游雁 4967c8bd8e feat: add Qwen3-ASR speaker diarization; remove unnecessary punc_model from SPK examples ...

- Qwen3-ASR: support spk_model with forced_aligner for speaker diarization
- Qwen3-ASR: fix list input handling (numpy/tensor to tuple format)
- Qwen3-ASR: output timestamp as [start_ms, end_ms] for pipeline compatibility
- Qwen3-ASR: graceful fallback when forced_aligner not configured
- Remove punc_model from Fun-ASR-Nano/SenseVoice/Qwen3-ASR SPK demos and tests
  (these models output punctuation natively)
- Fix test scripts to handle both text/sentence keys in sentence_info
- auto_model: set return_time_stamps=True when spk_model present

2026-05-21 04:38:21 +00:00

..

auto

feat: add Qwen3-ASR speaker diarization; remove unnecessary punc_model from SPK examples

2026-05-21 04:38:21 +00:00

bin

fix: fix undefined variable in FSDP auto_wrap_policy (train.py)

2026-05-19 17:52:57 +00:00

datasets

fix: resolve multiple issues (#2689, #2720, #2755, #2780)

2026-05-19 14:14:10 +00:00

download

bugfix

2025-12-16 00:50:07 +08:00

frontends

fix: replace mutable default argument cache: dict = {} with None

2026-05-19 17:56:00 +00:00

losses

Dev gzf exp (#1654)

2024-04-24 16:03:38 +08:00

metrics

fixbug

2024-06-24 10:20:05 +08:00

models

feat: add Qwen3-ASR speaker diarization; remove unnecessary punc_model from SPK examples

2026-05-21 04:38:21 +00:00

optimizers

Dev gzf exp (#1654)

2024-04-24 16:03:38 +08:00

schedulers

Dev gzf exp (#1678)

2024-04-29 14:52:20 +08:00

tokenizer

Fix HuggingfaceTokenizer import error handling (#2832)

2026-03-17 20:14:01 +08:00

train_utils

fix: raise NotImplementedError for FSDP checkpoint saving

2026-05-19 17:53:53 +00:00

utils

[Fix] fix is_sublist function (#2826)

2026-03-12 09:45:32 +08:00

__init__.py

Improve auto-registration import diagnostics

2026-05-20 12:51:56 +00:00

register.py

v1.1.18

2024-12-12 11:37:23 +08:00

version.txt

1.3.1

2026-01-26 21:01:13 +08:00