_examples: 3147
-        num_bytes: 694254
+        num_examples: 2932
+        num_bytes: 640182
   - config_name: synthetic_distributional
     features:
       - name: run_id
@@ -129,10 +125,6 @@ dataset_info:
         dtype: int64
       - name: duration_s
         dtype: float64
-      - name: successful_requests
-        dtype: int64
-      - name: failed_requests
-        dtype: int64
       - name: request_throughput
         dtype: float64
       - name: input_token_throughput
@@ -175,8 +167,62 @@ dataset_info:
         dtype: float64
     splits:
       - name: summary
-        num_examples: 245
+        num_examples: 223
         num_bytes: 70836
+  - config_name: per_layer_kernel
+    features:
+      - name: record_type
+        dtype: string
+      - name: model
+        dtype: string
+      - name: hardware
+        dtype: string
+      - name: phase
+        dtype: string
+      - name: batch_size
+        dtype: int64
+      - name: sequence_length
+        dtype: int64
+      - name: component_name
+        dtype: string
+      - name: bound
+        dtype: string
+      - name: flops
+        dtype: float64
+      - name: bytes_accessed
+        dtype: float64
+      - name: operational_intensity
+        dtype: float64
+      - name: ridge_point
+        dtype: float64
+      - name: kernel_id
+        dtype: int64
+      - name: kernel_name
+        dtype: string
+      - name: block_size
+        dtype: string
+      - name: grid_size
+        dtype: string
+      - name: duration_us
+        dtype: float64
+      - name: compute_sm_throughput_pct
+        dtype: float64
+      - name: dram_throughput_pct
+        dtype: float64
+      - name: memory_throughput_pct
+        dtype: float64
+      - name: l1_tex_cache_throughput_pct
+        dtype: float64
+      - name: l2_cache_throughput_pct
+        dtype: float64
+      - name: sm_frequency_ghz
+        dtype: float64
+      - name: dram_frequency_ghz
+        dtype: float64
+    splits:
+      - name: summary
+        num_examples: 37
+        num_bytes: 12000
   - config_name: kernels_labeled
     features:
       - name: source
@@ -228,25 +274,7 @@ dataset_info:
         num_examples: 148077
   - config_name: mse_validation
     features:
-      - name: validation_id
-        dtype: string
-      - name: tier
-        dtype: string
-      - name: stage
-        dtype: string
-      - name: stage_label
-        dtype: string
-      - name: comparison_group
-        dtype: string
-      - name: raw_json_r2_uri
-        dtype: string
-      - name: per_turn_json_r2_uri
-        dtype: string
-      - name: replay_kind
-        dtype: string
-      - name: workload
-        dtype: string
-      - name: profile
+      - name: run_id
         dtype: string
       - name: model
         dtype: string
@@ -254,187 +282,52 @@ dataset_info:
         dtype: string
       - name: engine
         dtype: string
-      - name: tensor_parallelism
-        dtype: int64
-      - name: concurrency
-        dtype: int64
-      - name: sessions
-        dtype: int64
-      - name: source_locked
-        dtype: bool
-      - name: prefix_caching_state
-        dtype: string
-      - name: chunked_prefill
-        dtype: string
-      - name: prefix_aware_synthetic
-        dtype: bool
-      - name: shared_prefix_tokens
-        dtype: int64
-      - name: shared_prefix_block_aligned
-        dtype: bool
-      - name: synthetic_filler_style
+      - name: profile
         dtype: string
-      - name: synthetic_target_chars_per_token
-        dtype: float64
-      - name: max_model_len
+      - name: concurrency
         dtype: int64
-      - name: success_rate
-        dtype: float64
       - name: num_requests
         dtype: int64
-      - name: duration_s
-        dtype: float64
       - name: successful_requests
         dtype: int64
       - name: failed_requests
         dtype: int64
-      - name: request_throughput
-        dtype: float64
-      - name: input_token_throughput
-        dtype: float64
-      - name: output_token_throughput
+      - name: duration_s
         dtype: float64
-      - name: total_token_throughput
+      - name: request_throughput
         dtype: float64
       - name: mean_ttft_ms
         dtype: float64
-      - name: median_ttft_ms
-        dtype: float64
-      - name: p90_ttft_ms
-        dtype: float64
-      - name: p99_ttft_ms
-        dtype: float64
       - name: mean_tpot_ms
         dtype: float64
-      - name: median_tpot_ms
-        dtype: float64
-      - name: p90_tpot_ms
-        dtype: float64
-      - name: p99_tpot_ms
-        dtype: float64
-      - name: mean_itl_ms
-        dtype: float64
-      - name: median_itl_ms
-        dtype: float64
-      - name: p90_itl_ms
-        dtype: float64
-      - name: p99_itl_ms
-        dtype: float64
       - name: mean_e2el_ms
         dtype: float64
-      - name: median_e2el_ms
-        dtype: float64
-      - name: p90_e2el_ms
-        dtype: float64
-      - name: p99_e2el_ms
-        dtype: float64
     splits:
       - name: summary
         num_examples: 28
-        num_bytes: 40940
-  - config_name: layer_roofline
-    features:
-      - name: record_type
-        dtype: string
-      - name: source_file
-        dtype: string
-      - name: model
-        dtype: string
-      - name: hardware
-        dtype: string
-      - name: engine
-        dtype: string
-      - name: dtype
-        dtype: string
-      - name: tensor_parallelism
-        dtype: int64
-      - name: phase
-        dtype: string
-      - name: batch_size
-        dtype: int64
-      - name: sequence_length
-        dtype: int64
-      - name: q_len
-        dtype: int64
-      - name: kv_len
-        dtype: int64
-      - name: component_name
-        dtype: string
-      - name: component_bound
-        dtype: string
-      - name: flops
-        dtype: float64
-      - name: bytes
-        dtype: float64
-      - name: operational_intensity_flop_per_byte
-        dtype: float64
-      - name: ridge_point_flop_per_byte
-        dtype: float64
-      - name: total_flops
-        dtype: float64
-      - name: total_bytes
-        dtype: float64
-      - name: overall_bound
-        dtype: string
-      - name: kernel_id
-        dtype: int64
-      - name: kernel_name
-        dtype: string
-      - name: block_size
-        dtype: string
-      - name: grid_size
-        dtype: string
-      - name: duration_us
-        dtype: float64
-      - name: compute_sm_throughput_pct
-        dtype: float64
-      - name: dram_throughput_pct
-        dtype: float64
-      - name: memory_throughput_pct
-        dtype: float64
-      - name: l1_tex_cache_throughput_pct
-        dtype: float64
-      - name: l2_cache_throughput_pct
-        dtype: float64
-      - name: sm_frequency_ghz
-        dtype: float64
-      - name: dram_frequency_ghz
-        dtype: float64
-      - name: artifact_path
-        dtype: string
-      - name: artifact_kind
-        dtype: string
-      - name: artifact_bytes
-        dtype: int64
-      - name: artifact_description
-        dtype: string
-      - name: notes
-        dtype: string
-    splits:
-      - name: summary
-        num_examples: 56
-        num_bytes: 27430
 ---
 
