Search results

Common Voice 15 EN Evaluation

Status: Open.

#27 In OpenMOSS/MOSS-Audio;

· PREPONDERANCE opened on Jun 19, 2026

sft lora finetune的loss和gradient norm的数值过于大，sft训练后性能反而显著下降

Status: Open.

#26 In OpenMOSS/MOSS-Audio;

· WeitengHu opened on Jun 11, 2026

Inference可以支持Batch处理吗？以及finetune.py是否可以支持自定义system prompt

Status: Open.

#25 In OpenMOSS/MOSS-Audio;

· WeitengHu opened on Jun 11, 2026

Add the LICENSE file — pyproject references one that doesn't exist (inference code has no licence)

Status: Open.

#24 In OpenMOSS/MOSS-Audio;

· slittycode opened on Jun 5, 2026

Hi, does this series of model have a quantized version like NVFP4 or FP8?

Status: Open.

#23 In OpenMOSS/MOSS-Audio;

· bi1101 opened on Jun 3, 2026

实测发现8B-instruct的asr质量不如4B-instruct的

Status: Open.

#22 In OpenMOSS/MOSS-Audio;

· zshnb opened on May 28, 2026

模型最大支持音频时长是多少？

Status: Open.

#21 In OpenMOSS/MOSS-Audio;

· zshnb opened on May 28, 2026

LoRA config in finetune.py

Status: Open.

#20 In OpenMOSS/MOSS-Audio;

· marcoyang1998 opened on May 25, 2026

8B think模型好像少文件？

Status: Open.

#19 In OpenMOSS/MOSS-Audio;

· qgzang opened on May 20, 2026

支持流式吗？

Status: Open.

#17 In OpenMOSS/MOSS-Audio;

· candy4290 opened on May 19, 2026

音频caption支持输出带时间戳的吗

Status: Open.

#16 In OpenMOSS/MOSS-Audio;

· 1aplace opened on May 12, 2026

现在audio模型有说话人识别的能力吗

Status: Open.

#15 In OpenMOSS/MOSS-Audio;

· mengyuanhui opened on May 11, 2026