Profile Job Results
Jobs
j5m0y1j9g
Results Ready
Name
distil_bert_base_uncased_hf
Target Device
- QCS8550 (Proxy)
- Android 12
- Qualcomm® Dragonwing™ QCS8550
Creator
ai-hub-support@qti.qualcomm.com
Input Specs
input_tokens: int32[1, 384]attention_masks: uint16[1, 384]mask_indices: int32[1]Completion Time
10/3/2026, 9:56:19 AM
Versions
- ONNX Runtime: 1.30.0
- QAIRT: v2.50.0.260828221209
- Android: 13 (TP1A.220624.014)
- AI Hub: aihub-2026.09.25.0
Estimated Inference Time
8.25 ms
Estimated Peak Memory Usage
0 ‑ 115 MB
Compute Units
NPU
219
| Stage | Time | Memory |
|---|---|---|
First App Load | 7.13 s | 665‑666 MB |
Subsequent App Load | 500 ms | 112‑521 MB |
Inference | 8.25 ms | 0‑115 MB |
| ONNX Runtime | Value |
|---|---|
| execution_mode | SEQUENTIAL |
| intra_op_num_threads | 0 |
| inter_op_num_threads | 0 |
| enable_memory_pattern | false |
| enable_cpu_memory_arena | false |
| graph_optimization_level | ENABLE_ALL |
| QNN Execution Provider | Value |
|---|---|
| htp_performance_mode | "burst" |
| htp_graph_finalization_optimization_mode | "3" |
| enable_htp_fp16_precision | "1" |
| capture_network_visualizations | false |
| context_priority | "normal" |
| offload_graph_io_quantization | "1" |
Sign up to run this model on a hosted Qualcomm® device!
Run on device