 # AgentPerfBench
 
-LLM inference benchmark: 3,392 main sweep rows measuring TTFT, TPOT, ITL, and throughput across 9 models, up to 14 GPU configurations, and 2 serving engines (vLLM 0.19.0, SGLang 0.5.9). All models served in BF16 except gpt-oss, which uses mxfp4 for projection weights. The dataset also includes 148,077 per-kernel NCU profiles, 28 curated tabular MSE validation rows for the distributional synthetic replay generator, and 56 layer-roofline validation rows.
+LLM inference benchmark: 3,155 main sweep rows and 37 per-layer kernel validation rows, plus 148,077 per-kernel NCU profiles, across 9 models, 14 GPU configurations, and 2 serving engines (vLLM 0.19.0, SGLang 0.5.9). All models served in BF16 except gpt-oss, which uses mxfp4 for projection weights.
 
 ## Dataset configurations
 
-The dataset provides five configurations. `trace_replay` replays exact input/output sequences from recorded agent sessions. `synthetic_distributional` samples from statistical distributions fitted to those same workloads, trading fidelity for faster sweeps across the hardware matrix. `kernels_labeled` contains per-kernel Nsight Compute labels. `mse_validation` contains tabular paired synthetic-vs-real validation rows and ablations for the final APC-aware synthetic generator, with raw JSON artifacts referenced in R2 rather than stored in the dataset repo. `layer_roofline` exposes the per-layer roofline validation artifacts as a loadable tabular subset.
-
-### trace_replay (3,147 rows)
+### trace_replay (2,932 rows)
 
-Replays exact ISL/OSL sequences from recorded agent sessions (SWE-Bench, TerminalBench, OSWorld, ShareGPT). Covers 77 unique (model, hardware, engine) combinations across 17 profiles and 6 concurrency levels. The full 5-dimensional matrix is 12.2% filled; not all models run on all hardware.
+Replays exact ISL/OSL sequences from recorded agent sessions (SWE-Bench, TerminalBench, OSWorld, ShareGPT). 77 unique (model, hardware, engine) combinations, 17 profiles, 6 concurrency levels {1, 5, 10, 20, 40, 80}, 11.4% matrix fill.
 
 17 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`
 
-### synthetic_distributional (245 rows)
+### synthetic_distributional (223 rows)
 
-Samples ISL/OSL from lognormal distributions fitted to real workload statistics. Covers 42 unique (model, hardware, engine) combinations across 6 profiles and 7 concurrency levels (3.0% matrix fill). `gpt-oss-120b`, `3090x8`, and `A100-40GBx8` are excluded from this configuration.
+ISL/OSL sampled from lognormal fits to real workload statistics. 46 unique (model, hardware, engine) combinations, 14 profiles, 3 concurrency levels {1, 5, 10}, 2.8% matrix fill.
 
-6 profiles: `chat-multiturn`, `chat-singleturn`, `coding-singleturn`, `osworld-multiturn`, `swebench-multiturn`, `terminalbench-multiturn`
+14 profiles: `chat-medium`, `chat-multiturn`, `chat-multiturn-long`, `chat-multiturn-medium`, `chat-multiturn-short`, `chat-short`, `chat-singleturn`, `coding-singleturn`, `osworld-multiturn`, `osworld-multiturn-long`, `osworld-multiturn-medium`, `swebench-multiturn`, `terminalbench-multiturn`, `terminalbench-multiturn-short`
+
+### per_layer_kernel (37 rows)
+
+Per-component operational intensity decomposition and Nsight Compute kernel profiles for Llama-3.1-8B on H100 (prefill phase). Analytical rows provide computed FLOPs, bytes, and OI per model component at batch sizes 1 and 80. NCU rows report measured SM and memory throughput per kernel from an 8-layer forward pass. Record types: `analytical_total`, `analytical_component`, `ncu_kernel`.
 
