Hugging Face's logo

qualcomm
/
Whisper-Medium

Automatic Speech Recognition
PyTorch
foundation
android
Model card Files
xet
 Community

Whisper-Medium: Optimized for Qualcomm Devices

HuggingFace Whisper-Medium ASR (Automatic Speech Recognition) model is a state-of-the-art system designed for transcribing spoken language into written text. This model is based on the transformer architecture and has been optimized for edge inference by replacing Multi-Head Attention (MHA) with Single-Head Attention (SHA) and linear layers with convolutional (conv) layers. It exhibits robust performance in realistic, noisy environments, making it highly reliable for real-world applications. Specifically, it excels in long-form transcription, capable of accurately transcribing audio clips up to 30 seconds long. Time to the first token is the encoder's latency, while time to each additional token is decoder's latency, where we assume a max decoded length specified below.

This is based on the implementation of Whisper-Medium found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up to run these models on a hosted Qualcomm® device.

Deploying Whisper-Medium on-device

This model is compatible with the Qualcomm Voice AI SDK. Download the SDK from the Qualcomm Package Manager to deploy this model on-device.

Getting Started

There are two ways to deploy this model on your device:

Option 1: Download Pre-Exported Models

Below are pre-exported model assets ready for deployment.

Runtime Precision Chipset SDK Versions Download
PRECOMPILED_QNN_ONNX float Snapdragon® X2 Elite QAIRT 2.45, ONNX Runtime 1.27.1 Download
PRECOMPILED_QNN_ONNX float Snapdragon® X Elite QAIRT 2.45, ONNX Runtime 1.27.1 Download
PRECOMPILED_QNN_ONNX float Snapdragon® 8 Gen 3 Mobile QAIRT 2.45, ONNX Runtime 1.27.1 Download
PRECOMPILED_QNN_ONNX float Snapdragon® 8 Gen 1 Mobile QAIRT 2.45 Download
PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-8275 QAIRT 2.45, ONNX Runtime 1.27.1 Download
PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) QAIRT 2.45, ONNX Runtime 1.27.1 Download
PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-9075 QAIRT 2.45, ONNX Runtime 1.27.1 Download
PRECOMPILED_QNN_ONNX float Snapdragon® 8 Elite Mobile QAIRT 2.45, ONNX Runtime 1.27.1 Download
PRECOMPILED_QNN_ONNX float Snapdragon® 8 Elite Gen 5 Mobile QAIRT 2.45, ONNX Runtime 1.27.1 Download
QNN_CONTEXT_BINARY float Snapdragon® X2 Elite QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Snapdragon® X Elite QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Snapdragon® 8 Gen 3 Mobile QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Snapdragon® 8 Gen 1 Mobile QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-8275 QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ QCS8550 (Proxy) QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Qualcomm® SA8775P QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-9075 QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Qualcomm® SA7255P QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Qualcomm® SA8295P QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Snapdragon® 8 Elite Mobile QAIRT 2.45 Download
QNN_CONTEXT_BINARY float Snapdragon® 8 Elite Gen 5 Mobile QAIRT 2.45 Download
VOICE_AI float Snapdragon® X2 Elite QAIRT 2.45 Download
VOICE_AI float Snapdragon® X Elite QAIRT 2.45 Download
VOICE_AI float Snapdragon® 8 Gen 3 Mobile QAIRT 2.45 Download
VOICE_AI float Snapdragon® 8 Gen 1 Mobile QAIRT 2.45 Download
VOICE_AI float Qualcomm® Dragonwing™ IQ-8275 QAIRT 2.45 Download
VOICE_AI float Qualcomm® Dragonwing™ QCS8550 (Proxy) QAIRT 2.45 Download
VOICE_AI float Qualcomm® SA8775P QAIRT 2.45 Download
VOICE_AI float Qualcomm® Dragonwing™ IQ-9075 QAIRT 2.45 Download
VOICE_AI float Qualcomm® SA7255P QAIRT 2.45 Download
VOICE_AI float Qualcomm® SA8295P QAIRT 2.45 Download
VOICE_AI float Snapdragon® 8 Elite Mobile QAIRT 2.45 Download
VOICE_AI float Snapdragon® 8 Elite Gen 5 Mobile QAIRT 2.45 Download

For more device-specific assets and performance metrics, visit Whisper-Medium on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models Python library to compile and export the model with your own:

  • Custom weights (e.g., fine-tuned checkpoints)
  • Custom input shapes
  • Target device and runtime configurations

This option is ideal if you need to customize the model beyond the default configuration provided here.

