Release WorldSonus C3 no-H0 150k inference weights

LICENSE

@@ -0,0 +1,18 @@
+Creative Commons Attribution-NonCommercial 4.0 International
+SPDX-License-Identifier: CC-BY-NC-4.0
+
+Copyright (c) 2026 WorldSonus contributors
+
+Except for third-party material identified in THIRD_PARTY_NOTICES.md,
+WorldSonus's original contributions are licensed under the Creative Commons
+Attribution-NonCommercial 4.0 International Public License.
+
+The full legal terms are incorporated by reference and are available at:
+https://creativecommons.org/licenses/by-nc/4.0/legalcode.en
+
+You may share and adapt the licensed material for non-commercial purposes,
+provided you give appropriate attribution, link to the license, and indicate
+changes. The legal terms linked above govern, not this short summary.
+
+Third-party components and source media retain their respective licenses.
+No warranties are provided. See the legal terms for the full disclaimer.

README.md

@@ -0,0 +1,54 @@
+---
+license: cc-by-nc-4.0
+library_name: pytorch
+tags:
+- worldsonus
+- video-to-audio
+- audio-generation
+- stereo
+---
+
+# WorldSonus: Bringing Sound to Worlds
+
+Inference weights for the final **C3 no-H0, 150k EMA** model.
+
+Code and usage: [NoizAI/WorldSonus](https://github.com/NoizAI/WorldSonus).
+
+## Contents
+
+- `worldsonus_150k.pt`: 638 inference tensors. No optimizer, training arguments,
+  inactive history adapters, or training-only alignment projector.
+- `audio_codec.pt`: causal stereo 48 kHz decoder, 64 latent channels at 30 Hz.
+- `z_stats.pt`: matching latent normalization statistics. Do not substitute
+  statistics from another model or recompute them for these weights.
+
+All files load with `torch.load(..., weights_only=True)`. DINOv3 and T5Gemma 2
+encoders are downloaded separately under their upstream licenses.
+
+## Usage
+
+Install the inference repository, then run:
+
+```bash
+python scripts/download_model.py --repo FF2416/WorldSonus --output assets
+python scripts/download_encoders.py --output assets
+```
+
+See the code repository for feature extraction and audio generation commands.
+Defaults: Euler15, AR visual/text CFG3/3, DINO APG8, eta0, momentum0.25,
+radius0.12421, AR noise0.1, no recache. Three latents per chunk (100 ms).
+
+## Limitations and responsible use
+
+Generated audio may not match every visual event and may contain artifacts.
+Do not present generated audio as authentic recorded evidence. Respect source
+media rights and avoid recreating identifiable voices without consent.
+
+This is a research inference release, not a fully qualified live-streaming
+deployment. Training code and datasets are not included.
+
+## License
+
+WorldSonus contributions and these released model weights are provided under
+CC BY-NC 4.0 for non-commercial use with attribution. Upstream third-party
+components retain their own terms; see LICENSE and THIRD_PARTY_NOTICES.md.

THIRD_PARTY_NOTICES.md

@@ -0,0 +1,37 @@
+# Third-party notices
+
+WorldSonus's original contributions use CC BY-NC 4.0 (see LICENSE). This does
+not replace upstream licenses or restrict rights independently granted by them.
+The original implementation's MIT copyright is retained below:
+
+Copyright (c) 2026 Koichi Saito
+
+The causal audio decoder follows the Oobleck codec implementation in
+[stable-audio-tools](https://github.com/Stability-AI/stable-audio-tools),
+Copyright (c) 2023 Stability AI. SnakeBeta follows
+[BigVGAN](https://github.com/NVIDIA/BigVGAN),
+Copyright (c) 2024 NVIDIA CORPORATION. Both are MIT-licensed under the grant below.
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+
+Model weights are distributed separately. The supplied demonstration video
+does not grant rights to third-party source footage depicted within it.
+No datasets or cached features are distributed here.
+DINOv3 and T5Gemma 2 assets have their own model licenses and access conditions;
+this code license does not grant access to or relicense those assets.

audio_codec.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b25deff42e6e1f0c060de1d84b31e031f362c90b801475ac0598773f172d4fdd
+size 301015270

worldsonus_150k.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6c197850cacda63d98eb47cdf4123fc23f7c900e2f4babc93a8c643142bd078
+size 4123796958

z_stats.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e25f63564240a84763a7926eb55bc64b24d0e866e57c791a2959bec9757d65e
+size 2016