 ### kernels_labeled (148,077 rows)
 
@@ -442,49 +335,27 @@ Per-kernel Nsight Compute (ncu) profiles across 4 GPUs (A100, H100, RTX 3090, RT
 
 ### mse_validation (28 rows)
 
-Curated H100 / Llama-3.1-8B / vLLM validation table for the distributional synthetic replay generator. The main rows keep paired MSE distributional replay and real trace replay runs with success rate at least 75%; supplementary rows preserve no-replacement and high-concurrency debug runs. Raw aggregate/per-turn JSON artifacts are referenced through R2 URI columns, not stored as dataset files.
-
-The headline SWE-bench C=5 source-locked cascade is:
-
-| Condition | Turn 10-19 E2EL delta |
-|-----------|----------------------:|
-| APC on, English filler, source-locked | +45.5% |
-| APC on, code-morph filler, no shared prefix | +31.1% |
-| APC off, code-morph filler, no shared prefix | +11.7% |
-| APC on, code-morph filler, 1024-token shared prefix | -3.2% |
-
-See `mse_validation/README.md`, `mse_validation/manifest.csv`, and `mse_validation/fidelity_deltas.csv` for the filtering rule and R2 raw-artifact URIs.
-
-### layer_roofline (56 rows)
-
-Tabular view over the per-layer roofline evidence for Llama-3.1-8B on H100. This subset combines analytical component OI rows, selected NCU kernel summary rows, and an artifact manifest pointing to the raw `per_layer_oi_cf/` evidence files.
-
-Record types: `analytical_total`, `analytical_component`, `ncu_kernel`, `artifact`.
+Curated H100 / Llama-3.1-8B / vLLM validation table for the distributional synthetic replay generator. Paired synthetic and real trace replay runs; supplementary rows preserve no-replacement and high-concurrency debug runs. Raw JSON artifacts referenced through R2 URI columns. Per-run successful/failed request counts retained.
 
 ### Concurrency filtering
 
-The benchmark harness capped actual concurrent connections at the session pool size. Rows where declared concurrency exceeded the pool were excluded:
+trace_replay: concurrency > 100 removed. synthetic_distributional: concurrency > 10 removed.
 
-- trace_replay: concurrency > 100 removed (session pool was 100). Remaining values: {1, 5, 10, 20, 40, 80}.
-- synthetic_distributional (pre-fix): concurrency > 10 removed (session pool was 10). Post-fix data has no cap. Remaining values: {1, 5, 10, 40, 80, 200, 320}.
+Configurations where fewer than 75% of requests completed successfully are excluded. Summary metrics are computed from successful requests only.
 
 | Config | Rows |
 |--------|------|
-| trace_replay | 3,147 |
-| synthetic_distributional | 245 |
-| mse_validation | 28 |
-| layer_roofline | 56 |
+| trace_replay | 2,932 |
+| synthetic_distributional | 223 |
+| per_layer_kernel | 37 |
 | kernels_labeled | 148,077 |
-
-### Failed requests
-
-Some runs produce request failures, typically at high concurrency where the engine hits memory or timeout limits. 30.8% of trace_replay rows and 42% of synthetic_distributional rows have `failed_requests > 0`. Summary metrics (TTFT, TPOT, throughput) are computed from successful requests only.
+| mse_validation | 28 |
 
 ## Coverage
 
 ### Hardware
 
-All benchmarks collected on PyTorch 2.10.0, CUDA 12.8.
+All benchmarks collected on PyTorch 2.10.0.
 
 | GPU | VRAM | HBM bandwidth | Peak half-precision TFLOPS |
 |-----|------|---------------|---------------------------|
@@ -493,7 +364,7 @@ All benchmarks collected on PyTorch 2.10.0, CUDA 12.8.
 | NVIDIA RTX 3090 | 24 GB | 936 GB/s | 71 |
 | NVIDIA RTX 2080 Ti | 11 GB | 616 GB/s | 27 |
 
-Multi-GPU configurations: 1, 2, 4, or 8 GPUs with tensor parallelism. TP degree depends on model size and available GPUs.
+Multi-GPU configurations: 1, 2, 4, or 8 GPUs with tensor parallelism.
 
 ### Models
 
@@ -511,8 +382,6 @@ All models served in BF16 unless noted.
 | gpt-oss-20b | GPT-OSS | 21B (3.6B active) | MoE | mxfp4 projections |
 | gpt-oss-120b | GPT-OSS | 117B (5.1B active) | MoE | mxfp4 projections |
 
-Model names in this table match the `model` column in the parquet files.
-
 ### Engines
 
 - vLLM 0.19.0
@@ -520,7 +389,7 @@ Model names in this table match the `model` column in the parquet files.
 
 ## Schema
 
-Each row in the serving `summary.parquet` configs:
+Each row in `summary.parquet` (trace_replay and synthetic_distributional):
 
 | Column | Type | Description |
 |--------|------|-------------|
@@ -534,8 +403,6 @@ Each row in the serving `summary.parquet` configs:
 | concurrency | int | Concurrent request level |
 | num_requests | int | Total requests in run |
 | duration_s | float | Total run duration |
-| successful_requests | int | Completed requests |
-| failed_requests | int | Failed requests |
 | request_throughput | float | Requests/second |
 | input_token_throughput | float | Input tokens/second |
 | output_token_throughput | float | Output tokens/second |
@@ -551,39 +418,21 @@ Each row in the serving `summary.parquet` configs:
 from datasets import load_dataset
 
