tom/kyutai-moshi
1
0
Fork 0
Code Issues Pull requests Packages 1 Activity Actions

Readme [5/N] (#34)

Browse source 
* wip

* plop

* fix

* readme

* fix

* update

* plop

* testing full title

* Update README.md

Co-authored-by: Christopher Fleetwood <45471420+FL33TW00D@users.noreply.github.com>

* Update README.md

Co-authored-by: Christopher Fleetwood <45471420+FL33TW00D@users.noreply.github.com>

* Update README.md

Co-authored-by: Christopher Fleetwood <45471420+FL33TW00D@users.noreply.github.com>

* Update README.md

Co-authored-by: Christopher Fleetwood <45471420+FL33TW00D@users.noreply.github.com>

* Update README.md

Co-authored-by: Christopher Fleetwood <45471420+FL33TW00D@users.noreply.github.com>

* Update README.md

Co-authored-by: Christopher Fleetwood <45471420+FL33TW00D@users.noreply.github.com>

* missing files

* wip simplify readme

* simplify

* fix

* test

* make pre commit always run

* revert

* fix

* nits

* fix

* test

* test

* test

* missing file

---------

Co-authored-by: Christopher Fleetwood <45471420+FL33TW00D@users.noreply.github.com>
This commit is contained in:
Alexandre Défossez 2024-09-18 03:33:53 +02:00 • committed by GitHub
parent 9b288dec20
commit ebd489f73c
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
19 changed files with 601 additions and 42 deletions
Download patch file Download diff file Expand all files Collapse all files

2
.github/actions/moshi_build/action.yml vendored
View file

@ -19,9 +19,9 @@ runs:
. env/bin/activate
python -m pip install --upgrade pip
pip install torch==2.4.0 --index-url https://download.pytorch.org/whl/cpu
pip install -e './moshi[dev]'
- name: Setup env
shell: bash
run: |
. env/bin/activate
pre-commit install
pip install -e './moshi[dev]'

8
.pre-commit-config.yaml
View file

@ -6,19 +6,23 @@ repos:
language: system
entry: bash -c 'cd moshi && flake8'
pass_filenames: false
always_run: true
- id: pyright-moshi
name: pyright on moshi package
language: system
entry: bash -c 'cd moshi && pyright'
entry: scripts/run_ci_when_installed.sh moshi 'cd moshi && pyright'
pass_filenames: false
always_run: true
- id: flake8-moshi_mlx
name: flake8 on moshi_mlx package
language: system
entry: bash -c 'cd moshi_mlx && flake8'
pass_filenames: false
always_run: true
- id: pyright-moshi_mlx
name: pyright on moshi_mlx package
language: system
entry: bash -c 'cd moshi_mlx && pyright'
entry: scripts/run_ci_when_installed.sh moshi_mlx 'cd moshi_mlx && pyright'
pass_filenames: false
always_run: true

