Skip to content Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗ Midashenglm 7b 1021 GGUF
This repository contains the GGUF versions of mispeech/midashenglm-7b-1021-fp32. Can be used with llama.cpp for efficient local inference.
This repository contains the GGUF versions of mispeech/midashenglm-7b-1021-fp32. Can be used with llama.cpp for efficient local inference.
KindOtherBase modelQwen/Qwen2.5-Omni-7BVersion52bee4098bf98d08e48d2270f9d12c7fffd9d401Bgrade
Model source
- Kind
- Other
- File size
- 1.90–14.2 GB
- Base model
- Qwen/Qwen2.5-Omni-7B
- Version
- v52bee4098bf98d08e48d2270f9d12c7fffd9d401
- Updated
- Aug 3, 2026
- Source
- Hugging Face
MiDashengLM-7B-1021 (GGUF)
This repository contains the GGUF versions of mispeech/midashenglm-7b-1021-fp32.
Can be used with llama.cpp for efficient local inference.
Usage
Model artifacts
10 artifacts
backbone-bf16.gguf
gguf · 14.2 GB · SHA-256 aa18539ed93c…a5a9 · Hugging Face
Downloadbackbone-IQ1_M.gguf
gguf · 1.90 GB · SHA-256 1b2d181ead27…76a5 · Hugging Face
DownloadSource excerpts
2 excerpts
One-shot inference
llama-cli --model backbone-bf16.gguf --mmproj mmproj-f32.gguf \
--reasoning off --temp 0 --audio audio.wav \
--prompt 'Write a detailed caption.' \
--single-turn --display-prompt
Interactive mode
llama-cli --model backbone-bf16.gguf --mmproj mmproj-f32.gguf \
--reasoning off --temp 0
OpenAI/Anthropic-compatible API server
llama-server --model backbone-bf16.gguf --mmproj mmproj-f32.gguf
Results
The following evaluation results are based on the model version: mispeech/midashenglm-7b-1021-fp32.
Audio Captioning Results
| Domain | Dataset | MiDashengLM | Qwen2.5-Omni-7B | Kimi-Audio-Instruct |
|---|
| Music | MusicCaps | 59.11 | 43.71 | 35.43 |
| Music | Songdescriber | 46.42 | 45.31 | 44.63 |
| Sound | AudioCaps | 62.13 | 60.79 | 49.00 |
| Sound | ClothoV2 | 49.35 | 47.55 | 48.01 |
| Sound | AutoACD | 67.13 | 55.93 |
Metrics: FENSE (higher is better).
Audio and Paralinguistic Classification
| Dataset | Metric | MiDashengLM | Qwen2.5-Omni-7B | Kimi-Audio-Instruct |
|---|
| VoxCeleb1 | ACC↑ | 92.66 | 59.71 | 82.72 |
| VoxLingua107 | ACC↑ | 93.72 | 51.03 | 73.65 |
| VoxCeleb-Gender | ACC↑ | 97.72 | 99.82 | 99.69 |
| VGGSound | ACC↑ | 52.19 | 0.97 | 2.20 |
| Cochlscene | ACC↑ | 75.81 | 23.88 |
ASR Performance
| Dataset | Language | MiDashengLM | Qwen2.5-Omni-7B | Kimi-Audio-Instruct |
|---|
| LibriSpeech test-clean | English | 3.6 | 1.7 | 1.3 |
| LibriSpeech test-other | English | 5.9 | 3.4 | 2.4 |
| People's Speech | English | 26.12 | 28.6 | 22.3 |
| AISHELL2 Mic | Chinese | 3.2 | 2.5 | 2.7 |
| AISHELL2 iOS | Chinese | 2.9 | |
Metrics: WER/CER (lower is better).
Question Answering Results
| Dataset | Subset | Metric | MiDashengLM | Qwen2.5-Omni-7B | Kimi-Audio-Instruct |
|---|
| MMAU-Pro | IF | ACC↑ | 37.93 | 61.30 | 42.30 |
| MMAU-Pro | Multi-Audio | ACC↑ | 42.33 | 24.30 | 17.20 |
| MMAU-Pro | Music | ACC↑ | 62.20 | 61.50 | 57.60 |
| MMAU-Pro | Open-ended | ACC↑ | 63.21 | 52.30 | 34.50 |
|
Metrics: Higher is better.
Citation
MiDashengLM is under the Apache License 2.0, and we encourage its use in both research and business applications.
If you find MiDashengLM useful in your research, please consider citing our work:
@techreport{midashenglm7b,
title = {MiDashengLM: Efficient Audio Understanding with General Audio Captions},
author = {{Horizon Team, MiLM Plus}},
institution= {Xiaomi Inc.},
year = {2025},
note = {Contributors: Heinrich Dinkel et al. (listed alphabetically in Appendix B)},
url = {https://arxiv.org/abs/2508.03983},
eprint = {2508.03983},
}
gguf · 2.59 GB · SHA-256 8db737dbeb74…be81 · Hugging Face
backbone-Q2_K_S.gguf
gguf · 2.64 GB · SHA-256 20473a49e796…2f8e · Hugging Face
Downloadbackbone-Q3_K_M.gguf
gguf · 3.55 GB · SHA-256 87d5e27200b3…8061 · Hugging Face
Downloadbackbone-Q4_K_M.gguf
gguf · 4.36 GB · SHA-256 6151abc3259f…6d7c · Hugging Face
Downloadbackbone-Q5_K_M.gguf
gguf · 5.07 GB · SHA-256 a93be2827910…2ce4 · Hugging Face
Downloadbackbone-Q6_K.gguf
gguf · 5.82 GB · SHA-256 1bbf982322de…2432 · Hugging Face
Downloadbackbone-Q8_0.gguf
gguf · 7.54 GB · SHA-256 87c46a566fb1…7812 · Hugging Face
Downloadmmproj-f32.gguf
gguf · 2.48 GB · SHA-256 689898523691…f52d · Hugging Face
Download---
license: apache-2.0
language:
- en
- zh
- th
- id
- vi
pipeline_tag: audio-text-to-text
tags:
- multimodal
- audio-language-model
- audio
base_model:
- mispeech/dasheng-0.6B
- Qwen/Qwen2.5-Omni-7B
base_model_relation: finetune
---
# MiDashengLM-7B-1021 (GGUF)
This repository contains the GGUF versions of [mispeech/midashenglm-7b-1021-fp32](https://huggingface.co/mispeech/midashenglm-7b-1021-fp32).
