Nomic-Embed-Text: Optimized for Qualcomm Devices
A text encoder that surpasses OpenAI text-embedding-ada-002 and text-embedding-3-small performance on short and long context tasks.
This is based on the implementation of Nomic-Embed-Text 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.
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 | float | Universal | QAIRT 2.45, ONNX Runtime 1.27.1 | Download |
| QNN_DLC | float | Universal | QAIRT 2.45 | Download |
| TFLITE | float | Universal | QAIRT 2.45 | Download |
For more device-specific assets and performance metrics, visit Nomic-Embed-Text 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 Nomic-Embed-Text on GitHub for usage instructions.
Model Details
Model Type: Model_use_case.text_generation
Model Stats:
- Model checkpoint: v1.5
- Input resolution: 1x128 (seqlen can vary)
- Number of parameters: 137M
- Model size (float): 523 MB
Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit |
|---|---|---|---|---|---|---|
| Nomic-Embed-Text | ONNX | float | Snapdragon® X2 Elite | 3.456 ms | 1 - 1 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Snapdragon® X Elite | 8.242 ms | 264 - 264 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 5.668 ms | 0 - 396 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 11.474 ms | 0 - 375 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 7.973 ms | 0 - 323 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Qualcomm® QCS8450 | 11.474 ms | 0 - 375 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 12.45 ms | 0 - 3 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 8.242 ms | 264 - 264 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 4.304 ms | 0 - 186 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Snapdragon® 8 Elite Mobile | 4.304 ms | 0 - 186 MB | NPU |
| Nomic-Embed-Text | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.883 ms | 0 - 195 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Snapdragon® X2 Elite | 3.842 ms | 1 - 1 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Snapdragon® X Elite | 8.474 ms | 0 - 0 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 5.607 ms | 0 - 422 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 11.798 ms | 0 - 372 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8275 | 29.038 ms | 0 - 198 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 7.944 ms | 0 - 2 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® SA8775P | 10.18 ms | 0 - 198 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® SA8650P | 10.18 ms | 0 - 198 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® SA8255P | 10.18 ms | 0 - 198 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® QCS8450 | 11.798 ms | 0 - 372 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 9.789 ms | 0 - 2 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 8.474 ms | 0 - 0 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 4.333 ms | 0 - 182 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® SA7255P | 29.038 ms | 0 - 198 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Qualcomm® SA8295P | 11.473 ms | 0 - 177 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 4.333 ms | 0 - 182 MB | NPU |
| Nomic-Embed-Text | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.908 ms | 0 - 188 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 5.607 ms | 0 - 400 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 11.624 ms | 0 - 372 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® Dragonwing™ QCS8275 | 29.035 ms | 0 - 202 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 7.757 ms | 0 - 2 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® SA8775P | 10.186 ms | 0 - 203 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® SA8650P | 10.186 ms | 0 - 203 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® SA8255P | 10.186 ms | 0 - 203 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® QCS8450 | 11.624 ms | 0 - 372 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 12.559 ms | 0 - 265 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 4.351 ms | 0 - 189 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® SA7255P | 29.035 ms | 0 - 202 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Qualcomm® SA8295P | 11.586 ms | 0 - 178 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Snapdragon® 8 Elite Mobile | 4.351 ms | 0 - 189 MB | NPU |
| Nomic-Embed-Text | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.912 ms | 0 - 196 MB | NPU |
License
- The license for the original implementation of Nomic-Embed-Text can be found here.
References
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.