Profile Job Results
Jobs
j563r7v75
Results Ready
Name
distil_whisper_decoder
Target Device
- Dragonwing IQ-9075 EVK
- Qualcomm Linux 1.7
- Qualcomm® Dragonwing™ IQ-9075 | QCS9075
Creator
ai-hub-support@qti.qualcomm.com
Target Model
Input Specs
input_ids: int32[1, 1]attention_mask: float32[1, 1, 1, 200]k_cache_self_0_in: float32[12, 1, 64, 199]v_cache_self_0_in: float32[12, 1, 199, 64]k_cache_self_1_in: float32[12, 1, 64, 199]v_cache_self_1_in: float32[12, 1, 199, 64]k_cache_self_2_in: float32[12, 1, 64, 199]v_cache_self_2_in: float32[12, 1, 199, 64]k_cache_self_3_in: float32[12, 1, 64, 199]v_cache_self_3_in: float32[12, 1, 199, 64]k_cache_cross_0: float32[12, 1, 64, 1500]v_cache_cross_0: float32[12, 1, 1500, 64]k_cache_cross_1: float32[12, 1, 64, 1500]v_cache_cross_1: float32[12, 1, 1500, 64]k_cache_cross_2: float32[12, 1, 64, 1500]v_cache_cross_2: float32[12, 1, 1500, 64]k_cache_cross_3: float32[12, 1, 64, 1500]v_cache_cross_3: float32[12, 1, 1500, 64]position_ids: int32[1]Completion Time
4/5/2026, 10:32:17 AM
Versions
- LiteRT: 1.4.2
- QAIRT: v2.43.0.260127150333_193827
- QNN TfLite Delegate: v2.43.0.260127150333_193827
- Qualcomm Linux: 1.7-ver.1.1
- AI Hub: aihub-2026.03.29.0
Estimated Inference Time
13.1 ms
Estimated Peak Memory Usage
2 ‑ 266 MB
Compute Units
NPU
1080
| Stage | Time | Memory |
|---|---|---|
First App Load | 6.27 s | 636‑637 MB |
Subsequent App Load | 321 ms | 221‑223 MB |
Inference | 13.1 ms | 2‑266 MB |
| TensorFlow Lite | Value |
|---|---|
| number_of_threads | 4 |
| QNN Delegate | Value |
|---|---|
| backend_type | kHtpBackend |
| log_level | kLogLevelWarn |
| htp_options.performance_mode | kHtpBurst |
| htp_options.precision | kHtpFp16 |
| htp_options.optimization_strategy | kHtpOptimizeForInference |
| htp_options.useConvHmx | true |
Sign up to run this model on a hosted Qualcomm® device!
Run on device