Whisper‑Small‑En is the English-only variant of OpenAI’s Whisper small model, featuring approximately 244 million parameters and optimized for English speech recognition via the small.en checkpoint. Trained on 680 k hours of labeled audio (including English), it accepts up to 30-second audio segments and demonstrates robust performance for English-speaking tasks.
- Model Statistics:
- Encoder: 102M parameters (~390 MB)
- Decoder: 139M parameters (~531 MB)
- Inference Performance: ~2.17 inferences/sec, throughput 110–141 MB, latency around 500–700 ms depending on hardware
- Use Cases: Suitable for real-time captioning, voice assistants, meeting transcription, and other English ASR tasks on mid-level devices (e.g., smartphones, AI chips)
The source model can be found here
When the user has fine-tuned the source model, the model conversion process must be performed again.
Users can refer to either of the following two methods to complete the model conversion:
Using AIMO for model conversion: Click Model Conversion Reference in the Performance Reference section on the right to view the conversion steps.
Using Qualcomm QNN for model conversion: Please refer to the Qualcomm QNN Documentation.
The model performance benchmarks and example code provided by Model Farm are all implemented based on the APLUX AidLite SDK.
For models in .bin format, you can use either of the following two inference engines to run inference on Qualcomm chips:
Inference using APLUX AidLite: please refer to the APLUX AidLite Developer Documentation
Inference using Qualcomm QNN: Please refer to the Qualcomm QNN Documentation
Inference Example Code
The inference example code is implemented using the AidLite SDK.
Click Model & Code to download the model files and the inference code package. The file structure is as follows:
/model_farm_{model_name}_aidlite
|__ models # folder where model files are stored
|__ python # aidlite python model inference example
|__ cpp # aidlite cpp model inference example
|__ README.md