anodyine/vllm-omni
1
0 Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity

[Performance] Improve MiMo-Audio tokenizer decoding performance (#2183)

Browse Source 
Signed-off-by: 齐保元 <qibaoyuan@xiaomi.com>
Co-authored-by: Jialong Liu <88185941+Galleons2029@users.noreply.github.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
This commit is contained in:
Baoyuan QiJialong LiuHongsheng Liu
 authored and GitHub committed 2026-05-11 10:05:46 +08:00
1 parent ef7f2f1cd2
commit e1088026fa
6 files changed
+775 -33
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

tests/model_executor/models/mimo_audio/test_mimo_audio_code2wav_batch_decode.py
+1 -1
View File
@@ -91,7 +91,7 @@ def _minimal_model(mocker: MockerFixture):
decoder=decoder,
config=SimpleNamespace(avg_pooler=2, stride_size=2, hop_length=240),
)
model._tokenizer_service = SimpleNamespace(audio_tokenizer=audio_tok)
model._tokenizer_service = SimpleNamespace(audio_tokenizer=audio_tok, cuda_graph_wrapper=None)
return model, audio_tok