 ds = load_dataset("agent-perf-bench/AgentPerfBench", "trace_replay")
-# or "synthetic_distributional", "kernels_labeled", "mse_validation", "layer_roofline"
+# or "synthetic_distributional", "per_layer_kernel", "kernels_labeled", "mse_validation"
 ```
 
 ## Benchmark methodology
 
 - Closed-loop concurrency with semaphore control.
-- Concurrency levels: {1, 5, 10, 20, 40, 80} (trace_replay), {1, 5, 10, 40, 80, 200, 320} (synthetic_distributional).
 - 3-request warmup before each configuration.
-- Metrics: TTFT, TPOT, ITL, E2EL, request throughput, token throughput.
-- Summary statistics: mean, median, p90, p99.
+- Metrics: TTFT, TPOT, ITL, E2EL, request throughput, token throughput (mean, median, p90, p99).
+- Metrics computed over successful requests only.
 - Collection period: March 2026 onwards.
-- PyTorch 2.10.0, CUDA 12.8 on all machines. All models served in BF16 (gpt-oss: mxfp4 projection weights).
-
-## Future releases
-
-- Full per-request and multi-turn granularity data for the main sweep (pending raw JSON availability from collection infrastructure). Curated raw JSONs are referenced from R2 for `mse_validation`.
-- Additional per-kernel roofline profiles beyond the included `kernels_labeled`, `layer_roofline`, and raw `per_layer_oi_cf` artifacts.
-- This is version 1.0. Updates will be tagged with semantic versions.
-
-## Intended uses
-
-- Inference engine comparison under controlled conditions.
-- Capacity planning for LLM deployments.
-- TTFT scaling with context length in multi-turn sessions.
 
 ## Limitations
 
-- Results are specific to tested hardware and software versions (vLLM 0.19.0, SGLang 0.5.9, PyTorch 2.10.0, CUDA 12.8).
-- Synthetic distributional profiles approximate but do not replicate production traffic patterns.
-- Consumer GPU coverage is limited to RTX 3090 and RTX 2080 Ti; no non-NVIDIA accelerators.
+- Distributional profiles are fitted approximations, not direct production replays.
 - Closed-loop concurrency only; no open-loop (Poisson) arrivals.
-- The model-hardware-concurrency matrix is sparse (12.2% fill for trace_replay, 3.0% for synthetic_distributional). Not all model-hardware combinations are represented.
-- No model quality metrics. This is a systems benchmark.
 
 ## Ethical considerations
 
@@ -600,6 +449,12 @@ Benchmark data released under Apache-2.0. Source datasets retain their original
 - [ShareGPT (Aeala/ShareGPT_Vicuna_unfiltered)](https://huggingface.co/datasets/Aeala/ShareGPT_Vicuna_unfiltered)
 - [OSWorld](https://github.com/xlang-ai/OSWorld)
 
+## Future releases
+
+- Additional hardware configurations and model families.
+- Open-loop (Poisson) arrival mode.
+- Additional per-kernel roofline profiles.
+
 ## Citation
 
