Zhou Zhao, Ziyue Jiang, Shengpeng Ji, Jialong Zuo, Hanting Wang
We lifted 12 functions out of this paper's own repositories and ran 10 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| enhuiz/vall-e | — | 10 of 12 |
| Function | Status | Where it lives |
|---|---|---|
| AdaLN | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("ec7cdfe06cf76a22") |
| Attention | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("8fe53745e4d5d22f") |
| Block | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("4d47845809fd8661") |
| Embedding | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("ccb069c6188deb3c") |
| MultiEmbedding | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("b1cc06b94bee5ada") |
| PrenormResidual | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("b8b6021719fa7eed") |
| SinusodialEmbedding | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("91e12b0e6dfa4060") |
| _create_mask | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("d59002e01a0b0518") |
| _join | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("f47e5f0f909741dd") |
| list_to_tensor | Ran | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("b06aa768d8e2dd43") |
| AR | Not yet run | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("894ecc1799b226be") |
| Base | Not yet run | enhuiz/vall-e/vall_e/vall_e/ar.py code served (permissive licence) · get_code("3c60927a0bd1494e") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Zero-shot text-to-speech (TTS) has gained significant attention due to its powerful voice cloning capabilities, requiring only a few seconds of unseen speaker voice prompts. However, all previous work has been developed for cloud-based systems. Taking autoregressive models as an example, although these approaches achieve high-fidelity voice cloning, they fall short in terms of inference speed, model size, and robustness. Therefore, we propose MobileSpeech, which is a fast, lightweight, and robust zero-shot text-tospeech system based on mobile devices for the first time. Specifically: 1) leveraging discrete codec, we design a parallel speech mask decoder module called SMD, which incorporates hierarchical information from the speech codec and weight mechanisms across different codec layers during the generation process. Moreover, to bridge the gap between text and speech, we introduce a high-level probabilistic mask that simulates the progression of information flow from less to more during speech generation. 2) For speaker prompts, we extract fine-grained prompt duration from the prompt speech and incorporate text, prompt speech by cross attention in SMD. We demonstrate the effectiveness of MobileSpeech on multilingual datasets at different levels, achieving state-ofthe-art results in terms of generating speed and speech quality. MobileSpeech achieves RTF of 0.09 on a single A100 GPU and we have successfully deployed MobileSpeech on mobile devices. Audio samples are available at https://mobilespeech.github.io/ .
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2402.09378")
get_code_for_paper("2402.09378")
have("2402.09378")
Connect an agent — have() is free.