AEmotionStudio / midashenglm-gen-models

huggingface.co
Total runs: 7
24-hour runs: -1
7-day runs: 2
30-day runs: -2
Model's Last Updated: August 31 2026
text-to-audio

Introduction of midashenglm-gen-models

Model Details of midashenglm-gen-models

MiDashengLM-Gen — MAESTRO mirror

Mirror of mispeech/midashenglm-gen (Xiaomi MiLM Plus, Apache-2.0, arXiv:2608.11804 ) maintained by AEmotionStudio for download-on-demand use inside MAESTRO.

MiDashengLM-Gen generates blended 16 kHz mono audio scenes — speech, music, sound effects and ambience in one pass — from a structured multi-view caption ( <|caption|> <|asr|> <|speech|> <|sfx|> <|music|> <|env|> ), 1–20 s, via Qwen3-1.7B + per-token conditional flow matching + a Vocos vocoder.

Differences from upstream
  • Precision : weights may be cast fp32 → bf16 (one model.safetensors ); NOTICE records the payload precision and per-file sha256 digests. Upstream fp32 was sha256-verified against the hub before casting, and a bf16 payload is published only after an A/B parity pass on GPU.
  • Bundled tokenizer : tokenizer/ carries the Qwen3-1.7B tokenizer + config pinned at the revision recorded in tokenizer/tokens_added.json (upstream fetches it from the network, unpinned, at construction time). tokens_added.json also records the 28 special tokens appended in training order — loader drift fails loudly instead of generating noise.
  • No remote code : the modeling_*.py trust-remote-code files are not mirrored; MAESTRO vendors the inference code (Apache-2.0) in-tree.
License

Apache-2.0 (see LICENSE ). Weights © 2026 MiLM Plus Xiaomi Inc.; tokenizer files © Alibaba (Qwen), Apache-2.0. See upstream for the model card, evaluation results and use-restriction notes.

Runs of AEmotionStudio midashenglm-gen-models on huggingface.co

7
Total runs
-1
24-hour runs
0
3-day runs
2
7-day runs
-2
30-day runs

More Information About midashenglm-gen-models huggingface.co Model

More midashenglm-gen-models license Visit here:

https://choosealicense.com/licenses/apache-2.0

midashenglm-gen-models huggingface.co

midashenglm-gen-models huggingface.co is an AI model on huggingface.co that provides midashenglm-gen-models's model effect (), which can be used instantly with this AEmotionStudio midashenglm-gen-models model. huggingface.co supports a free trial of the midashenglm-gen-models model, and also provides paid use of the midashenglm-gen-models. Support call midashenglm-gen-models model through api, including Node.js, Python, http.

AEmotionStudio midashenglm-gen-models online free

midashenglm-gen-models huggingface.co is an online trial and call api platform, which integrates midashenglm-gen-models's modeling effects, including api services, and provides a free online trial of midashenglm-gen-models, you can try midashenglm-gen-models online for free by clicking the link below.

AEmotionStudio midashenglm-gen-models online free url in huggingface.co:

https://huggingface.co/AEmotionStudio/midashenglm-gen-models

midashenglm-gen-models install

midashenglm-gen-models is an open source model from GitHub that offers a free installation service, and any user can find midashenglm-gen-models on GitHub to install. At the same time, huggingface.co provides the effect of midashenglm-gen-models install, users can directly use midashenglm-gen-models installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

midashenglm-gen-models install url in huggingface.co:

https://huggingface.co/AEmotionStudio/midashenglm-gen-models

Url of midashenglm-gen-models

midashenglm-gen-models huggingface.co Url

Provider of midashenglm-gen-models huggingface.co

AEmotionStudio
ORGANIZATIONS

Other API from AEmotionStudio

huggingface.co

Total runs: 213
Run Growth: -90
Growth Rate: -42.25%
Updated:February 25 2026