Home Creators Posts Import Register Favorites Logout

Videos

  • 433173382.mp4

Downloads

Content

1-Click installers for MMAudio on Windows, RunPod and Massed Compute. Especially useful for generating sound for AI videos, Games and such

Patreon exclusive posts index to find our scripts easily, Patreon scripts updates history to see which updates arrived to which scripts and amazing Patreon special generative scripts list that you can use in any of your task.

Join discord to get help, chat, discuss and also tell me your discord username to get your special rank : SECourses Discord

Please also Star, Watch and Fork our Stable Diffusion & Generative AI  GitHub repository and join our Reddit subreddit and follow me on LinkedIn (my real profile)

=======

Latest installer zip file : MMAudio_v12.zip

Full Tutorial : https://youtu.be/504f8S4MLTw

How To Generate Prompts

You can use Google Studio AI for free and upload Video or Image

https://aistudio.google.com/prompts/new_chat

analyze attached video. i will generate audio for this video via text to audio model. describe audio for this video in a single sentence. so write me a great sentence to generate a great audio matching to this video. use passive sentence like "an intense, sparking buzz of electric hair clippers loudly cutting through something tough, like metal, rather than hair."

Few Text-to-Audio examples - Make duration accordingly to the sound you expect

rhythmic splashing and lapping of water

creamy sound of mashed potatoes being scooped

10 April 2026 Update V12

Installers upgraded to latest CUDA 12.9, Torch 2.8, Flash Attention 2.8.3, xFormers, Sage Attention

All compiled for RTX 1000 to 5000 series including cloud GPUs like A100, H100, B200, etc

Both Windows, RunPod and Massed Compute installers upgraded

Also now we support SimplePod

Installers upgraded to uv install thus now like 100x faster on RunPod

There was a bug that caused app to fail and it is fixed

Overwrite older files, delete venv folder inside MMAudio and run Windows_Install_or_Update.bat file or do a fresh install

Windows Requirements

Python 3.10, FFmpeg, CUDA 12.9, C++ tools and Git

If it doesn't work make sure to below tutorial and install everything exactly as shown in this below tutorial

https://youtu.be/DrhUHnYfwC0

How To Install and Use:

20 March 2025 Full tutorial : https://youtu.be/504f8S4MLTw

Windows:

First watch above requirements tutorial and follow every step

Extract zip archive into a drive like c:/MMAudio_v12

Longer directory path causes issues and don't use space characters in folder names

Double click and install with Windows_Install.bat and that is it

Then run with Windows_Start_App.bat to start MMAudio SECourses APP

And follow the above tutorial

Massed Compute (Recommend Cloud) :

Please register via this link : https://vm.massedcompute.com/signup?linkId=lp_034338&sourceId=secourses&tenantId=massed-compute

Use our coupon SECourses

Our coupon now works on all GPUs

Then select our image SECourses from Creator dropdown

Then follow Massed_Compute_Instructions_READ.txt

Same as my any other Massed Compute installer script

Example tutorial for learn how to install and use Massed Compute

(Starts at 12:58) : https://youtu.be/KW-MHmoNcqo?si=G1WbG-Qw4ujWvOtG&t=778

RunPod (Cloud):

