Issues

How do you quantify “10x fewer data” — is this based on total hours of audio, number of unique speakers, or linguistic diversity?

PolyAI-LDN/pheme#25 · opened

on Oct 16, 2025

Conversational style outputs

PolyAI-LDN/pheme#24 · opened

on Oct 15, 2024

Does it support Chinese?

PolyAI-LDN/pheme#22 · opened

on Jun 28, 2024

What is the minimum level of graphics card required for model training?

PolyAI-LDN/pheme#21 · opened

on May 6, 2024

Possible to infer word by word?

PolyAI-LDN/pheme#20 · opened

on Apr 5, 2024

hallucination in the T2S stage.

PolyAI-LDN/pheme#19 · opened

on Mar 28, 2024

1

There is an issue with the pronunciation of numbers and symbols.

PolyAI-LDN/pheme#17 · opened

on Feb 21, 2024

Only <eos> generated after training new model.

PolyAI-LDN/pheme#16 · opened

on Feb 19, 2024

When are you going to release tensorrt-llm image?

PolyAI-LDN/pheme#15 · opened

on Feb 15, 2024

1

Does it support streaming output?

PolyAI-LDN/pheme#14 · opened

on Feb 15, 2024

1

unique_text_tokens.k2symbols for non-english languages

PolyAI-LDN/pheme#13 · opened

on Feb 9, 2024

1

Segmentation fault when running `python transformer_infer.py`

PolyAI-LDN/pheme#12 · opened

on Feb 5, 2024

How to train model in other languages?

PolyAI-LDN/pheme#8 · opened

on Jan 14, 2024

3

Replicate information about the model weights on the 🤗 Hub!

PolyAI-LDN/pheme#7 · opened

on Jan 14, 2024

1

Why semantic_tokens generation is so incredibly slow? :(

PolyAI-LDN/pheme#6 · opened

on Jan 13, 2024

3

Alternative to Soundstorm (S2A) model

PolyAI-LDN/pheme#5 · opened

on Jan 12, 2024

Demo Colab Notebook for Inference

PolyAI-LDN/pheme#4 · opened