Search results

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

Status: Open.

#25 In PolyAI-LDN/pheme;

· NeuroPeakX opened on Oct 16, 2025

Conversational style outputs

Status: Open.

#24 In PolyAI-LDN/pheme;

· jonhilgart22 opened on Oct 15, 2024

Does it support Chinese?

Status: Open.

#22 In PolyAI-LDN/pheme;

· wjddd opened on Jun 28, 2024

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

Status: Open.

#21 In PolyAI-LDN/pheme;

· Moonmore opened on May 6, 2024

Possible to infer word by word?

Status: Open.

#20 In PolyAI-LDN/pheme;

· rumbleFTW opened on Apr 5, 2024

hallucination in the T2S stage.

Status: Open.

#19 In PolyAI-LDN/pheme;

· sos1sos2Sixteen opened on Mar 28, 2024

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

Status: Open.

#17 In PolyAI-LDN/pheme;

· Bambuuai opened on Feb 21, 2024

Only <eos> generated after training new model.

Status: Open.

#16 In PolyAI-LDN/pheme;

· kip6000 opened on Feb 19, 2024

When are you going to release tensorrt-llm image?

Status: Open.

#15 In PolyAI-LDN/pheme;

· Bambuuai opened on Feb 15, 2024

Does it support streaming output?

Status: Open.

#14 In PolyAI-LDN/pheme;

· Bambuuai opened on Feb 15, 2024

unique_text_tokens.k2symbols for non-english languages

Status: Open.

#13 In PolyAI-LDN/pheme;

· paulovasconcellos-hotmart opened on Feb 9, 2024

Segmentation fault when running python transformer_infer.py

Status: Open.

#12 In PolyAI-LDN/pheme;

· CircleCly opened on Feb 5, 2024