phonon:40m-fr

1,605 yesterday

Phonon is Gradium’s text-to-speech family. It turns written text into spoken audio as it generates, making it useful for reading assistant replies, narrating passages, and adding a voice to local apps. 

ollama run phonon:40m-fr

Details

yesterday

da29147f2cba · 123MB

json
{ "languages": [ "fr" ], "embeddings_ext": [ "ba7a5a45@300" ], "flow_lm": { "d_model": 512, "num_hea
json
{ "do_sample": true, "eos_threshold": 4.0, "min_frames_before_eos": 0, "num_flow_steps": 1, "pad_tok
json
{ "audio_tokenizer": { "audio_tokenizer_class": "PocketTTSAudioTokenizerModel", "audio_tokenizer_nam
json
{ "version": "1.0", "truncation": null, "padding": null, "added_tokens": [ { "id": 0, "content": "\u
json
{ "version": "1.0", "truncation": null, "padding": null, "added_tokens": [ { "id": 0, "content": "\u
json
{ "tokenizer_class": "TokenizersBackend", "fast_tokenizer_files": [ "tokenizer.5.0.0.json" ], "unk_t
license
Gradium gradium/phonon-40m-fr Source: https://huggingface.co/gradium/phonon-40m-fr/tree/87f56c5c66cf
210 tensors

Readme

Phonon is Gradium’s text-to-speech family. It turns written text into spoken audio as it generates, making it useful for reading assistant replies, narrating passages, and adding a voice to local apps.

ollama run phonon "Good morning. Here is a short update before your first meeting."

The default model is Phonon 31M English. It includes one voice, named default. Choose a 40M model for another language or for the multilingual checkpoint:

Ollama tag Language Loaded memory on Apple M5 Max
phonon (phonon:31m-en, phonon:en) English, 31M 121 MB
phonon:40m-en English, 40M 141 MB
phonon:40m-fr (phonon:fr) French, 40M 141 MB
phonon:40m-de (phonon:de) German, 40M 141 MB
phonon:40m-es (phonon:es) Spanish, 40M 141 MB
phonon:40m-pt (phonon:pt) Portuguese, 40M 141 MB
phonon:40m-multilingual (phonon:multilingual) All five languages, 40M 141 MB

The 31M English checkpoint is released under Apache 2.0. The 40M English, French, German, Spanish, Portuguese, and multilingual checkpoints are released under CC BY-NC 4.0, which limits use to noncommercial purposes.

Phonon produces 24 kHz mono audio. These Ollama packages require Apple Silicon and Ollama 0.40.9 or later. Give the model text intended to be spoken; formatting marks such as Markdown headings may be read aloud. For saving audio or using the speech API, see the Ollama speech guide.