Profile Job Results

Jobs
j1p8jeyx5
Results Ready
Name
whisper_tiny_en_WhisperDecoder
Target Device
  • SA8775 (Proxy)
  • Android 13
  • Qualcomm® SA8775P
Creator
ai-hub-support@qti.qualcomm.com
Target Model
Input Specs
x: int32[1, 1]
index: int32[1, 1]
k_cache_cross: float32[4, 6, 64, 1500]
v_cache_cross: float32[4, 6, 1500, 64]
k_cache_self: float32[4, 6, 64, 224]
v_cache_self: float32[4, 6, 224, 64]
Completion Time
8/11/2024, 5:48:33 AM
Versions
  • TensorFlow Lite: 2.16.1
  • QNN TfLite Delegate: v2.24.0.240626131148_96320
  • Android: 13 (TP1A.220624.014)
  • AI Hub: aihub-2024.08.01.0
Estimated Inference Time
10.2 ms
Estimated Peak Memory Usage
21 - 26 MB
Compute Units
NPU
552
CPU
4
GPU
1
StageTimeMemory
First App Load
1.72 s288-289 MB
Subsequent App Load
377 ms62-124 MB
Inference
10.2 ms21-26 MB
TensorFlow LiteValue
number_of_threads4
QNN DelegateValue
backend_typekHtpBackend
log_levelkLogLevelWarn
htp_options.performance_modekHtpBurst
htp_options.precisionkHtpFp16
htp_options.useConvHmxtrue
GPUv2 DelegateValue
inference_preferenceTFLITE_GPU_INFERENCE_PREFERENCE_SUSTAINED_SPEED
inference_priority1TFLITE_GPU_INFERENCE_PRIORITY_MIN_LATENCY
inference_priority2TFLITE_GPU_INFERENCE_PRIORITY_MIN_MEMORY_USAGE
inference_priority3TFLITE_GPU_INFERENCE_PRIORITY_MAX_PRECISION
XNNPACK DelegateValue

Sign up to run this model on a hosted Qualcomm® device!

Run on device