 ```bibtex

croissant.json

@@ -1,501 +1,952 @@
-{
-  "@context": {
-    "@language": "en",
-    "@vocab": "https://schema.org/",
-    "citeAs": "cr:citeAs",
-    "column": "cr:column",
-    "conformsTo": "dct:conformsTo",
-    "cr": "http://mlcommons.org/croissant/",
-    "data": {
-      "@id": "cr:data",
-      "@type": "@json"
-    },
-    "dataType": {
-      "@id": "cr:dataType",
-      "@type": "@vocab"
-    },
-    "dct": "http://purl.org/dc/terms/",
-    "equivalentProperty": "cr:equivalentProperty",
-    "examples": {
-      "@id": "cr:examples",
-      "@type": "@json"
-    },
-    "extract": "cr:extract",
-    "field": "cr:field",
-    "fileObject": "cr:fileObject",
-    "fileProperty": "cr:fileProperty",
-    "fileSet": "cr:fileSet",
-    "format": "cr:format",
-    "includes": "cr:includes",
-    "isLiveDataset": "cr:isLiveDataset",
-    "jsonPath": "cr:jsonPath",
-    "key": "cr:key",
-    "md5": "cr:md5",
-    "parentField": "cr:parentField",
-    "path": "cr:path",
-    "rai": "http://mlcommons.org/croissant/RAI/",
-    "recordSet": "cr:recordSet",
-    "references": "cr:references",
-    "regex": "cr:regex",
-    "repeated": "cr:repeated",
-    "replace": "cr:replace",
-    "samplingRate": "cr:samplingRate",
-    "prov": "http://www.w3.org/ns/prov#",
-    "sc": "https://schema.org/",
-    "separator": "cr:separator",
-    "source": "cr:source",
-    "subField": "cr:subField",
-    "transform": "cr:transform"
-  },
-  "@type": "sc:Dataset",
-  "name": "AgentPerfBench",
-  "description": "LLM inference benchmark dataset measuring serving performance (TTFT, TPOT, throughput) across 9 models, 4 GPU platforms, 2 serving engines under agentic and chat workloads. Three configurations: trace_replay, synthetic_distributional, and kernels_labeled.",
-  "url": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench",
-  "license": "https://spdx.org/licenses/Apache-2.0.html",
-  "conformsTo": "http://mlcommons.org/croissant/1.1",
-  "datePublished": "2026-05-04",
-  "version": "1.0",
-  "citeAs": "@inproceedings{agentperfbench2026, title={AgentPerfBench: A Benchmarking and Evaluation Suite for Inference Performance of Agentic LLMs}, author={Anonymous}, booktitle={NeurIPS 2026 Evaluations and Datasets Track}, year={2026}}",
-  "creator": {
-    "@type": "sc:Organization",
-    "name": "Anonymous (under review)"
-  },
-  "distribution": [
-    {
-      "@type": "cr:FileObject",
-      "@id": "trace-replay-summary-parquet",
-      "name": "trace_replay/summary.parquet",
-      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/trace_replay/summary.parquet",
-      "encodingFormat": "application/x-parquet",
-      "sha256": "19dfc3ed8e2f09c52449045cded6bce29a00c0b9c4e563a407ed8e9960333da3"
-    },
-    {
-      "@type": "cr:FileObject",
-      "@id": "synthetic-distributional-summary-parquet",
-      "name": "synthetic_distributional/summary.parquet",
-      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/synthetic_distributional/summary.parquet",
-      "encodingFormat": "application/x-parquet",
-      "sha256": "67d63072831143e929e0b6d6464744756f783b53ac30fed0232bc01c601676ec"
-    },
-    {
-      "@type": "cr:FileObject",
-      "@id": "kernels-labeled-parquet",
-      "name": "kernel_profiles/kernels_labeled.parquet",
-      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/kernel_profiles/kernels_labeled.parquet",
-      "encodingFormat": "application/x-parquet",
-      "sha256": "05571ee61fe97beac7f3d0b0e8be07a0feeca00cee9b308cd469a57cd4c9b671"
-    }
-  ],
-  "recordSet": [
-    {
-      "@type": "cr:RecordSet",
-      "@id": "trace-replay-summary",
-      "name": "Trace Replay Summary",
-      "description": "One row per benchmark configuration from trace replay runs (empirical agent session replays).",
-      "field": [
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/run_id",
-          "name": "run_id",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "run_id"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/model",
-          "name": "model",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "model"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/hardware",
-          "name": "hardware",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "hardware"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/engine",
-          "name": "engine",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "engine"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/profile",
-          "name": "profile",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "profile"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/concurrency",
-          "name": "concurrency",
-          "dataType": "sc:Integer",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "concurrency"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/request_throughput",
-          "name": "request_throughput",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "request_throughput"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/median_ttft_ms",
-          "name": "median_ttft_ms",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "median_ttft_ms"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "trace-replay-summary/median_tpot_ms",
-          "name": "median_tpot_ms",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "trace-replay-summary-parquet"
-            },
-            "extract": {
-              "column": "median_tpot_ms"
-            }
-          }
-        }
-      ]
-    },
-    {
-      "@type": "cr:RecordSet",
-      "@id": "synthetic-distributional-summary",
-      "name": "Synthetic Distributional Summary",
-      "description": "One row per benchmark configuration from synthetic distributional runs (statistical sampling from lognormal fits to recorded workload statistics).",
-      "field": [
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/run_id",
-          "name": "run_id",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "run_id"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/model",
-          "name": "model",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "model"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/hardware",
-          "name": "hardware",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "hardware"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/engine",
-          "name": "engine",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "engine"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/profile",
-          "name": "profile",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "profile"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/concurrency",
-          "name": "concurrency",
-          "dataType": "sc:Integer",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "concurrency"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/request_throughput",
-          "name": "request_throughput",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "request_throughput"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/median_ttft_ms",
-          "name": "median_ttft_ms",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "median_ttft_ms"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "synthetic-distributional-summary/median_tpot_ms",
-          "name": "median_tpot_ms",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "synthetic-distributional-summary-parquet"
-            },
-            "extract": {
-              "column": "median_tpot_ms"
-            }
-          }
-        }
-      ]
-    },
-    {
-      "@type": "cr:RecordSet",
-      "@id": "kernels-labeled",
-      "name": "Kernels Labeled",
-      "description": "Per-kernel Nsight Compute (ncu) hardware counter profiles. Individual kernel invocations across 4 GPUs and 13 model/sweep sources.",
-      "field": [
-        {
-          "@type": "cr:Field",
-          "@id": "kernels-labeled/source",
-          "name": "source",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "kernels-labeled-parquet"
-            },
-            "extract": {
-              "column": "source"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "kernels-labeled/gpu",
-          "name": "gpu",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "kernels-labeled-parquet"
-            },
-            "extract": {
-              "column": "gpu"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "kernels-labeled/model",
-          "name": "model",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "kernels-labeled-parquet"
-            },
-            "extract": {
-              "column": "model"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "kernels-labeled/kernel_family",
