Search Results for "human"

...Specifically, the base model (MetaVoice-1B) uses around 1.2 billion parameters and has been trained on a massive dataset — reportedly around 100,000 hours of speech data. The goal is to provide human-like, expressive, and flexible TTS: able to generate natural-sounding speech that can handle diverse inputs and likely generalize over voice styles, intonation, prosody, and perhaps multiple languages or accents. With that scale and dataset volume, MetaVoice aims to push the boundary of what open-source TTS models can achieve: high fidelity, natural prosody, and robustness even for edge cases. ...

Orpheus TTS is a state-of-the-art open-source text-to-speech system built on a Llama-3B backbone, treating speech synthesis as a large language model problem instead of a traditional TTS pipeline. It is designed to produce human-like speech with natural intonation, emotion, and rhythm, targeting quality comparable to or better than many closed-source systems. The project ships both pretrained and finetuned English models, as well as a family of multilingual models released as a research preview, and includes data-processing scripts so users can train or finetune their own variants. ...

...Built from the ground up for scalability, composability, and research flexibility, fairseq2 supports a broad range of language, speech, and multimodal content generation tasks, including instruction fine-tuning, reinforcement learning from human feedback (RLHF), and large-scale multilingual modeling. Unlike the original fairseq—which evolved into a large, monolithic codebase—fairseq2 introduces a clean, plugin-oriented architecture designed for long-term maintainability and rapid experimentation. It supports multi-GPU and multi-node distributed training using DDP, FSDP, and tensor parallelism, capable of scaling up to 70B+ parameter models. ...

GLM-4-Voice is an open-source speech-enabled model from ZhipuAI, extending the GLM-4 family into the audio domain. It integrates advanced voice recognition and generation with the multimodal reasoning capabilities of GLM-4, enabling smooth natural interaction via spoken input and output. The model supports real-time speech-to-text transcription, spoken dialogue understanding, and text-to-speech synthesis, making it suitable for conversational AI, virtual assistants, and accessibility...

...It can also produce nonverbal vocalizations like laughter, coughs, clearing the throat, and similar sounds, which are crucial for making synthetic conversations feel human. Dia is released with pretrained checkpoints and inference code, with weights hosted on Hugging Face, so researchers and developers can quickly try it or integrate it into pipelines. The base model currently targets English and has around 1.6 billion parameters, offering a strong balance between realism and computational cost, while the ecosystem also includes Dia2.

StyleTTS2 is a state-of-the-art text-to-speech system that aims for human-level naturalness by combining style diffusion, adversarial training, and large speech language models. It extends the original StyleTTS idea by introducing a style diffusion model that can sample rich, realistic speaking styles conditioned on reference speech, allowing highly expressive and diverse prosody. The architecture uses a two-stage training process and leverages an auxiliary speech language model to guide generation toward more natural and coherent utterances. ...