Hugging Face's logo

qualcomm
/
Whisper-Small

Automatic Speech Recognition
PyTorch
foundation
android
Model card Files
xet
Community
Whisper-Small / README.md
qaihm-bot's picture
qaihm-bot
v0.43.0
70b5b2e verified
preview code
|
raw
history blame contribute delete
13.1 kB
metadata
library_name: pytorch
license: other
tags:
  - foundation
  - android
pipeline_tag: automatic-speech-recognition

Whisper-Small: Optimized for Mobile Deployment

Transformer-based automatic speech recognition (ASR) model for multilingual transcription and translation available on HuggingFace

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 model is an implementation of Whisper-Small found here.

This repository provides scripts to run Whisper-Small on Qualcomm® devices. More details on model performance across various devices, can be found here.

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
    • Number of parameters (HfWhisperEncoder): 102M
    • Model size (HfWhisperEncoder) (float): 391 MB
    • Number of parameters (HfWhisperDecoder): 139M
    • Model size (HfWhisperDecoder) (float): 533 MB
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
HfWhisperEncoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_CONTEXT_BINARY 427.348 ms 0 - 10 MB NPU Use Export Script
HfWhisperEncoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_CONTEXT_BINARY 278.013 ms 0 - 19 MB NPU Use Export Script
HfWhisperEncoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_CONTEXT_BINARY 119.313 ms 0 - 3 MB NPU Use Export Script
HfWhisperEncoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) PRECOMPILED_QNN_ONNX 136.472 ms 0 - 258 MB NPU Use Export Script
HfWhisperEncoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_CONTEXT_BINARY 139.056 ms 1 - 11 MB NPU Use Export Script
HfWhisperEncoder float SA7255P ADP Qualcomm® SA7255P QNN_CONTEXT_BINARY 427.348 ms 0 - 10 MB NPU Use Export Script
HfWhisperEncoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_CONTEXT_BINARY 119.219 ms 1 - 3 MB NPU Use Export Script
HfWhisperEncoder float SA8295P ADP Qualcomm® SA8295P QNN_CONTEXT_BINARY 248.829 ms 0 - 15 MB NPU Use Export Script
HfWhisperEncoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_CONTEXT_BINARY 119.885 ms 0 - 3 MB NPU Use Export Script
HfWhisperEncoder float SA8775P ADP Qualcomm® SA8775P QNN_CONTEXT_BINARY 139.056 ms 1 - 11 MB NPU Use Export Script
HfWhisperEncoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_CONTEXT_BINARY 88.332 ms 0 - 20 MB NPU Use Export Script
HfWhisperEncoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile PRECOMPILED_QNN_ONNX 104.703 ms 131 - 150 MB NPU Use Export Script
HfWhisperEncoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_CONTEXT_BINARY 62.133 ms 0 - 14 MB NPU Use Export Script
HfWhisperEncoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile PRECOMPILED_QNN_ONNX 79.937 ms 130 - 145 MB NPU Use Export Script
HfWhisperEncoder float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_CONTEXT_BINARY 47.339 ms 0 - 11 MB NPU Use Export Script
HfWhisperEncoder float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile PRECOMPILED_QNN_ONNX 63.036 ms 110 - 120 MB NPU Use Export Script
HfWhisperEncoder float Snapdragon X Elite CRD Snapdragon® X Elite QNN_CONTEXT_BINARY 119.879 ms 0 - 0 MB NPU Use Export Script
HfWhisperEncoder float Snapdragon X Elite CRD Snapdragon® X Elite PRECOMPILED_QNN_ONNX 130.558 ms 225 - 225 MB NPU Use Export Script
HfWhisperDecoder float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_CONTEXT_BINARY 18.72 ms 52 - 61 MB NPU Use Export Script
HfWhisperDecoder float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_CONTEXT_BINARY 18.264 ms 56 - 74 MB NPU Use Export Script
HfWhisperDecoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_CONTEXT_BINARY 11.746 ms 60 - 62 MB NPU Use Export Script
HfWhisperDecoder float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) PRECOMPILED_QNN_ONNX 12.662 ms 0 - 319 MB NPU Use Export Script
HfWhisperDecoder float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_CONTEXT_BINARY 13.112 ms 60 - 70 MB NPU Use Export Script
HfWhisperDecoder float SA7255P ADP Qualcomm® SA7255P QNN_CONTEXT_BINARY 18.72 ms 52 - 61 MB NPU Use Export Script
HfWhisperDecoder float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_CONTEXT_BINARY 11.908 ms 57 - 59 MB NPU Use Export Script
HfWhisperDecoder float SA8295P ADP Qualcomm® SA8295P QNN_CONTEXT_BINARY 14.753 ms 43 - 58 MB NPU Use Export Script
HfWhisperDecoder float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_CONTEXT_BINARY 11.919 ms 56 - 59 MB NPU Use Export Script
HfWhisperDecoder float SA8775P ADP Qualcomm® SA8775P QNN_CONTEXT_BINARY 13.112 ms 60 - 70 MB NPU Use Export Script
HfWhisperDecoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_CONTEXT_BINARY 9.458 ms 55 - 74 MB NPU Use Export Script
HfWhisperDecoder float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile PRECOMPILED_QNN_ONNX 9.942 ms 75 - 94 MB NPU Use Export Script
HfWhisperDecoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_CONTEXT_BINARY 8.151 ms 12 - 27 MB NPU Use Export Script
HfWhisperDecoder float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile PRECOMPILED_QNN_ONNX 8.554 ms 19 - 30 MB NPU Use Export Script
HfWhisperDecoder float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_CONTEXT_BINARY 7.273 ms 60 - 71 MB NPU Use Export Script
HfWhisperDecoder float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile PRECOMPILED_QNN_ONNX 7.545 ms 75 - 89 MB NPU Use Export Script
HfWhisperDecoder float Snapdragon X Elite CRD Snapdragon® X Elite QNN_CONTEXT_BINARY 9.741 ms 60 - 60 MB NPU Use Export Script
HfWhisperDecoder float Snapdragon X Elite CRD Snapdragon® X Elite PRECOMPILED_QNN_ONNX 10.386 ms 286 - 286 MB NPU Use Export Script