See our repository for Whisper-Medium on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.speech_recognition

Model Stats:

  • Input resolution: 80x3000 (30 seconds audio)
  • Max decoded sequence length: 200 tokens
  • Model checkpoint: openai/whisper-medium
  • Model size (decoder) (float): 1.54GB
  • Model size (encoder) (float): 1.54GB
  • Number of parameters (decoder): 385M
  • Number of parameters (encoder): 384M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
decoder PRECOMPILED_QNN_ONNX float Snapdragon® X2 Elite 18.082 ms 159 - 159 MB NPU
decoder PRECOMPILED_QNN_ONNX float Snapdragon® X Elite 27.964 ms 725 - 725 MB NPU
decoder PRECOMPILED_QNN_ONNX float Snapdragon® 8 Gen 3 Mobile 28.729 ms 200 - 211 MB NPU
decoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-8275 39.629 ms 159 - 323 MB NPU
decoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 33.702 ms 0 - 733 MB NPU
decoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-9075 37.766 ms 160 - 322 MB NPU
decoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-X7181 27.964 ms 725 - 725 MB NPU
decoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ Q-8750 24.063 ms 150 - 162 MB NPU
decoder PRECOMPILED_QNN_ONNX float Snapdragon® 8 Elite Mobile 24.063 ms 150 - 162 MB NPU
decoder PRECOMPILED_QNN_ONNX float Snapdragon® 8 Elite Gen 5 Mobile 22.031 ms 153 - 166 MB NPU
decoder QNN_CONTEXT_BINARY float Snapdragon® X2 Elite 18.97 ms 160 - 160 MB NPU
decoder QNN_CONTEXT_BINARY float Snapdragon® X Elite 27.533 ms 160 - 160 MB NPU
decoder QNN_CONTEXT_BINARY float Snapdragon® 8 Gen 3 Mobile 28.835 ms 134 - 143 MB NPU
decoder QNN_CONTEXT_BINARY float Snapdragon® 8 Gen 1 Mobile 58.465 ms 160 - 177 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-8275 39.416 ms 160 - 340 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ QCS8550 (Proxy) 33.723 ms 121 - 141 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® SA8775P 36.697 ms 160 - 172 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® SA8650P 36.697 ms 160 - 172 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® SA8255P 36.697 ms 160 - 172 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® QCS8450 58.465 ms 160 - 177 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-9075 37.0 ms 160 - 340 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-X7181 27.533 ms 160 - 160 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ Q-8750 23.6 ms 84 - 98 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® SA7255P 51.657 ms 144 - 155 MB NPU
decoder QNN_CONTEXT_BINARY float Qualcomm® SA8295P 38.566 ms 109 - 115 MB NPU
decoder QNN_CONTEXT_BINARY float Snapdragon® 8 Elite Mobile 23.6 ms 84 - 98 MB NPU
decoder QNN_CONTEXT_BINARY float Snapdragon® 8 Elite Gen 5 Mobile 21.24 ms 158 - 173 MB NPU
decoder VOICE_AI float Snapdragon® X2 Elite 18.785 ms 160 - 160 MB NPU
decoder VOICE_AI float Snapdragon® X Elite 27.354 ms 160 - 160 MB NPU
decoder VOICE_AI float Snapdragon® 8 Gen 3 Mobile 28.295 ms 135 - 143 MB NPU
decoder VOICE_AI float Snapdragon® 8 Gen 1 Mobile 57.513 ms 160 - 177 MB NPU
decoder VOICE_AI float Qualcomm® Dragonwing™ IQ-8275 39.434 ms 160 - 340 MB NPU
decoder VOICE_AI float Qualcomm® Dragonwing™ QCS8550 (Proxy) 33.318 ms 136 - 138 MB NPU
decoder VOICE_AI float Qualcomm® SA8775P 36.842 ms 146 - 158 MB NPU
decoder VOICE_AI float Qualcomm® SA8650P 36.842 ms 146 - 158 MB NPU
decoder VOICE_AI float Qualcomm® SA8255P 36.842 ms 146 - 158 MB NPU
decoder VOICE_AI float Qualcomm® QCS8450 57.513 ms 160 - 177 MB NPU
decoder VOICE_AI float Qualcomm® Dragonwing™ IQ-9075 36.538 ms 160 - 340 MB NPU
decoder VOICE_AI float Qualcomm® Dragonwing™ IQ-X7181 27.354 ms 160 - 160 MB NPU
decoder VOICE_AI float Qualcomm® Dragonwing™ Q-8750 23.51 ms 89 - 103 MB NPU
decoder VOICE_AI float Qualcomm® SA7255P 51.806 ms 147 - 158 MB NPU
decoder VOICE_AI float Qualcomm® SA8295P 37.873 ms 112 - 118 MB NPU