143
README.md
View file

@ -1,9 +1,85 @@
# moshi
# Moshi: a speech-text fundation model for real time dialogue
![precommit badge](https://github.com/kyutai-labs/moshi/workflows/precommit/badge.svg)
![rust ci badge](https://github.com/kyutai-labs/moshi/workflows/Rust%20CI/badge.svg)
[Moshi][moshi] is a speech-text foundation model and **full-duplex** spoken dialogue framework.
It uses [Mimi][moshi], a state-of-the-art streaming neural audio codec. Mimi operates at 12.5 Hz, and compresses
audio down to 1.1 kbps, in a fully streaming manner (latency of 80ms, the frame size),
yet performs better than existing, non-streaming, codec like
[SpeechTokenizer](https://github.com/ZhangXInFD/SpeechTokenizer) (50 Hz, 4 kbps), or [SemantiCodec](https://github.com/haoheliu/SemantiCodec-inference) (50 Hz, 1kbps).
Moshi models **two streams of audio**: one corresponds to Moshi, and one to the user.
At inference, the stream from the user is taken from the audio input,
and the one for Moshi is sampled from. Along that, Moshi predicts text tokens corresponding to its own speech, its **inner monologue**,
which greatly improves the quality of its generation. A small depth transformer models inter codebook dependencies for a given step,
while a large, 7B parameter Transformer models the temporal dependencies. Moshi achieves a theoretical latency
of 160ms (80ms for the frame size of Mimi + 80ms of acoustic delay), with a practical overall latency as low as 200ms.
[Talk to Moshi](https://moshi.chat) now on our live demo.
<p align="center">
<img src="./moshi.png" alt="Schema representing the structure Moshi. Moshi models two streams of audio:
one corresponds to Moshi, and one to the user. At inference, the one from the user is taken from the audio input,
and the one for Moshi is sampled from. Along that, Moshi predicts text tokens corresponding to its own speech
for improved accuracy. A small depth transformer models inter codebook dependencies for a given step."
width="650px"></p>
Mimi builds on previous neural audio codecs such as [SoundStream](https://arxiv.org/abs/2107.03312)
and [EnCodec](https://github.com/facebookresearch/encodec), adding a Transformer both in the encoder and decoder,
and adapting the strides to match an overall frame rate of 12.5 Hz. This allows Mimi to get closer to the
average frame rate of text tokens (~3-4 Hz), and limit the number of auto-regressive steps in Moshi.
Similarly to SpeechTokenizer, Mimi uses a distillation loss so that the first codebook tokens match
a self-supervised representation from [WavLM](https://arxiv.org/abs/2110.13900). Interestingly, while
Mimi is fully causal and streaming, it learns to match sufficiently well the non causal representation from WavLM,
without introducing any delays. Finally, and similary to [EBEN](https://arxiv.org/pdf/2210.14090), Mimi
uses **only an adversarial training loss**, along with feature matching, showing strong improvements in terms of subjective quality
despite its low bitrate.
<p align="center">
<img src="./mimi.png" alt="Schema representing the structure Moshi. Moshi models two streams of audio:
one corresponds to Moshi, and one to the user. At inference, the one from the user is taken from the audio input,
and the one for Moshi is sampled from. Along that, Moshi predicts text tokens corresponding to its own speech
for improved accuracy. A small depth transformer models inter codebook dependencies for a given step."
width="800px"></p>
## Organisation of the repository
There are three separate versions of the moshi inference stack in this repo.
- The python version using PyTorch is in the `moshi` directory.
- The python version using MLX is in the `moshi_mlx` directory.
- The rust version used in production is in the `rust` directory.
- The python version using PyTorch is in the [`moshi/`](moshi/) directory.
- The python version using MLX for M series Macs is in the [`moshi_mlx/`](moshi_mlx/) directory.
- The rust version used in production is in the [`rust/`](rust/) directory.
Finally, the code for the live demo is provided in the [`client/`](client/) directory.
## Requirements
You will need at least Python 3.10. For using the rust backend, you will need a recent version of
the [Rust toolchain](https://rustup.rs/). For specific requirements, please check the individual backends
directories. You can install the PyTorch and MLX clients with the following:
```bash
pip install moshi # moshi PyTorch, from PyPI
pip install moshi_mlx # moshi MLX, from PyPI
# Or the bleeding edge versions for Moshi and Moshi-MLX.
pip install -e "git+https://git@github.com/kyutai-labs/moshi.git#egg=moshi&subdirectory=moshi"
