+

Deformable DETR Multi-Scale Deformable Attention Benchmarks - Aggregated Results

+

This document combines benchmark results from multiple Deformable DETR implementations.

+

Combined Summary and Visualization

+
+ + + + + + + 2025-10-31T20:14:23.345627 + image/svg+xml + + + Matplotlib v3.10.7, https://matplotlib.org/ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + cuda_B1_Q100_H8_E256_L4_P4 + + + + + + + + + + + + + cuda_B1_Q300_H8_E256_L4_P4 + + + + + + + + + + + + + cuda_B2_Q100_H8_E256_L4_P4 + + + + + + + + + + + + + cuda_B2_Q300_H8_E256_L4_P4 + + + + Workload + + + + + + + + + + + + + + + + + 0.0 + + + + + + + + + + + + + 0.5 + + + + + + + + + + + + + 1.0 + + + + + + + + + + + + + 1.5 + + + + + + + + + + + + + 2.0 + + + + + + + + + + + + + 2.5 + + + + + + + + + + + + + 3.0 + + + + + + + + + + + + + 3.5 + + + + + + + + + + + + + 4.0 + + + + Latency P50 (ms) + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Attention Implementation Latency + + + + + + + + + + + + + hf_kernels_deformable_detr + + + + + + + + + torch_eager + + + + + + + + + + +
+ +
+
+ +▶ code +▼ output + ▶ uv-logs + | +Cell: combine | 4.34s + | + +Raw +
+ +
+
======================================================================
+LOADING BENCHMARK DATA
+======================================================================
+✓ HF Kernels Deformable DETR    : /__w/kernels-benchmarks/kernels-benchmarks/benches/deformable_detr/impls/.uvnote/cache/8ab95d7f8f4c6a375b95806e646e4e6f12f0749960d319cf7587215b378ccfa9
+✓ PyTorch Deformable DETR       : /__w/kernels-benchmarks/kernels-benchmarks/benches/deformable_detr/impls/.uvnote/cache/9c0a40cf66719a0b460ebb0ca3b41bcaf6c5486905bbf2045a65be2710694dfa
+
+  ✓ Found HF Kernels Deformable DETR
+     Path: /__w/kernels-benchmarks/kernels-benchmarks/benches/deformable_detr/impls/.uvnote/cache/8ab95d7f8f4c6a375b95806e646e4e6f12f0749960d319cf7587215b378ccfa9/deformable_detr.jsonl
+  ✓ Found PyTorch Deformable DETR
+     Path: /__w/kernels-benchmarks/kernels-benchmarks/benches/deformable_detr/impls/.uvnote/cache/9c0a40cf66719a0b460ebb0ca3b41bcaf6c5486905bbf2045a65be2710694dfa/deformable_detr.jsonl
+
+======================================================================
+Summary: 2 found, 0 skipped, 0 missing
+======================================================================
+
+COMBINED BENCHMARK SUMMARY
+
+impl                     wl                  p50(ms)  ok
+hf_kernels_deformable_detr cuda_B1_Q100_H8_E256_L4_P4     0.04  True
+hf_kernels_deformable_detr cuda_B1_Q300_H8_E256_L4_P4     0.05  True
+hf_kernels_deformable_detr cuda_B2_Q100_H8_E256_L4_P4     0.05  True
+hf_kernels_deformable_detr cuda_B2_Q300_H8_E256_L4_P4     0.05  True
+torch_eager              cuda_B1_Q100_H8_E256_L4_P4     3.39  True
+torch_eager              cuda_B1_Q300_H8_E256_L4_P4     4.01  True
+torch_eager              cuda_B2_Q100_H8_E256_L4_P4     4.02  True
+torch_eager              cuda_B2_Q300_H8_E256_L4_P4     4.02  True
+
+GENERATING COMBINED VISUALIZATION
+
+Loaded 8 records
+✓ Visualization saved as latency.svg
+Saved latency.png
+✓ Visualization saved as latency.svg
+✓ SVG visualization ready!
+
+ANALYSIS COMPLETE
+Total implementations analyzed: 2
+
+Implementations included:
+  ✓ HF Kernels Deformable DETR
+  ✓ PyTorch Deformable DETR
+
+
+
▶ UV Install Logs
+ +
+
+

Artifacts:

+latency.svg +
+ + + + + + + 2025-10-31T20:14:23.345627 + image/svg+xml + + + Matplotlib v3.10.7, https://matplotlib.org/ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + cuda_B1_Q100_H8_E256_L4_P4 + + + + + + + + + + + + + cuda_B1_Q300_H8_E256_L4_P4 + + + + + + + + + + + + + cuda_B2_Q100_H8_E256_L4_P4 + + + + + + + + + + + + + cuda_B2_Q300_H8_E256_L4_P4 + + + + Workload + + + + + + + + + + + + + + + + + 0.0 + + + + + + + + + + + + + 0.5 + + + + + + + + + + + + + 1.0 + + + + + + + + + + + + + 1.5 + + + + + + + + + + + + + 2.0 + + + + + + + + + + + + + 2.5 + + + + + + + + + + + + + 3.0 + + + + + + + + + + + + + 3.5 + + + + + + + + + + + + + 4.0 + + + + Latency P50 (ms) + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Attention Implementation Latency + + + + + + + + + + + + + hf_kernels_deformable_detr + + + + + + + + + torch_eager + + + + + + + + + + +
+
+
+
+