0 Fork 0

mirror of https://github.com/k2-fsa/sherpa-onnx.git synced 2026-11-10 21:11:03 +08:00

Files

T

Hassan Kazi c65f321699 Fix decoder positions of generated tokens in Canary greedy search (#4097 )

The prompt is fed at positions 0..n-1, but generated tokens were fed at
1, 2, ... instead of n, n+1, ..., overlapping the prompt. Feed them at
decoder_input.size() + i - 1, as NeMo's GreedySequenceGenerator does.

Fixes #4096

2026-10-24 16:50:59 +08:00

canary

Fix decoder positions of generated tokens in Canary greedy search (#4097 )

2026-10-24 16:50:59 +08:00

fast-conformer-hybrid-transducer-ctc

Support Portuguese and German ASR models from NeMo (#2394 )

2025-07-14 15:00:18 +08:00

GigaAM

Fix typos in URL (#2905 )

2025-12-17 14:24:01 +08:00

nemotron-3.5-asr-streaming-0.6b

Export nemotron-3.5-asr-streaming-0.6b to QNN (#3741 )

2026-07-09 18:01:21 +08:00

nemotron-speech-streaming-en-0.6b

Update nemotron-speech-streaming-en-0.6b (#3555 )

2026-04-27 10:21:08 +08:00

parakeet-tdt_ctc-0.6b-ja

Refactor exporting NeMo models (#2362 )

2025-07-09 16:02:12 +08:00

parakeet-tdt-0.6b-v2

Add modified beam search and hotwords support for NeMo transducer models (#3077 )

2026-02-05 12:49:27 +08:00

parakeet-tdt-0.6b-v3

Add modified beam search and hotwords support for NeMo transducer models (#3077 )

2026-02-05 12:49:27 +08:00

parakeet-unified-en-0.6b

Export NeMo unified streaming 0.6b models (#3602 )

2026-05-12 11:57:05 +08:00

qnn

fix(scripts/nemo/qnn/parakeet-ctc): bind feat_dim before use (#3937 )

2026-09-09 12:33:19 +08:00

speaker-verification

Upload speaker embedding models to huggingface (#1428 )

2024-10-14 16:20:00 +08:00

.gitignore

Export non-streaming NeMo faster conformer hybrid transducer and ctc to sherpa-onnx (#847 )

2024-05-09 13:59:47 +08:00

generate_bpe_vocab.py

Add modified beam search and hotwords support for NeMo transducer models (#3077 )

2026-02-05 12:49:27 +08:00

README.md

Export speaker verification models from NeMo to ONNX (#526 )

2024-01-13 19:49:45 +08:00

README.md

Introduction

This directory contains scripts for exporting models from NeMo to onnx so that you can use them in sherpa-onnx.

./speaker-verification contains models for speaker verification.