Please register via this link : https://get.runpod.io/955rkuppqv4h

Then follow Runpod_Instructions_READ.txt

Same as my any other RunPod installer script

Use the template written in Runpod_Instructions_READ.txt file

Example tutorial for learn how to install and use RunPod

(starts at 22:03) : https://youtu.be/KW-MHmoNcqo?si=QN8X8Sjn13ZYu-EU&t=1323

Files

Comments

The90sPope

During installation I get this, but it says all installed fine in the end. Attempting uninstall: torch Found existing installation: torch 2.5.1+cu124 Uninstalling torch-2.5.1+cu124: Successfully uninstalled torch-2.5.1+cu124 Attempting uninstall: torchvision Found existing installation: torchvision 0.20.1+cu124 Uninstalling torchvision-0.20.1+cu124: Successfully uninstalled torchvision-0.20.1+cu124 ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts. xformers 0.0.28.post3 requires torch==2.5.1, but you have torch 2.6.0 which is incompatible. torchaudio 2.5.1+cu124 requires torch==2.5.1+cu124, but you have torch 2.6.0 which is incompatible. But when i try starting app i get this: C:\MMAudio\MMAudio>call .\venv\Scripts\activate.bat || exit /b Traceback (most recent call last): File "C:\MMAudio\MMAudio\gradio_demo.py", line 10, in import torchaudio File "C:\MMAudio\MMAudio\venv\lib\site-packages\torchaudio\__init__.py", line 2, in from . import _extension # noqa # usort: skip File "C:\MMAudio\MMAudio\venv\lib\site-packages\torchaudio\_extension\__init__.py", line 38, in _load_lib("libtorchaudio") File "C:\MMAudio\MMAudio\venv\lib\site-packages\torchaudio\_extension\utils.py", line 60, in _load_lib torch.ops.load_library(path) File "C:\MMAudio\MMAudio\venv\lib\site-packages\torch\_ops.py", line 1357, in load_library ctypes.CDLL(path) File "C:\Python310\lib\ctypes\__init__.py", line 374, in __init__ self._handle = _dlopen(self._name, mode) FileNotFoundError: Could not find module 'C:\MMAudio\MMAudio\venv\Lib\site-packages\torchaudio\lib\libtorchaudio.pyd' (or one of its dependencies). Try using the full path with constructor syntax.

Hockey

Sound effects work but text to speech is not good at all. Is there another software that can do text to speech or perhaps I need to prompt it better?

The90sPope

nice, works now, it would be nice to just be able to output audio or download audio only, I'm gonna proccess in in a DAW, so getting video with audio is extra work for me.

The90sPope

Nice, since the video output has also reduced quality compared to the original, so audio only is nice.

The90sPope

it says v5 after I updated to v7, also no new option, everything the same. I also forgot to move output folder content out before updating so 7 hours of work are gone XD

The90sPope

I fucked up lol, I deleted folder before backing up haha. Happens, was very tired, my brain did not work lol.

The90sPope

Thanks for fixing stuff fast, you're very reactive to all the feedback. Greatly appreciated