Installation

Install the package via pip:

# NOTE: 3.10 <= PYTHON_VERSION < 3.14 is supported.
pip install "qai-hub-models[whisper-small]"

Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device

Sign-in to Qualcomm® AI Hub Workbench with your Qualcomm® ID. Once signed in navigate to Account -> Settings -> API Token.

With this API token, you can configure your client to run models on the cloud hosted devices.

qai-hub configure --api_token API_TOKEN

Navigate to docs for more information.

Demo off target

The package contains a simple end-to-end demo that downloads pre-trained weights and runs this model on a sample input.

python -m qai_hub_models.models.whisper_small.demo

The above demo runs a reference implementation of pre-processing, model inference, and post processing.

NOTE: If you want running in a Jupyter Notebook or Google Colab like environment, please add the following to your cell (instead of the above).

%run -m qai_hub_models.models.whisper_small.demo

Run model on a cloud-hosted device

In addition to the demo, you can also run the model on a cloud-hosted Qualcomm® device. This script does the following:

  • Performance check on-device on a cloud-hosted device
  • Downloads compiled assets that can be deployed on-device for Android.
  • Accuracy check between PyTorch and on-device outputs.
python -m qai_hub_models.models.whisper_small.export

How does this work?

This export script leverages Qualcomm® AI Hub to optimize, validate, and deploy this model on-device. Lets go through each step below in detail:

Step 1: Compile model for on-device deployment

To compile a PyTorch model for on-device deployment, we first trace the model in memory using the jit.trace and then call the submit_compile_job API.

import torch

import qai_hub as hub
from qai_hub_models.models.whisper_small import Model

# Load the model
torch_model = Model.from_pretrained()

# Device
device = hub.Device("Samsung Galaxy S25")

# Trace model
input_shape = torch_model.get_input_spec()
sample_inputs = torch_model.sample_inputs()

pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])

# Compile model on a specific device
compile_job = hub.submit_compile_job(
    model=pt_model,
    device=device,
    input_specs=torch_model.get_input_spec(),
)

# Get target model to run on-device
target_model = compile_job.get_target_model()

Step 2: Performance profiling on cloud-hosted device

After compiling models from step 1. Models can be profiled model on-device using the target_model. Note that this scripts runs the model on a device automatically provisioned in the cloud. Once the job is submitted, you can navigate to a provided job URL to view a variety of on-device performance metrics.

profile_job = hub.submit_profile_job(
    model=target_model,
    device=device,
)

Step 3: Verify on-device accuracy

To verify the accuracy of the model on-device, you can run on-device inference on sample input data on the same cloud hosted device.

input_data = torch_model.sample_inputs()
inference_job = hub.submit_inference_job(
    model=target_model,
    device=device,
    inputs=input_data,
)
    on_device_output = inference_job.download_output_data()

With the output of the model, you can compute like PSNR, relative errors or spot check the output with expected output.

Note: This on-device profiling and inference requires access to Qualcomm® AI Hub Workbench. Sign up for access.

Deploying compiled model to Android

The models can be deployed using multiple runtimes:

  • TensorFlow Lite (.tflite export): This tutorial provides a guide to deploy the .tflite model in an Android application.

  • QNN (.so export ): This sample app provides instructions on how to use the .so shared library in an Android application.

View on Qualcomm® AI Hub

Get more details on Whisper-Small's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

  • The license for the original implementation of Whisper-Small can be found here.
  • The license for the compiled assets for on-device deployment can be found here

References

Community