Distil-Bert-Base-Uncased-Hf: Optimized for Qualcomm Devices

DistilBERT is a lightweight BERT model designed for efficient self-supervised learning of language representations. It can be used for masked language modeling and as a backbone for various NLP tasks.

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.

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
ONNX	w8a16	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a16	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit Distil-Bert-Base-Uncased-Hf 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 Distil-Bert-Base-Uncased-Hf on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.text_generation

Model Stats:

• Input resolution: 1x384
• Model checkpoint: distil_bert_base_uncased_hf
• Model size (float): 43.3 MB
• Number of parameters: 11.3M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Snapdragon® X2 Elite	3.722 ms	1 - 1 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Snapdragon® X Elite	8.764 ms	112 - 112 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	5.769 ms	0 - 370 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	14.084 ms	0 - 371 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	90.689 ms	60 - 63 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-8275	8.054 ms	0 - 4 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	8.408 ms	0 - 114 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® QCS8450	14.084 ms	0 - 371 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	8.409 ms	0 - 3 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	8.764 ms	112 - 112 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ Q-6690	174.542 ms	60 - 427 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ Q-7790	29.639 ms	60 - 386 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	4.31 ms	0 - 316 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Snapdragon® 8 Elite Mobile	4.31 ms	0 - 316 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	3.43 ms	0 - 325 MB	NPU
Distil-Bert-Base-Uncased-Hf	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	29.639 ms	60 - 386 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Snapdragon® X2 Elite	5.052 ms	1 - 1 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Snapdragon® X Elite	11.478 ms	1 - 1 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	7.522 ms	0 - 373 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	23.712 ms	0 - 402 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	14.554 ms	0 - 2 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	10.716 ms	0 - 2 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® SA8775P	15.288 ms	0 - 249 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® SA8650P	15.288 ms	0 - 249 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® SA8255P	15.288 ms	0 - 249 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® QCS8450	23.712 ms	0 - 402 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	15.498 ms	0 - 2 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	11.478 ms	1 - 1 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	5.732 ms	0 - 267 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® SA7255P	45.93 ms	0 - 247 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Qualcomm® SA8295P	17.848 ms	0 - 276 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Snapdragon® 8 Elite Mobile	5.732 ms	0 - 267 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	4.199 ms	0 - 237 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Snapdragon® X2 Elite	3.183 ms	1 - 1 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Snapdragon® X Elite	7.047 ms	1 - 1 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	4.622 ms	0 - 366 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	6.205 ms	0 - 2 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	6.482 ms	0 - 2 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® SA8775P	6.55 ms	0 - 308 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® SA8650P	6.55 ms	0 - 308 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® SA8255P	6.55 ms	0 - 308 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	6.383 ms	2 - 4 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	7.047 ms	1 - 1 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	3.665 ms	0 - 314 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Qualcomm® SA7255P	16.106 ms	0 - 308 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	3.665 ms	0 - 314 MB	NPU
Distil-Bert-Base-Uncased-Hf	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.535 ms	0 - 318 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	7.606 ms	0 - 377 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	23.81 ms	0 - 402 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	14.804 ms	0 - 177 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	10.413 ms	0 - 2 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® SA8775P	15.495 ms	0 - 255 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® SA8650P	15.495 ms	0 - 255 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® SA8255P	15.495 ms	0 - 255 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® QCS8450	23.81 ms	0 - 402 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	15.834 ms	0 - 176 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	6.001 ms	0 - 284 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® SA7255P	46.35 ms	0 - 255 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Qualcomm® SA8295P	17.854 ms	0 - 276 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Snapdragon® 8 Elite Mobile	6.001 ms	0 - 284 MB	NPU
Distil-Bert-Base-Uncased-Hf	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	4.467 ms	0 - 242 MB	NPU

License

• The license for the original implementation of Distil-Bert-Base-Uncased-Hf can be found here.

References

• DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter

Community

• Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
• For questions or feedback please reach out to us.