-          "name": "kernel_family",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "kernels-labeled-parquet"
-            },
-            "extract": {
-              "column": "kernel_family"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "kernels-labeled/kernel_name",
-          "name": "kernel_name",
-          "dataType": "sc:Text",
-          "source": {
-            "fileObject": {
-              "@id": "kernels-labeled-parquet"
-            },
-            "extract": {
-              "column": "kernel_name"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "kernels-labeled/gpu_time_duration_ms",
-          "name": "gpu_time_duration_ms",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "kernels-labeled-parquet"
-            },
-            "extract": {
-              "column": "gpu_time_duration_ms"
-            }
-          }
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "kernels-labeled/dram_bytes_sum",
-          "name": "dram_bytes_sum",
-          "dataType": "sc:Float",
-          "source": {
-            "fileObject": {
-              "@id": "kernels-labeled-parquet"
-            },
-            "extract": {
-              "column": "dram_bytes_sum"
-            }
-          }
-        }
-      ]
-    }
-  ],
-  "rai:dataLimitations": "Results cover NVIDIA H100, A100, RTX 3090, and RTX 2080 Ti GPUs only and may not generalize to other accelerators (AMD, Intel, TPU). Benchmark configurations are pinned to vLLM 0.19.0 and SGLang 0.5.9; results do not represent other engine versions. Concurrency levels (1-320) may not cover extreme-scale deployments. Not recommended as sole basis for hardware purchasing decisions or for comparing model task quality.",
-  "rai:dataBiases": "Model selection over-represents Meta Llama and Alibaba Qwen families. Hardware is exclusively NVIDIA GPUs. Workload profiles are author-designed approximations of production traffic; real deployment patterns may differ.",
-  "rai:personalSensitiveInformation": "No personally identifiable information is present. All API endpoints and credentials are stripped. Workload traces use synthetic random tokens or publicly available coding benchmarks.",
-  "rai:dataUseCases": "Established uses: relative comparison of inference engine throughput, latency benchmarking under controlled conditions, TTFT scaling with context length in multi-turn sessions, per-kernel performance modelling. Not established: absolute latency prediction for production, model quality comparison, cost estimation.",
-  "rai:dataSocialImpact": "Enables reproducible comparison of open-source LLM serving systems, supporting infrastructure research and reducing vendor lock-in.",
-  "rai:hasSyntheticData": true,
-  "prov:wasDerivedFrom": [
-    {
-      "@id": "https://huggingface.co/datasets/princeton-nlp/SWE-bench",
-      "name": "SWE-Bench"
-    },
-    {
-      "@id": "https://github.com/TerminalBench/TerminalBench",
-      "name": "TerminalBench"
-    },
-    {
-      "@id": "https://github.com/xlang-ai/OSWorld",
-      "name": "OSWorld"
-    },
-    {
-      "@id": "https://huggingface.co/datasets/Aeala/ShareGPT_Vicuna_unfiltered",
-      "name": "ShareGPT"
-    }
-  ],
-  "prov:wasGeneratedBy": {
-    "@type": "prov:Activity",
-    "name": "AgentPerfBench benchmark collection",
-    "description": "Deploy model on target GPU with specified engine and tensor parallelism. Send requests per configuration after warmup using closed-loop concurrency control. Record per-request TTFT, TPOT, ITL, E2EL, and token counts. Compute summary statistics. Sanitize credentials and convert to Parquet."
-  },
-  "rai:dataCollection": "Automated benchmark collection. Models deployed on target GPU with specified engine and tensor parallelism. Requests sent per workload profile using closed-loop concurrency control with semaphore. Per-request TTFT, TPOT, ITL, E2EL, and token counts recorded. Kernel profiles collected via Nsight Compute (ncu) with hardware counter metrics.",
-  "rai:dataCollectionType": "Automated measurement via benchmarking scripts. No human subjects or crowdsourcing.",
-  "rai:dataCollectionMissingData": "Some model-hardware-concurrency combinations are absent due to OOM or engine incompatibility. Concurrency levels exceeding the session-pool size were filtered (see Concurrency filtering in README).",
-  "rai:dataPreprocessing": "Credentials and API endpoints stripped. Concurrency-buggy rows filtered. Configurations with completion rates below 75% are excluded. Summary statistics (mean, median, p90, p99) computed from per-request measurements. Kernel names demangled from NCU output.",
-  "rai:dataAnnotationProtocol": "Not applicable. Fully automated benchmark collection with no human annotation.",
-  "rai:dataAnnotationPlatform": "Not applicable. No human annotation.",
-  "rai:dataAnnotationAnalysis": "Not applicable. No human annotation."
-}
+{
+  "@context": {
+    "@language": "en",
+    "@vocab": "https://schema.org/",
+    "citeAs": "cr:citeAs",
+    "column": "cr:column",
+    "conformsTo": "dct:conformsTo",
+    "cr": "http://mlcommons.org/croissant/",
+    "data": {
+      "@id": "cr:data",
+      "@type": "@json"
+    },
+    "dataType": {
+      "@id": "cr:dataType",
+      "@type": "@vocab"
+    },
+    "dct": "http://purl.org/dc/terms/",
+    "equivalentProperty": "cr:equivalentProperty",
+    "examples": {
+      "@id": "cr:examples",
+      "@type": "@json"
+    },
+    "extract": "cr:extract",
+    "field": "cr:field",
+    "fileObject": "cr:fileObject",
+    "fileProperty": "cr:fileProperty",
+    "fileSet": "cr:fileSet",
+    "format": "cr:format",
+    "includes": "cr:includes",
+    "isLiveDataset": "cr:isLiveDataset",
+    "jsonPath": "cr:jsonPath",
+    "key": "cr:key",
+    "md5": "cr:md5",
+    "parentField": "cr:parentField",
+    "path": "cr:path",
+    "rai": "http://mlcommons.org/croissant/RAI/",
+    "recordSet": "cr:recordSet",
+    "references": "cr:references",
+    "regex": "cr:regex",
+    "repeated": "cr:repeated",
+    "replace": "cr:replace",
+    "samplingRate": "cr:samplingRate",
+    "prov": "http://www.w3.org/ns/prov#",
+    "sc": "https://schema.org/",
+    "separator": "cr:separator",
+    "source": "cr:source",
+    "subField": "cr:subField",
+    "transform": "cr:transform"
+  },
+  "@type": "sc:Dataset",
+  "name": "AgentPerfBench",
+  "description": "LLM inference benchmark dataset: 3,155 main sweep rows measuring TTFT, TPOT, ITL, and throughput across 9 models, up to 14 GPU configurations, and 2 serving engines (vLLM 0.19.0, SGLang 0.5.9). Also includes 148,077 per-kernel NCU profiles, 28 MSE validation rows, and 37 per-layer kernel validation rows.",
+  "url": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench",
+  "license": "https://spdx.org/licenses/Apache-2.0.html",
+  "conformsTo": "http://mlcommons.org/croissant/1.1",
+  "datePublished": "2026-05-04",
+  "version": "1.0",
+  "citeAs": "@inproceedings{agentperfbench2026, title={AgentPerfBench: A Benchmarking and Evaluation Suite for Inference Performance of Agentic LLMs}, author={Anonymous}, booktitle={NeurIPS 2026 Evaluations and Datasets Track}, year={2026}}",
+  "creator": {
+    "@type": "sc:Organization",
+    "name": "Anonymous (under review)"
+  },
+  "distribution": [
+    {
+      "@type": "cr:FileObject",
+      "@id": "trace-replay-summary-parquet",
+      "name": "trace_replay/summary.parquet",
+      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/trace_replay/summary.parquet",
+      "encodingFormat": "application/x-parquet",
+      "sha256": "568efbdae473a830ba792317fb09cfba0ea4f667db67b93d28e72637a5c71d1e"
+    },
+    {
+      "@type": "cr:FileObject",
+      "@id": "synthetic-distributional-summary-parquet",
+      "name": "synthetic_distributional/summary.parquet",
+      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/synthetic_distributional/summary.parquet",
+      "encodingFormat": "application/x-parquet",
+      "sha256": "0fd6ff9f802f9465a4618dfe8d5d426013028565e08ef5024f4302b9126a05a5"
+    },
+    {
+      "@type": "cr:FileObject",
+      "@id": "kernels-labeled-parquet",
+      "name": "kernel_profiles/kernels_labeled.parquet",
+      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/kernel_profiles/kernels_labeled.parquet",
+      "encodingFormat": "application/x-parquet",
+      "sha256": "05571ee61fe97beac7f3d0b0e8be07a0feeca00cee9b308cd469a57cd4c9b671"
+    },
+    {
+      "@type": "cr:FileObject",
+      "@id": "mse-validation-summary-parquet",
+      "name": "mse_validation/summary.parquet",
+      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/mse_validation/summary.parquet",