PeroTech Community

Requirement already satisfied: py-cpuinfo in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from deepspeed==0.16.4) (9.0.0) Requirement already satisfied: pydantic>=2.0.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from deepspeed==0.16.4) (2.11.3) Requirement already satisfied: torch in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from deepspeed==0.16.4) (2.7.0.dev20250311+cu128) Requirement already satisfied: tqdm in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from deepspeed==0.16.4) (4.67.1) Requirement already satisfied: nvidia-ml-py in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from deepspeed==0.16.4) (12.570.86) Requirement already satisfied: annotated-types>=0.6.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from pydantic>=2.0.0->deepspeed==0.16.4) (0.7.0) Requirement already satisfied: pydantic-core==2.33.1 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from pydantic>=2.0.0->deepspeed==0.16.4) (2.33.1) Requirement already satisfied: typing-extensions>=4.12.2 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from pydantic>=2.0.0->deepspeed==0.16.4) (4.12.2) Requirement already satisfied: typing-inspection>=0.4.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from pydantic>=2.0.0->deepspeed==0.16.4) (0.4.0) Requirement already satisfied: filelock in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from torch->deepspeed==0.16.4) (3.16.1) Requirement already satisfied: sympy>=1.13.3 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from torch->deepspeed==0.16.4) (1.13.3) Requirement already satisfied: networkx in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from torch->deepspeed==0.16.4) (3.4.2) Requirement already satisfied: jinja2 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from torch->deepspeed==0.16.4) (3.1.4) Requirement already satisfied: fsspec in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from torch->deepspeed==0.16.4) (2024.10.0) Requirement already satisfied: colorama in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from tqdm->deepspeed==0.16.4) (0.4.6) Requirement already satisfied: mpmath<1.4,>=1.1.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from sympy>=1.13.3->torch->deepspeed==0.16.4) (1.3.0) Requirement already satisfied: MarkupSafe>=2.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from jinja2->torch->deepspeed==0.16.4) (2.1.5) Requirement already satisfied: triton-windows in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (3.3.0a0.post18) Requirement already satisfied: setuptools>=40.8.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from triton-windows) (65.5.0) Requirement already satisfied: huggingface_hub in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (0.30.2) Requirement already satisfied: ipywidgets in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (8.1.6) Requirement already satisfied: hf_transfer in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (0.1.9) Requirement already satisfied: filelock in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from huggingface_hub) (3.16.1) Requirement already satisfied: fsspec>=2023.5.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from huggingface_hub) (2024.10.0) Requirement already satisfied: packaging>=20.9 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from huggingface_hub) (24.2) Requirement already satisfied: pyyaml>=5.1 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from huggingface_hub) (6.0.2) Requirement already satisfied: requests in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from huggingface_hub) (2.32.3) Requirement already satisfied: tqdm>=4.42.1 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from huggingface_hub) (4.67.1) Requirement already satisfied: typing-extensions>=3.7.4.3 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from huggingface_hub) (4.12.2) Requirement already satisfied: comm>=0.1.3 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipywidgets) (0.2.2) Requirement already satisfied: ipython>=6.1.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipywidgets) (8.35.0) Requirement already satisfied: traitlets>=4.3.1 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipywidgets) (5.14.3) Requirement already satisfied: widgetsnbextension~=4.0.14 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipywidgets) (4.0.14) Requirement already satisfied: jupyterlab_widgets~=3.0.14 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipywidgets) (3.0.14) Requirement already satisfied: colorama in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (0.4.6) Requirement already satisfied: decorator in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (5.2.1) Requirement already satisfied: exceptiongroup in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (1.2.2) Requirement already satisfied: jedi>=0.16 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (0.19.2) Requirement already satisfied: matplotlib-inline in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (0.1.7) Requirement already satisfied: prompt_toolkit<3.1.0,>=3.0.41 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (3.0.51) Requirement already satisfied: pygments>=2.4.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (2.19.1) Requirement already satisfied: stack_data in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ipython>=6.1.0->ipywidgets) (0.6.3) Requirement already satisfied: charset-normalizer<4,>=2 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from requests->huggingface_hub) (3.4.1) Requirement already satisfied: idna<4,>=2.5 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from requests->huggingface_hub) (3.10) Requirement already satisfied: urllib3<3,>=1.21.1 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from requests->huggingface_hub) (2.4.0) Requirement already satisfied: certifi>=2017.4.17 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from requests->huggingface_hub) (2025.1.31) Requirement already satisfied: parso<0.9.0,>=0.8.4 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from jedi>=0.16->ipython>=6.1.0->ipywidgets) (0.8.4) Requirement already satisfied: wcwidth in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from prompt_toolkit<3.1.0,>=3.0.41->ipython>=6.1.0->ipywidgets) (0.2.13) Requirement already satisfied: executing>=1.2.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from stack_data->ipython>=6.1.0->ipywidgets) (2.2.0) Requirement already satisfied: asttokens>=2.1.0 in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from stack_data->ipython>=6.1.0->ipywidgets) (3.0.0) Requirement already satisfied: pure-eval in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from stack_data->ipython>=6.1.0->ipywidgets) (0.2.3) Requirement already satisfied: ffmpeg-python in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (0.2.0) Requirement already satisfied: future in c:\mmaudio_v8\mmaudio\venv\lib\site-packages (from ffmpeg-python) (1.0.0) mmaudio_large_44k_v2.pth: 19%|█████████▋ | 765M/4.12G [10:42<46:56, 1.19MB/s] Traceback (most recent call last): File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\file_download.py", line 423, in http_get hf_transfer.download( Exception: Error while removing corrupted file: The process cannot access the file because it is being used by another process. (os error 32) The above exception was the direct cause of the following exception: Traceback (most recent call last): File "C:\MMAudio_v8\Download_Models.py", line 3, in snapshot_download( File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\utils\_validators.py", line 114, in _inner_fn return fn(*args, **kwargs) File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\_snapshot_download.py", line 294, in snapshot_download _inner_hf_hub_download(file) File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\_snapshot_download.py", line 270, in _inner_hf_hub_download return hf_hub_download( File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\utils\_validators.py", line 114, in _inner_fn return fn(*args, **kwargs) File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\file_download.py", line 941, in hf_hub_download return _hf_hub_download_to_local_dir( File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\file_download.py", line 1241, in _hf_hub_download_to_local_dir _download_to_tmp_and_move( File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\file_download.py", line 1675, in _download_to_tmp_and_move http_get( File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\huggingface_hub\file_download.py", line 434, in http_get raise RuntimeError( RuntimeError: An error occurred while downloading using `hf_transfer`. Consider disabling HF_HUB_ENABLE_HF_TRANSFER for better error handling. Virtual environment made and installed properly Press any key to continue . . .