pip install -e "git+https://git@github.com/kyutai-labs/moshi.git#egg=moshi_mlx&subdirectory=moshi_mlx"
```
While we hope that the present codebase will work on Windows, we do not provide official support for it.
We have tested the MLX version with MacBook Pro M3. At the moment, we do not support quantization
for the PyTorch version, so you will need a GPU with a significant amount of memory (24GB).
## Development
If you wish to install from a clone of this repository, maybe to further develop Moshi, you can do the following:
```
# From the root of the clone of the repo
pip install -e 'moshi[dev]'
pip install -e 'moshi_mlx[dev]'
pre-commit install
```
## Python (PyTorch)
@ -15,38 +91,37 @@ run the model, you can then use either the web UI or a command line client.
Start the server with:
```bash
PYTHONPATH=moshi python -m moshi.server
python -m moshi.server [--gradio_tunnel]
```
And then access the web UI on [localhost:8998](http://localhost:8998).
If the server is running on a remote box, you may want to forward the 8998 port
via your ssh connection so as to be able to access the web UI locally.
And then access the web UI on [localhost:8998](http://localhost:8998). If your GPU is on a distant machine
with no direct access, `--gradio_tunnel` will create a tunnel with a URL accessible from anywhere.
Keep in mind that this tunnel goes through the US and can add significant latency (up to 500ms from Europe).
Alternatively, you might want to use SSH to redirect your connection.
Accessing a server that is not localhost via http may cause issues around using
the microphone in the web UI (in some browsers this is only allowed using
https).
A local client is also available, as
```bash
python -m moshi.client [--url URL_TO_GRADIO]
```
However note, that unlike the web browser, this client is bare bone. It doesn't do any echo cancellation,
nor does it try to compensate for a growing lag by skipping frames.
## Python (MLX) for local inference on macOS
You can either compile and install the `rustymimi` extension or install it via
pip.
Once you have installed `moshi_mlx`, you can run
```bash
# Install from pip:
pip install rustymimi==0.1.1
# Alternatively, if you want to compile the package run:
maturin dev -r -m rust/mimi-pyo3/Cargo.toml
python -m moshi_mlx.local -q 4 # weights quantized to 4 bits
python -m moshi_mlx.local -q 8 # weights quantized to 8 bits
```
Then the model can be run with:
```bash
PYTHONPATH=moshi_mlx python -m moshi_mlx.local \
--model ~/tmp/moshiko_mlx_301e30bf@120.q8.safetensors \
--mimi ~/tmp/tokenizer-e351c8d8-checkpoint125.safetensors \
--quantized 8
```
This uses a command line interface, which is bare bone. It doesn't do any echo cancellation,
nor does it try to compensate for a growing lag by skipping frames.
This uses a command line interface, alternatively you can use `local_web` to use
Alternatively you can use `python -m moshi_mlx.local_web` to use
the web UI, connection is via http on [localhost:8998](http://localhost:8998).
## Rust
@ -102,3 +177,25 @@ npm run build
```
The web UI can then be found in the `client/dist` directory.
## License
The present code is provided under the MIT license for the Python parts, and Apache license for the Rust backend.
The web client code is provided under the MIT license.
Note that parts of this code is based on [AudioCraft](https://github.com/facebookresearch/audiocraft), released under
the MIT license.
## Citation
If you use either Mimi or Moshi, please cite the following paper,
```
@article{defossez2024moshi,
title={Moshi: a speech-text foundation model for real-time dialogue},
author={Alexandre Défossez and Laurent Mazaré and Manu Orsini and Amélie Royer and Patrick Pérez and Hervé Jégou and Edouard Grave and Neil Zeghidour},
journal={arXiv:TBC},
year={2024},
}
```
[moshi]: https://arxiv.org/

23
client/LICENSE Normal file
View file

@ -0,0 +1,23 @@
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.

4
client/README.md
View file

@ -14,3 +14,7 @@ Frontend for the demo.
## Skipping the queue
To skip the queue for standalone use, once the project is running go to `/?worker_addr={WORKER_ADDR}` where `WORKER_ADDR` is your worker instance address.
For example : `https://localhost:5173/?worker_addr=0.0.0.0:8088`
## License
The present code is provided under the MIT license.

BIN
mimi.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 566 KiB

BIN
moshi.png Normal file
View file

Binary file not shown.
Side by side

After

Width:  |  Height:  |  Size: 552 KiB

0
moshi/moshi/LICENSE.audiocraft → moshi/LICENSE.audiocraft
View file

5
moshi/MANIFEST.in Normal file
View file

@ -0,0 +1,5 @@