decoder VOICE_AI float Snapdragon® 8 Elite Mobile 23.51 ms 89 - 103 MB NPU
decoder VOICE_AI float Snapdragon® 8 Elite Gen 5 Mobile 21.027 ms 86 - 97 MB NPU
encoder PRECOMPILED_QNN_ONNX float Snapdragon® X2 Elite 163.054 ms 528 - 528 MB NPU
encoder PRECOMPILED_QNN_ONNX float Snapdragon® X Elite 367.443 ms 698 - 698 MB NPU
encoder PRECOMPILED_QNN_ONNX float Snapdragon® 8 Gen 3 Mobile 273.953 ms 305 - 317 MB NPU
encoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-8275 470.986 ms 163 - 167 MB NPU
encoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 365.448 ms 0 - 770 MB NPU
encoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-9075 440.552 ms 163 - 166 MB NPU
encoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ IQ-X7181 367.443 ms 698 - 698 MB NPU
encoder PRECOMPILED_QNN_ONNX float Qualcomm® Dragonwing™ Q-8750 229.982 ms 307 - 318 MB NPU
encoder PRECOMPILED_QNN_ONNX float Snapdragon® 8 Elite Mobile 229.982 ms 307 - 318 MB NPU
encoder PRECOMPILED_QNN_ONNX float Snapdragon® 8 Elite Gen 5 Mobile 151.677 ms 304 - 317 MB NPU
encoder QNN_CONTEXT_BINARY float Snapdragon® X2 Elite 153.148 ms 0 - 0 MB NPU
encoder QNN_CONTEXT_BINARY float Snapdragon® X Elite 364.918 ms 0 - 0 MB NPU
encoder QNN_CONTEXT_BINARY float Snapdragon® 8 Gen 3 Mobile 276.544 ms 61 - 69 MB NPU
encoder QNN_CONTEXT_BINARY float Snapdragon® 8 Gen 1 Mobile 758.474 ms 0 - 13 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-8275 471.127 ms 0 - 144 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ QCS8550 (Proxy) 362.368 ms 0 - 135 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® SA8775P 441.603 ms 1 - 12 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® SA8650P 441.603 ms 1 - 12 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® SA8255P 441.603 ms 1 - 12 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® QCS8450 758.474 ms 0 - 13 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-9075 437.179 ms 0 - 143 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ IQ-X7181 364.918 ms 0 - 0 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® Dragonwing™ Q-8750 220.298 ms 1 - 14 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® SA7255P 1339.536 ms 1 - 11 MB NPU
encoder QNN_CONTEXT_BINARY float Qualcomm® SA8295P 508.791 ms 1 - 6 MB NPU
encoder QNN_CONTEXT_BINARY float Snapdragon® 8 Elite Mobile 220.298 ms 1 - 14 MB NPU
encoder QNN_CONTEXT_BINARY float Snapdragon® 8 Elite Gen 5 Mobile 151.752 ms 25 - 34 MB NPU
encoder VOICE_AI float Snapdragon® X2 Elite 154.199 ms 0 - 0 MB NPU
encoder VOICE_AI float Snapdragon® X Elite 365.046 ms 0 - 0 MB NPU
encoder VOICE_AI float Snapdragon® 8 Gen 3 Mobile 275.659 ms 1 - 8 MB NPU
encoder VOICE_AI float Snapdragon® 8 Gen 1 Mobile 748.409 ms 0 - 11 MB NPU
encoder VOICE_AI float Qualcomm® Dragonwing™ IQ-8275 471.328 ms 0 - 144 MB NPU
encoder VOICE_AI float Qualcomm® Dragonwing™ QCS8550 (Proxy) 363.144 ms 0 - 2 MB NPU
encoder VOICE_AI float Qualcomm® SA8775P 441.333 ms 1 - 12 MB NPU
encoder VOICE_AI float Qualcomm® SA8650P 441.333 ms 1 - 12 MB NPU
encoder VOICE_AI float Qualcomm® SA8255P 441.333 ms 1 - 12 MB NPU
encoder VOICE_AI float Qualcomm® QCS8450 748.409 ms 0 - 11 MB NPU
encoder VOICE_AI float Qualcomm® Dragonwing™ IQ-9075 506.869 ms 0 - 143 MB NPU
encoder VOICE_AI float Qualcomm® Dragonwing™ IQ-X7181 365.046 ms 0 - 0 MB NPU
encoder VOICE_AI float Qualcomm® Dragonwing™ Q-8750 220.789 ms 1 - 13 MB NPU
encoder VOICE_AI float Qualcomm® SA7255P 1340.474 ms 1 - 11 MB NPU
encoder VOICE_AI float Qualcomm® SA8295P 509.071 ms 1 - 6 MB NPU
encoder VOICE_AI float Snapdragon® 8 Elite Mobile 220.789 ms 1 - 13 MB NPU
encoder VOICE_AI float Snapdragon® 8 Elite Gen 5 Mobile 149.542 ms 1 - 14 MB NPU

License

  • The license for the original implementation of Whisper-Medium can be found here.

References

Community

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
Automatic Speech Recognition
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support