Hugging Face's logo

qualcomm
/
Whisper-Small-Quantized

Automatic Speech Recognition
PyTorch
foundation
android
Model card Files
xet
 Community

Whisper-Small-Quantized: Optimized for Qualcomm Devices

We have applied w8a16 quantization to significantly enhance performance and efficiency. HuggingFace Whisper-Small 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-Small-Quantized 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-Small-Quantized 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 w8a16 Snapdragon® 8 Elite Gen 5 Mobile QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Snapdragon® X2 Elite QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Snapdragon® X Elite QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Gen 3 Mobile QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS8550 (Proxy) QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Elite For Galaxy Mobile QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 7 Gen 4 Mobile QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCM6690 QAIRT 2.42, ONNX Runtime 1.25.0 Download
PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS9075 QAIRT 2.42, ONNX Runtime 1.25.0 Download
QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Elite Gen 5 Mobile QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Snapdragon® X2 Elite QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Snapdragon® X Elite QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Gen 3 Mobile QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS8550 (Proxy) QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Qualcomm® SA8775P QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Elite For Galaxy Mobile QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Snapdragon® 7 Gen 4 Mobile QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Qualcomm® SA7255P QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Qualcomm® QCM6690 QAIRT 2.45 Download
QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS9075 QAIRT 2.45 Download
VOICE_AI w8a16 Snapdragon® 8 Elite Gen 5 Mobile QAIRT 2.45 Download
VOICE_AI w8a16 Snapdragon® X2 Elite QAIRT 2.45 Download
VOICE_AI w8a16 Snapdragon® X Elite QAIRT 2.45 Download
VOICE_AI w8a16 Snapdragon® 8 Gen 3 Mobile QAIRT 2.45 Download
VOICE_AI w8a16 Qualcomm® QCS8550 (Proxy) QAIRT 2.45 Download
VOICE_AI w8a16 Qualcomm® SA8775P QAIRT 2.45 Download
VOICE_AI w8a16 Snapdragon® 8 Elite For Galaxy Mobile QAIRT 2.45 Download
VOICE_AI w8a16 Snapdragon® 7 Gen 4 Mobile QAIRT 2.45 Download
VOICE_AI w8a16 Qualcomm® SA7255P QAIRT 2.45 Download
VOICE_AI w8a16 Qualcomm® QCM6690 QAIRT 2.45 Download
VOICE_AI w8a16 Qualcomm® QCS9075 QAIRT 2.45 Download

For more device-specific assets and performance metrics, visit Whisper-Small-Quantized 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-Small-Quantized on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.speech_recognition

Model Stats:

  • Model checkpoint: openai/whisper-small
  • Input resolution: 80x3000 (30 seconds audio)
  • Max decoded sequence length: 200 tokens

