Repair dataset card configs after MSE merge

README.md

@@ -12,7 +12,7 @@ tags:
   - sglang
   - agentic-workloads
 size_categories:
-  - 1K<n<10K
+  - 100K<n<1M
 pretty_name: AgentPerfBench
 version: "1.0"
 configs:
@@ -20,10 +20,14 @@ configs:
     data_files:
       - split: summary
         path: trace_replay/summary.parquet
-  - config_name: distributional
+  - config_name: synthetic_distributional
     data_files:
       - split: summary
-        path: distributional/summary.parquet
+        path: synthetic_distributional/summary.parquet
+  - config_name: kernels_labeled
+    data_files:
+      - split: train
+        path: kernel_profiles/kernels_labeled.parquet
   - config_name: mse_validation
     data_files:
       - split: summary
@@ -99,7 +103,7 @@ dataset_info:
       - name: summary
         num_examples: 3147
         num_bytes: 694254
-  - config_name: distributional
+  - config_name: synthetic_distributional
     features:
       - name: run_id
         dtype: string
@@ -169,6 +173,55 @@ dataset_info:
       - name: summary
         num_examples: 245
         num_bytes: 70836
+  - config_name: kernels_labeled
+    features:
+      - name: source
+        dtype: string
+      - name: gpu
+        dtype: string
+      - name: model
+        dtype: string
+      - name: kernel_family
+        dtype: string
+      - name: kernel_name
+        dtype: string
+      - name: dtype
+        dtype: string
+      - name: held_out
+        dtype: bool
+      - name: M
+        dtype: float64
+      - name: N
+        dtype: float64
+      - name: K
+        dtype: float64
+      - name: bs
+        dtype: float64
+      - name: seq
+        dtype: float64
+      - name: n_heads
+        dtype: float64
+      - name: head_dim
+        dtype: float64
+      - name: kv_heads
+        dtype: float64
+      - name: numel
+        dtype: float64
+      - name: op_type
+        dtype: string
+      - name: gpu_time_duration_ms
+        dtype: float64
+      - name: launch_block_size
+        dtype: float64
+      - name: launch_grid_size
+        dtype: float64
+      - name: dram_bytes_sum
+        dtype: float64
+      - name: launch_registers_per_thread
+        dtype: float64
+    splits:
+      - name: train
+        num_examples: 148077
   - config_name: mse_validation
     features:
       - name: validation_id
@@ -279,11 +332,11 @@ dataset_info:
 
 # AgentPerfBench
 
-LLM inference benchmark: 3,392 main sweep runs 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 28 curated MSE validation rows for the distributional synthetic replay generator.
+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 and 28 curated MSE validation rows for the distributional synthetic replay generator.
 
 ## Dataset configurations
 
-The dataset provides three configurations. *trace_replay* replays exact input/output sequences from recorded agent sessions. *distributional* samples from statistical distributions fitted to those same workloads, trading fidelity for faster sweeps across the hardware matrix. *mse_validation* contains paired synthetic-vs-real validation runs and ablations for the final APC-aware synthetic generator.
+The dataset provides four 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 paired synthetic-vs-real validation runs and ablations for the final APC-aware synthetic generator.
 
 ### trace_replay (3,147 rows)
 
@@ -291,12 +344,16 @@ Replays exact ISL/OSL sequences from recorded agent sessions (SWE-Bench, Termina
 
 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`
 
-### distributional (245 rows)
+### synthetic_distributional (245 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.
 
 6 profiles: `chat-multiturn`, `chat-singleturn`, `coding-singleturn`, `osworld-multiturn`, `swebench-multiturn`, `terminalbench-multiturn`
 
+### kernels_labeled (148,077 rows)
+
+Per-kernel Nsight Compute (ncu) profiles across 4 GPUs (A100, H100, RTX 3090, RTX 2080 Ti) and 13 model/sweep sources.
+
 ### mse_validation (28 rows)
 
 Curated H100 / Llama-3.1-8B / vLLM validation artifacts 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.
@@ -317,18 +374,18 @@ See `mse_validation/README.md`, `mse_validation/manifest.csv`, and `mse_validati
 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 (session pool was 100). Remaining values: {1, 5, 10, 20, 40, 80}.
-- 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}.
+- 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}.
 
 | Config | Rows |
 |--------|------|
 | trace_replay | 3,147 |
-| distributional | 245 |
+| synthetic_distributional | 245 |
 | mse_validation | 28 |
-| **Total** | **3,420** |
+| 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 distributional rows have `failed_requests > 0`. Summary metrics (TTFT, TPOT, throughput) are computed from successful requests only.
+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.
 
 ## Coverage
 
@@ -370,7 +427,7 @@ Model names in this table match the `model` column in the parquet files.
 
 ## Schema
 
-Each row in `summary.parquet` (both configs):
+Each row in the serving `summary.parquet` configs:
 
 | Column | Type | Description |
 |--------|------|-------------|
@@ -401,13 +458,13 @@ Each row in `summary.parquet` (both configs):
 from datasets import load_dataset
 
 ds = load_dataset("agent-perf-bench/AgentPerfBench", "trace_replay")
-# or "distributional" / "mse_validation"
+# or "synthetic_distributional", "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} (distributional).
+- 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.
@@ -417,7 +474,7 @@ ds = load_dataset("agent-perf-bench/AgentPerfBench", "trace_replay")
 ## 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 included for `mse_validation`.
-- Per-kernel CUDA roofline profiles (PyTorch profiler, 2-layer forward passes, batch sizes 1/4/8/32/64).
+- Additional per-kernel roofline profiles beyond the included `kernels_labeled` and `per_layer_oi_cf` artifacts.
 - This is version 1.0. Updates will be tagged with semantic versions.
 
 ## Intended uses
@@ -429,10 +486,10 @@ ds = load_dataset("agent-perf-bench/AgentPerfBench", "trace_replay")
 ## 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).
-- Distributional profiles approximate but do not replicate production traffic patterns.
-- No consumer GPUs beyond RTX 3090; no non-NVIDIA accelerators.
+- 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.
 - 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 distributional). Not all model-hardware combinations are represented.
+- 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