Can be used with llama.cpp for efficient local inference.
## Usage
> [!IMPORTANT]
> This model currently requires our [llama.cpp fork](https://github.com/zhoukezi/llama.cpp) to run.
> See the [build instructions](https://github.com/zhoukezi/llama.cpp/blob/master/docs/build.md) to get started.
### One-shot inference
```bash
llama-cli --model backbone-bf16.gguf --mmproj mmproj-f32.gguf \
--reasoning off --temp 0 --audio audio.wav \
--prompt 'Write a detailed caption.' \
--single-turn --display-prompt
```
### Interactive mode
```bash
llama-cli --model backbone-bf16.gguf --mmproj mmproj-f32.gguf \
--reasoning off --temp 0
```
For more CLI options, see the [CLI documentation](https://github.com/zhoukezi/llama.cpp/blob/master/tools/cli/README.md).
### OpenAI/Anthropic-compatible API server
```bash
llama-server --model backbone-bf16.gguf --mmproj mmproj-f32.gguf
```
For more server options, see the [server documentation](https://github.com/zhoukezi/llama.cpp/blob/master/tools/server/README.md).
## Results
The following evaluation results are based on the model version: `mispeech/midashenglm-7b-1021-fp32`.
### Audio Captioning Results
| Domain | Dataset | MiDashengLM | Qwen2.5-Omni-7B | Kimi-Audio-Instruct |
|:--------:|:--------------:|:--------------:|:----------------:|:-------------------:|
| Music | MusicCaps | **59.11** | 43.71 | 35.43 |
| Music | Songdescriber | **46.42** | 45.31 | 44.63 |
| Sound | AudioCaps | **62.13** | 60.79 | 49.00 |
| Sound | ClothoV2 | **49.35** | 47.55 | 48.01 |
| Sound | AutoACD | **67.13** | 55.93 | 44.76 |
*Metrics: FENSE (higher is better).*
### Audio and Paralinguistic Classification
| Dataset | Metric | MiDashengLM | Qwen2.5-Omni-7B | Kimi-Audio-Instruct |
|:----------------:|:------:|:--------------:|:----------------:|:------------------:|
| VoxCeleb1 | ACC↑ | **92.66** | 59.71 | 82.72 |
| VoxLingua107 | ACC↑ | **93.72** | 51.03 | 73.65 |
| VoxCeleb-Gender | ACC↑ | 97.72 | **99.82** | 99.69 |
| VGGSound...
Source context: 344 downloads · 2 likes · Pipeline audio-text-to-text · Repo mispeech/midashenglm-7b-1021-gguf
| NSynth | ACC↑ | 80.32 | 60.45 | 38.09 |
| FSDKaggle2018 | ACC↑ | 73.38 | 31.38 | 24.75 |
| FSD50K | mAP↑ | 38.10 | 23.87 | 27.23 |
2.6
| AISHELL2 Android | Chinese | 3.1 | 2.7 | 2.6 |
| GigaSpeech2 | Indonesian | 22.3 | 21.2 | >100 |
| GigaSpeech2 | Thai | 38.4 | 53.8 | >100 |
| GigaSpeech2 | Viet | 17.7 | 18.6 | >100 |
| MMAU-Pro | Sound–Music | ACC↑ | 42.00 | 40.00 | 46.00 |
| MMAU-Pro | Sound–Music–Speech | ACC↑ | 71.43 | 28.50 | 42.80 |
| MMAU-Pro | Spatial | ACC↑ | 18.77 | 41.20 | 43.70 |
| MMAU-Pro | Speech | ACC↑ | 61.17 | 57.40 | 52.20 |
| MMAU-Pro | Speech–Music | ACC↑ | 58.70 | 53.20 | 54.30 |
| MMAU-Pro | Speech–Sound | ACC↑ | 51.14 | 60.20 | 48.90 |
| MMAU-Pro | Voice | ACC↑ | 54.83 | 60.00 | 50.60 |
| MMAU-Pro | Average | ACC↑ | 55.92 | 52.20 | 46.60 |
| MMAU-v05.15.25 | Sound | ACC↑ | 77.48 | 78.10 | 75.68 |
| MMAU-v05.15.25 | Music | ACC↑ | 70.96 | 65.90 | 66.77 |
| MMAU-v05.15.25 | Speech | ACC↑ | 76.28 | 70.60 | 62.16 |
| MMAU-v05.15.25 | Average | ACC↑ | 74.90 | 71.50 | 68.20 |
| MuChoMusic | | ACC↑ | 73.04 | 64.79 | 67.40 |
| MusicQA | | FENSE↑ | 61.56 | 60.60 | 40.00 |
| AudioCaps-QA | | FENSE↑ | 54.20 | 53.28 | 47.34 |