PeroTech Community

======================================================== Select GPU Option: 1: Above 8 GB GPUs - Full Speed 2: Below 8 GB GPUs - VRAM Optimized - Same Quality ======================================================== Enter your selection: [1,2]?1 Running in Full Speed mode... [INFO ]: Loaded hf-hub:apple/DFN5B-CLIP-ViT-H-14-384 model config. [INFO ]: Loading pretrained hf-hub:apple/DFN5B-CLIP-ViT-H-14-384 weights (C:\Users\USER PC\.cache\huggingface\hub\models--apple--DFN5B-CLIP-ViT-H-14-384\snapshots\01b771ed0d1395ca5ffdd279897d665ebe00dfd2\open_clip_pytorch_model.bin). Traceback (most recent call last): File "C:\MMAudio_v8\MMAudio\app.py", line 89, in net, feature_utils, seq_cfg = get_model() File "C:\MMAudio_v8\MMAudio\app.py", line 76, in get_model feature_utils = FeaturesUtils( File "C:\MMAudio_v8\MMAudio\mmaudio\model\utils\features_utils.py", line 49, in __init__ self.clip_model = create_model_from_pretrained('hf-hub:apple/DFN5B-CLIP-ViT-H-14-384', File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\open_clip\factory.py", line 562, in create_model_from_pretrained model = create_model( File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\open_clip\factory.py", line 410, in create_model load_checkpoint(model, checkpoint_path, weights_only=load_weights_only) File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\open_clip\factory.py", line 190, in load_checkpoint state_dict = load_state_dict(checkpoint_path, device=device, weights_only=weights_only) File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\open_clip\factory.py", line 160, in load_state_dict checkpoint = torch.load(checkpoint_path, map_location=device, weights_only=weights_only) File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\torch\serialization.py", line 1541, in load return _legacy_load( File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\torch\serialization.py", line 1787, in _legacy_load return legacy_load(f) File "C:\MMAudio_v8\MMAudio\venv\lib\site-packages\torch\serialization.py", line 1654, in legacy_load raise RuntimeError( RuntimeError: Cannot use ``weights_only=True`` with files saved in the legacy .tar format. In PyTorch 2.6, we changed the default value of the `weights_only` argument in `torch.load` from `False` to `True`. Re-running `torch.load` with `weights_only` set to `False` will likely succeed, but it can result in arbitrary code execution. Do it only if you got the file from a trusted source. Press any key to continue . . .

JP LONB

Tried to install it on my laptop and then run it on the low VRAM option (my RTX has only 8GB) and I got the error message below, unfortunately. Chat says that it's "it’s a code/library version mismatch" because "the local Hugging Face Hub library (huggingface_hub) has changed its internal API. Newer versions now require two keyword-only arguments (proxies and resume_download) that older code didn’t provide.". I tried to solve the issue with its help but did not manage. Can you please check?. The message I got is: "Traceback (most recent call last): File "C:\AI\MMAudio\MMAudio\app.py", line 89, in net, feature_utils, seq_cfg = get_model() File "C:\AI\MMAudio\MMAudio\app.py", line 76, in get_model feature_utils = FeaturesUtils( File "C:\AI\MMAudio\MMAudio\mmaudio\model\utils\features_utils.py", line 67, in __init__ self.tod = AutoEncoderModule(vae_ckpt_path=tod_vae_ckpt, File "C:\AI\MMAudio\MMAudio\mmaudio\ext\autoencoder\autoencoder.py", line 30, in __init__ self.vocoder = BigVGANv2.from_pretrained('nvidia/bigvgan_v2_44khz_128band_512x', File "C:\AI\MMAudio\MMAudio\venv\lib\site-packages\huggingface_hub\utils\_validators.py", line 89, in _inner_fn return fn(*args, **kwargs) File "C:\AI\MMAudio\MMAudio\venv\lib\site-packages\huggingface_hub\hub_mixin.py", line 559, in from_pretrained instance = cls._from_pretrained( TypeError: BigVGAN._from_pretrained() missing 2 required keyword-only arguments: 'proxies' and 'resume_download'"

JP LONB

Thanks a lot dear Furkan! I will try this asap!

JP LONB

Tested it. Works perfectly! Thks again!

Hockey

Maybe it's just me but I don't think the results are any good, anybody else having success?