Search results

How to match onnx and pth?

Status: Open.

#39 In pipecat-ai/smart-turn;

· cenwurong opened on May 9, 2026

Training information

Status: Open.

#38 In pipecat-ai/smart-turn;

· Benjoyo opened on Mar 28, 2026

Does your model support real-time semantic end-of-speech detection for audio?

Status: Open.

#34 In pipecat-ai/smart-turn;

· lzl-mt opened on Dec 8, 2025

Quantized ONNX model fails on WebGPU due to invalid int32 DequantizeLinear (bias zero_point)

Status: Open.

#33 In pipecat-ai/smart-turn;

· urim-thummim opened on Nov 28, 2025

Higher confidence scores for turn_completion only after 1200ms of silence at end of audio sample.

Status: Open.

#32 In pipecat-ai/smart-turn;

· atharvabagde opened on Nov 19, 2025

Which format is appropriate for the input of turn prediction

Status: Open.

#25 In pipecat-ai/smart-turn;

· yoimiya122 opened on Jul 23, 2025

Add new language

Status: Open.

#21 In pipecat-ai/smart-turn;

· karimi12 opened on Jul 19, 2025

Turn detection ends prematurely on short Hindi utterances like "theek hai", "haan", "hmm"

Status: Open.

#18 In pipecat-ai/smart-turn;

· Anudeeprao1 opened on Jul 16, 2025

Could you kindly provide a pipeline for generating synthetic training data?

Status: Open.

#17 In pipecat-ai/smart-turn;

· mumuyeye opened on May 11, 2025

Silenced audio gives different results

Status: Open.

#12 In pipecat-ai/smart-turn;

· rootux opened on Mar 23, 2025

Whispering does not work well

Status: Open.

#11 In pipecat-ai/smart-turn;

· rootux opened on Mar 22, 2025

Benchmarks and smol LLM model

Status: Open.

#5 In pipecat-ai/smart-turn;

· bevenky opened on Mar 7, 2025