+      "encodingFormat": "application/x-parquet"
+    },
+    {
+      "@type": "cr:FileObject",
+      "@id": "per-layer-kernel-summary-parquet",
+      "name": "per_layer_kernel/summary.parquet",
+      "contentUrl": "https://huggingface.co/datasets/agent-perf-bench/AgentPerfBench/resolve/main/per_layer_kernel/summary.parquet",
+      "encodingFormat": "application/x-parquet",
+      "sha256": "c78e4694009da435587bc9e7e3e4821fca6a3bf881284f873b71ccdb5ed47691"
+    }
+  ],
+  "recordSet": [
+    {
+      "@type": "cr:RecordSet",
+      "@id": "trace-replay-summary",
+      "name": "Trace Replay Summary",
+      "description": "One row per benchmark configuration from trace replay runs (empirical agent session replays).",
+      "field": [
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/run_id",
+          "name": "run_id",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "run_id"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/model",
+          "name": "model",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "model"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/hardware",
+          "name": "hardware",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "hardware"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/engine",
+          "name": "engine",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "engine"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/profile",
+          "name": "profile",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "profile"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/concurrency",
+          "name": "concurrency",
+          "dataType": "sc:Integer",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "concurrency"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/request_throughput",
+          "name": "request_throughput",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "request_throughput"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/median_ttft_ms",
+          "name": "median_ttft_ms",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "median_ttft_ms"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "trace-replay-summary/median_tpot_ms",
+          "name": "median_tpot_ms",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "trace-replay-summary-parquet"
+            },
+            "extract": {
+              "column": "median_tpot_ms"
+            }
+          }
+        }
+      ]
+    },
+    {
+      "@type": "cr:RecordSet",
+      "@id": "synthetic-distributional-summary",
+      "name": "Synthetic Distributional Summary",
+      "description": "One row per benchmark configuration from synthetic distributional runs (statistical sampling from lognormal fits to recorded workload statistics).",
+      "field": [
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/run_id",
+          "name": "run_id",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "run_id"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/model",
+          "name": "model",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "model"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/hardware",
+          "name": "hardware",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "hardware"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/engine",
+          "name": "engine",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "engine"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/profile",
+          "name": "profile",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "profile"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/concurrency",
+          "name": "concurrency",
+          "dataType": "sc:Integer",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "concurrency"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/request_throughput",
+          "name": "request_throughput",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "request_throughput"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/median_ttft_ms",
+          "name": "median_ttft_ms",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "median_ttft_ms"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "synthetic-distributional-summary/median_tpot_ms",
+          "name": "median_tpot_ms",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "synthetic-distributional-summary-parquet"
+            },
+            "extract": {
+              "column": "median_tpot_ms"
+            }
+          }
+        }
+      ]
+    },
+    {
+      "@type": "cr:RecordSet",
+      "@id": "kernels-labeled",
+      "name": "Kernels Labeled",
+      "description": "Per-kernel Nsight Compute (ncu) hardware counter profiles. Individual kernel invocations across 4 GPUs and 13 model/sweep sources.",
+      "field": [
+        {
+          "@type": "cr:Field",
+          "@id": "kernels-labeled/source",
+          "name": "source",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "kernels-labeled-parquet"
+            },
+            "extract": {
+              "column": "source"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "kernels-labeled/gpu",
+          "name": "gpu",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "kernels-labeled-parquet"
+            },
+            "extract": {
+              "column": "gpu"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "kernels-labeled/model",
+          "name": "model",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "kernels-labeled-parquet"
+            },
+            "extract": {
+              "column": "model"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "kernels-labeled/kernel_family",
+          "name": "kernel_family",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "kernels-labeled-parquet"
+            },
+            "extract": {
+              "column": "kernel_family"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "kernels-labeled/kernel_name",
+          "name": "kernel_name",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "kernels-labeled-parquet"
+            },
+            "extract": {
+              "column": "kernel_name"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "kernels-labeled/gpu_time_duration_ms",
+          "name": "gpu_time_duration_ms",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "kernels-labeled-parquet"
+            },
+            "extract": {
+              "column": "gpu_time_duration_ms"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "kernels-labeled/dram_bytes_sum",
+          "name": "dram_bytes_sum",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "kernels-labeled-parquet"
+            },
+            "extract": {
+              "column": "dram_bytes_sum"
+            }
+          }
+        }
+      ]
+    },
+    {
+      "@type": "cr:RecordSet",
+      "@id": "mse-validation-summary",
+      "name": "MSE Validation Summary",
+      "description": "Curated validation rows for the distributional synthetic replay generator. Paired synthetic and real trace replay runs on H100 / Llama-3.1-8B / vLLM.",
+      "field": [
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/validation_id",
+          "name": "validation_id",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "validation_id"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/tier",
+          "name": "tier",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "tier"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/stage",
+          "name": "stage",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "stage"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/workload",
+          "name": "workload",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "workload"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/profile",
+          "name": "profile",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "profile"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/model",
+          "name": "model",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "model"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/hardware",
+          "name": "hardware",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "hardware"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/engine",
+          "name": "engine",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "engine"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/concurrency",
+          "name": "concurrency",
+          "dataType": "sc:Integer",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "concurrency"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/success_rate",