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
decoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Elite Gen 5 Mobile 3.982 ms 39 - 50 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® X2 Elite 3.881 ms 33 - 33 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® X Elite 7.638 ms 187 - 187 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 6.375 ms 36 - 49 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS8550 (Proxy) 8.37 ms 27 - 30 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS9075 9.105 ms 24 - 57 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCM6690 32.297 ms 29 - 39 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Elite For Galaxy Mobile 4.79 ms 26 - 36 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 7 Gen 4 Mobile 10.855 ms 29 - 35 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS7790 10.855 ms 29 - 35 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS8750 4.79 ms 26 - 36 MB NPU
decoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS7181 7.638 ms 187 - 187 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Elite Gen 5 Mobile 3.909 ms 30 - 40 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Snapdragon® X2 Elite 4.173 ms 30 - 30 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Snapdragon® X Elite 7.651 ms 30 - 30 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Gen 3 Mobile 5.959 ms 30 - 38 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS8275 12.752 ms 20 - 27 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS8550 (Proxy) 7.913 ms 30 - 32 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA8775P 8.959 ms 20 - 30 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA8650P 8.959 ms 20 - 30 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA8255P 8.959 ms 20 - 30 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS9075 8.655 ms 25 - 60 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCM6690 30.17 ms 29 - 36 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Elite For Galaxy Mobile 4.589 ms 17 - 27 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA7255P 12.752 ms 20 - 27 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 7 Gen 4 Mobile 10.47 ms 29 - 36 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS7790 10.47 ms 29 - 36 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS8750 4.589 ms 17 - 27 MB NPU
decoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS7181 7.651 ms 30 - 30 MB NPU
decoder VOICE_AI w8a16 Snapdragon® 8 Elite Gen 5 Mobile 3.922 ms 30 - 40 MB NPU
decoder VOICE_AI w8a16 Snapdragon® X2 Elite 4.235 ms 30 - 30 MB NPU
decoder VOICE_AI w8a16 Snapdragon® X Elite 7.326 ms 30 - 30 MB NPU
decoder VOICE_AI w8a16 Snapdragon® 8 Gen 3 Mobile 5.884 ms 30 - 38 MB NPU
decoder VOICE_AI w8a16 Qualcomm® QCS8275 12.731 ms 30 - 39 MB NPU
decoder VOICE_AI w8a16 Qualcomm® QCS8550 (Proxy) 7.857 ms 30 - 32 MB NPU
decoder VOICE_AI w8a16 Qualcomm® SA8775P 8.958 ms 26 - 36 MB NPU
decoder VOICE_AI w8a16 Qualcomm® SA8650P 8.958 ms 26 - 36 MB NPU
decoder VOICE_AI w8a16 Qualcomm® SA8255P 8.958 ms 26 - 36 MB NPU
decoder VOICE_AI w8a16 Qualcomm® QCS9075 8.649 ms 25 - 60 MB NPU
decoder VOICE_AI w8a16 Qualcomm® QCM6690 35.596 ms 30 - 37 MB NPU
decoder VOICE_AI w8a16 Snapdragon® 8 Elite For Galaxy Mobile 4.659 ms 30 - 44 MB NPU
decoder VOICE_AI w8a16 Qualcomm® SA7255P 12.731 ms 30 - 39 MB NPU
decoder VOICE_AI w8a16 Snapdragon® 7 Gen 4 Mobile 10.445 ms 29 - 36 MB NPU
decoder VOICE_AI w8a16 Qualcomm® QCS7790 10.445 ms 29 - 36 MB NPU
decoder VOICE_AI w8a16 Qualcomm® QCS8750 4.659 ms 30 - 44 MB NPU
decoder VOICE_AI w8a16 Qualcomm® QCS7181 7.326 ms 30 - 30 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Elite Gen 5 Mobile 175.315 ms 63 - 73 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® X2 Elite 152.81 ms 67 - 67 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® X Elite 264.986 ms 128 - 128 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 260.066 ms 59 - 65 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS8550 (Proxy) 372.355 ms 54 - 57 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS9075 254.886 ms 63 - 66 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCM6690 4348.872 ms 2 - 12 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 8 Elite For Galaxy Mobile 198.029 ms 63 - 74 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Snapdragon® 7 Gen 4 Mobile 455.084 ms 57 - 69 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS7790 455.084 ms 57 - 69 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS8750 198.029 ms 63 - 74 MB NPU
encoder PRECOMPILED_QNN_ONNX w8a16 Qualcomm® QCS7181 264.986 ms 128 - 128 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Elite Gen 5 Mobile 190.571 ms 1 - 10 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Snapdragon® X2 Elite 156.327 ms 0 - 0 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Snapdragon® X Elite 301.833 ms 0 - 0 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Gen 3 Mobile 268.377 ms 1 - 7 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS8275 525.354 ms 1 - 7 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS8550 (Proxy) 381.169 ms 1 - 3 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA8775P 317.35 ms 1 - 9 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA8650P 317.35 ms 1 - 9 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA8255P 317.35 ms 1 - 9 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS9075 296.199 ms 0 - 29 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCM6690 4411.617 ms 1 - 13 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 8 Elite For Galaxy Mobile 226.548 ms 1 - 10 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® SA7255P 525.354 ms 1 - 7 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Snapdragon® 7 Gen 4 Mobile 479.389 ms 0 - 7 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS7790 479.389 ms 0 - 7 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS8750 226.548 ms 1 - 10 MB NPU
encoder QNN_CONTEXT_BINARY w8a16 Qualcomm® QCS7181 301.833 ms 0 - 0 MB NPU
encoder VOICE_AI w8a16 Snapdragon® 8 Elite Gen 5 Mobile 176.799 ms 1 - 10 MB NPU
encoder VOICE_AI w8a16 Snapdragon® X2 Elite 155.785 ms 0 - 0 MB NPU
encoder VOICE_AI w8a16 Snapdragon® X Elite 303.584 ms 0 - 0 MB NPU
encoder VOICE_AI w8a16 Snapdragon® 8 Gen 3 Mobile 270.02 ms 1 - 7 MB NPU
encoder VOICE_AI w8a16 Qualcomm® QCS8275 519.085 ms 1 - 9 MB NPU
encoder VOICE_AI w8a16 Qualcomm® QCS8550 (Proxy) 367.279 ms 1 - 3 MB NPU
encoder VOICE_AI w8a16 Qualcomm® SA8775P 317.111 ms 0 - 9 MB NPU
encoder VOICE_AI w8a16 Qualcomm® SA8650P 317.111 ms 0 - 9 MB NPU
encoder VOICE_AI w8a16 Qualcomm® SA8255P 317.111 ms 0 - 9 MB NPU
encoder VOICE_AI w8a16 Qualcomm® QCS9075 297.079 ms 0 - 29 MB NPU
encoder VOICE_AI w8a16 Qualcomm® QCM6690 4382.54 ms 1 - 12 MB NPU
encoder VOICE_AI w8a16 Snapdragon® 8 Elite For Galaxy Mobile 226.25 ms 1 - 14 MB NPU
encoder VOICE_AI w8a16 Qualcomm® SA7255P 519.085 ms 1 - 9 MB NPU
encoder VOICE_AI w8a16 Snapdragon® 7 Gen 4 Mobile 472.954 ms 0 - 7 MB NPU
encoder VOICE_AI w8a16 Qualcomm® QCS7790 472.954 ms 0 - 7 MB NPU
encoder VOICE_AI w8a16 Qualcomm® QCS8750 226.25 ms 1 - 14 MB NPU
encoder VOICE_AI w8a16 Qualcomm® QCS7181 303.584 ms 0 - 0 MB NPU

License

  • The license for the original implementation of Whisper-Small-Quantized 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