library_name: pytorch
license: other
tags:
- real_time
- android
pipeline_tag: text-to-audio
MeloTTS-ES: Optimized for Qualcomm Devices
MeloTTS is a high-quality multi-lingual text-to-speech library for English, Chinese and Spanish language.
This is based on the implementation of MeloTTS-ES 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 |
|---|---|---|---|---|
| VOICE_AI | mixed_with_float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Snapdragon® X2 Elite | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Snapdragon® X Elite | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | qualcomm-qcs8275 | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Qualcomm® QCS8550 (Proxy) | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Qualcomm® SA8775P | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Qualcomm® SA7255P | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Qualcomm® SA8295P | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Qualcomm® QCS9075 | QAIRT 2.45 | Download |
| VOICE_AI | mixed_with_float | Qualcomm® QCS8450 (Proxy) | QAIRT 2.45 | Download |
For more device-specific assets and performance metrics, visit MeloTTS-ES 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 MeloTTS-ES on GitHub for usage instructions.
Model Details
Model Type: Model_use_case.audio_generation
Model Stats:
- Model checkpoint: myshell-ai/MeloTTS-Spanish
- Max decoded sequence length: 512 tokens
- Number of parameters (encoder): 8.36M
- Model size (encoder) (float): 32.0 MB
- Number of parameters (flow): 20.1M
- Model size (flow) (float): 76.9 MB
- Number of parameters (decoder): 14.5M
- Model size (decoder) (float): 55.5 MB
- Number of parameters (t5_encoder): 15.1M
- Model size (t5_encoder) (float): 57.5 MB
- Number of parameters (t5_decoder): 5.72M
- Model size (t5_decoder) (float): 21.8 MB
Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit |
|---|---|---|---|---|---|---|
| decoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite Gen 5 Mobile | 43.603 ms | 0 - 10 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Snapdragon® X2 Elite | 40.761 ms | 0 - 0 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Snapdragon® X Elite | 82.709 ms | 0 - 0 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Gen 3 Mobile | 61.145 ms | 0 - 8 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8275 (Proxy) | 134.528 ms | 0 - 9 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8550 (Proxy) | 85.581 ms | 0 - 2 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Qualcomm® SA8775P | 83.845 ms | 0 - 9 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS9075 | 83.281 ms | 0 - 2 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8450 (Proxy) | 113.955 ms | 1 - 9 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Qualcomm® SA7255P | 134.528 ms | 0 - 9 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Qualcomm® SA8295P | 100.024 ms | 0 - 6 MB | NPU |
| decoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite For Galaxy Mobile | 47.962 ms | 0 - 9 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite Gen 5 Mobile | 19.517 ms | 4 - 13 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Snapdragon® X2 Elite | 20.57 ms | 4 - 4 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Snapdragon® X Elite | 32.975 ms | 4 - 4 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Gen 3 Mobile | 25.204 ms | 4 - 11 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8275 (Proxy) | 52.553 ms | 2 - 10 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8550 (Proxy) | 34.626 ms | 4 - 5 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Qualcomm® SA8775P | 36.213 ms | 2 - 11 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS9075 | 35.679 ms | 4 - 9 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8450 (Proxy) | 44.062 ms | 4 - 13 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Qualcomm® SA7255P | 52.553 ms | 2 - 10 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Qualcomm® SA8295P | 40.778 ms | 0 - 5 MB | NPU |
| encoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite For Galaxy Mobile | 20.302 ms | 2 - 15 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite Gen 5 Mobile | 64.613 ms | 2 - 11 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Snapdragon® X2 Elite | 61.925 ms | 2 - 2 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Snapdragon® X Elite | 122.177 ms | 2 - 2 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Snapdragon® 8 Gen 3 Mobile | 90.362 ms | 2 - 10 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Qualcomm® QCS8275 (Proxy) | 235.037 ms | 2 - 11 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Qualcomm® QCS8550 (Proxy) | 123.207 ms | 3 - 5 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Qualcomm® SA8775P | 121.083 ms | 2 - 11 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Qualcomm® QCS9075 | 120.456 ms | 4 - 8 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Qualcomm® QCS8450 (Proxy) | 205.615 ms | 2 - 12 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Qualcomm® SA7255P | 235.037 ms | 2 - 11 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Qualcomm® SA8295P | 150.571 ms | 0 - 5 MB | NPU |
| flow | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite For Galaxy Mobile | 79.131 ms | 2 - 11 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite Gen 5 Mobile | 0.256 ms | 0 - 9 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Snapdragon® X2 Elite | 0.328 ms | 1 - 1 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Snapdragon® X Elite | 0.429 ms | 1 - 1 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Gen 3 Mobile | 0.305 ms | 0 - 8 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8275 (Proxy) | 0.983 ms | 0 - 9 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8550 (Proxy) | 0.401 ms | 1 - 2 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Qualcomm® SA8775P | 0.667 ms | 0 - 10 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS9075 | 0.512 ms | 1 - 3 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8450 (Proxy) | 0.578 ms | 1 - 10 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Qualcomm® SA7255P | 0.983 ms | 0 - 9 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Qualcomm® SA8295P | 0.796 ms | 0 - 5 MB | NPU |
| t5_decoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite For Galaxy Mobile | 0.271 ms | 0 - 9 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite Gen 5 Mobile | 0.482 ms | 0 - 10 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Snapdragon® X2 Elite | 0.654 ms | 0 - 0 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Snapdragon® X Elite | 1.052 ms | 0 - 0 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Gen 3 Mobile | 0.636 ms | 0 - 7 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8275 (Proxy) | 2.839 ms | 0 - 9 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8550 (Proxy) | 0.877 ms | 0 - 1 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Qualcomm® SA8775P | 1.275 ms | 0 - 9 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS9075 | 1.118 ms | 0 - 2 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Qualcomm® QCS8450 (Proxy) | 1.358 ms | 0 - 9 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Qualcomm® SA7255P | 2.839 ms | 0 - 9 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Qualcomm® SA8295P | 1.721 ms | 0 - 5 MB | NPU |
| t5_encoder | VOICE_AI | mixed_with_float | Snapdragon® 8 Elite For Galaxy Mobile | 0.523 ms | 0 - 9 MB | NPU |
License
- The license for the original implementation of MeloTTS-ES 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.