+          "name": "success_rate",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "success_rate"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/request_throughput",
+          "name": "request_throughput",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "request_throughput"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/successful_requests",
+          "name": "successful_requests",
+          "dataType": "sc:Integer",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "successful_requests"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/failed_requests",
+          "name": "failed_requests",
+          "dataType": "sc:Integer",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "failed_requests"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/median_ttft_ms",
+          "name": "median_ttft_ms",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "median_ttft_ms"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "mse-validation-summary/median_tpot_ms",
+          "name": "median_tpot_ms",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "mse-validation-summary-parquet"
+            },
+            "extract": {
+              "column": "median_tpot_ms"
+            }
+          }
+        }
+      ]
+    },
+    {
+      "@type": "cr:RecordSet",
+      "@id": "per-layer-kernel-summary",
+      "name": "Per-Layer Kernel Summary",
+      "description": "Per-component analytical OI decomposition and NCU kernel profiles for Llama-3.1-8B on H100 (prefill phase).",
+      "field": [
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/record_type",
+          "name": "record_type",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "record_type"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/model",
+          "name": "model",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "model"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/hardware",
+          "name": "hardware",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "hardware"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/phase",
+          "name": "phase",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "phase"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/batch_size",
+          "name": "batch_size",
+          "dataType": "sc:Integer",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "batch_size"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/sequence_length",
+          "name": "sequence_length",
+          "dataType": "sc:Integer",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "sequence_length"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/component_name",
+          "name": "component_name",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "component_name"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/bound",
+          "name": "bound",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "bound"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/flops",
+          "name": "flops",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "flops"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/bytes_accessed",
+          "name": "bytes_accessed",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "bytes_accessed"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/operational_intensity",
+          "name": "operational_intensity",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "operational_intensity"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/kernel_name",
+          "name": "kernel_name",
+          "dataType": "sc:Text",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "kernel_name"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/duration_us",
+          "name": "duration_us",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "duration_us"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/compute_sm_throughput_pct",
+          "name": "compute_sm_throughput_pct",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "compute_sm_throughput_pct"
+            }
+          }
+        },
+        {
+          "@type": "cr:Field",
+          "@id": "per-layer-kernel-summary/dram_throughput_pct",
+          "name": "dram_throughput_pct",
+          "dataType": "sc:Float",
+          "source": {
+            "fileObject": {
+              "@id": "per-layer-kernel-summary-parquet"
+            },
+            "extract": {
+              "column": "dram_throughput_pct"
+            }
+          }
+        }
+      ]
+    }
+  ],
+  "rai:dataLimitations": "Results cover NVIDIA H100, A100, RTX 3090, and RTX 2080 Ti GPUs only and may not generalize to other accelerators (AMD, Intel, TPU). Benchmark configurations are pinned to vLLM 0.19.0 and SGLang 0.5.9; results do not represent other engine versions. Concurrency levels (1-320) may not cover extreme-scale deployments. Not recommended as sole basis for hardware purchasing decisions or for comparing model task quality.",
+  "rai:dataBiases": "Model selection over-represents Meta Llama and Alibaba Qwen families. Hardware is exclusively NVIDIA GPUs. Workload profiles are author-designed approximations of production traffic; real deployment patterns may differ.",
+  "rai:personalSensitiveInformation": "No personally identifiable information is present. All API endpoints and credentials are stripped. Workload traces use synthetic random tokens or publicly available coding benchmarks.",
+  "rai:dataUseCases": "Established uses: relative comparison of inference engine throughput, latency benchmarking under controlled conditions, TTFT scaling with context length in multi-turn sessions, per-kernel performance modelling. Not established: absolute latency prediction for production, model quality comparison, cost estimation.",
+  "rai:dataSocialImpact": "Enables reproducible comparison of open-source LLM serving systems, supporting infrastructure research and reducing vendor lock-in.",
+  "rai:hasSyntheticData": true,
+  "prov:wasDerivedFrom": [
+    {
+      "@id": "https://huggingface.co/datasets/princeton-nlp/SWE-bench",
+      "name": "SWE-Bench"
+    },
+    {
+      "@id": "https://github.com/TerminalBench/TerminalBench",
+      "name": "TerminalBench"
+    },
+    {
+      "@id": "https://github.com/xlang-ai/OSWorld",
+      "name": "OSWorld"
+    },
+    {
+      "@id": "https://huggingface.co/datasets/Aeala/ShareGPT_Vicuna_unfiltered",
+      "name": "ShareGPT"
+    }
+  ],
+  "prov:wasGeneratedBy": {
+    "@type": "prov:Activity",
+    "name": "AgentPerfBench benchmark collection",
+    "description": "Deploy model on target GPU with specified engine and tensor parallelism. Send requests per configuration after warmup using closed-loop concurrency control. Record per-request TTFT, TPOT, ITL, E2EL, and token counts. Compute summary statistics. Sanitize credentials and convert to Parquet."
+  },
+  "rai:dataCollection": "Automated benchmark collection. Models deployed on target GPU with specified engine and tensor parallelism. Requests sent per workload profile using closed-loop concurrency control with semaphore. Per-request TTFT, TPOT, ITL, E2EL, and token counts recorded. Kernel profiles collected via Nsight Compute (ncu) with hardware counter metrics.",
+  "rai:dataCollectionType": "Automated measurement via benchmarking scripts. No human subjects or crowdsourcing.",
+  "rai:dataCollectionMissingData": "Some model-hardware-concurrency combinations are absent due to OOM or engine incompatibility. Concurrency levels exceeding the session-pool size were filtered (see Concurrency filtering in README).",
+  "rai:dataPreprocessing": "Credentials and API endpoints stripped. Configurations with fewer than 75% of requests completing successfully are excluded. Summary statistics computed from successful requests. Kernel names demangled from NCU output.",
+  "rai:dataAnnotationProtocol": "Not applicable. Fully automated benchmark collection with no human annotation.",
+  "rai:dataAnnotationPlatform": "Not applicable. No human annotation.",
+  "rai:dataAnnotationAnalysis": "Not applicable. No human annotation."
+}

per_layer_kernel/summary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c78e4694009da435587bc9e7e3e4821fca6a3bf881284f873b71ccdb5ed47691
+size 17496

synthetic_distributional/summary.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:67d63072831143e929e0b6d6464744756f783b53ac30fed0232bc01c601676ec
-size 68747
+oid sha256:0fd6ff9f802f9465a4618dfe8d5d426013028565e08ef5024f4302b9126a05a5
+size 64517

trace_replay/summary.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:19dfc3ed8e2f09c52449045cded6bce29a00c0b9c4e563a407ed8e9960333da3
-size 633996
+oid sha256:568efbdae473a830ba792317fb09cfba0ea4f667db67b93d28e72637a5c71d1e
+size 639880