include LICENSE*
include *.md
include *.cfg
include requirements.txt
include moshi/py.typed

98
moshi/README.md
View file

@ -1 +1,97 @@
# moshi - pytorch
# Moshi - PyTorch
See the [top-level README.md][main_repo] for more information on Moshi.
[Moshi][moshi] is a speech-text foundation model and full-duplex spoken dialogue framework.
It uses [Mimi][moshi], a state-of-the-art streaming neural audio codec. Mimi operates at 12.5 Hz, and compress
audio down to 1.1 kbps, in a fully streaming manner (latency of 80ms, the frame size), yet performs better than existing, non-streaming, codec.
This is the PyTorch implementation for Moshi and Mimi.
## Requirements
You will need at least Python 3.10. We kept a minimal set of dependencies for the current project.
It was tested with PyTorch 2.2 or 2.4. If you need a specific CUDA version, please make sure
to have PyTorch properly installed before installing Moshi.
```bash
pip install moshi # moshi PyTorch, from PyPI
# Or the bleeding edge versions for Moshi
pip install -e "git+https://git@github.com/kyutai-labs/moshi#egg=moshi&subdirectory=moshi"
```
While we hope that the present codebase will work on Windows, we do not provide official support for it.
At the moment, we do not support quantization for the PyTorch version, so you will need a GPU with a significant amount of memory (24GB).
## Usage
This package provides a streaming version of the audio tokenizer (Mimi) and the lm model (Moshi).
In order to run in interactive mode, you need to start a server which will
run the model, you can then use either the web UI or a command line client.
Start the server with:
```bash
python -m moshi.server [--gradio_tunnel]
```
And then access the web UI on [localhost:8998](http://localhost:8998). If your GPU is on a distant machine
with no direct access, `--gradio_tunnel` will create a tunnel with a URL accessible from anywhere.
Keep in mind that this tunnel goes through the US and can add significant latency (up to 500ms from Europe).
Alternatively, you might want to use SSH to redirect your connection.
Accessing a server that is not localhost via http may cause issues around using
the microphone in the web UI (in some browsers this is only allowed using
https).
A local client is also available, as
```bash
python -m moshi.client [--url URL_TO_GRADIO]
```
However note, that unlike the web browser, this client is bare bone. It doesn't do any echo cancellation,
nor does it try to compensate for a growing lag by skipping frames.
## Development
If you wish to install from a clone of this repository, maybe to further develop Moshi, you can do the following:
```bash
# From the current folder (e.g. `moshi/`)
pip install -e '.[dev]'
pre-commit install
```
Once locally installed, Mimi can be tested with the following command, from **the root** of the repository,
```bash
wget https://github.com/metavoiceio/metavoice-src/raw/main/assets/bria.mp3
python scripts/mimi_test.py
```
Similary, Moshi can be tested (with a GPU) with
```bash
python scripts/moshi_benchmark.py
```
## License
The present code is provided under the MIT license.
Note that parts of this code is based on [AudioCraft](https://github.com/facebookresearch/audiocraft), released under
the MIT license.
## Citation
If you use either Mimi or Moshi, please cite the following paper,
```
@article{defossez2024moshi,
title={Moshi: a speech-text foundation model for real-time dialogue},
author={Alexandre Défossez and Laurent Mazaré and Manu Orsini and Amélie Royer and Patrick Pérez and Hervé Jégou and Edouard Grave and Neil Zeghidour},
journal={arXiv:TBC},
year={2024},
}
```
[moshi]: https://arxiv.org/

15
moshi/moshi/testing.md
View file

@ -1,15 +0,0 @@
# Testing
In order to test the audio tokenizer, you can run the following command.
```bash
wget https://github.com/metavoiceio/metavoice-src/raw/main/assets/bria.mp3
PYTHONPATH=. python scripts/mimi_test.py --weights tokenizer-e351c8d8-checkpoint125.safetensors
```
In order to test moshi, run the following.
```bash
PYTHONPATH=. python scripts/moshi_test.py \
--mimi-weights tokenizer-e351c8d8-checkpoint125.safetensors \
--tokenizer tokenizer_spm_32k_3.model \
--moshi-weights moshiko_pt_301e30bf@120.safetensors
```

4
moshi/setup.cfg
View file

@ -4,3 +4,7 @@ max-line-length = 120
[flake8]
max-line-length = 120
ignore = E203,E704
exclude =
dist
build

5
moshi_mlx/MANIFEST.in Normal file
View file

@ -0,0 +1,5 @@
include LICENSE*
include *.md
include *.cfg
include requirements.txt
include moshi_mlx/py.typed

57
moshi_mlx/README.md Normal file
View file

@ -0,0 +1,57 @@
# Moshi - MLX
See the [top-level README.md][main_repo] for more information on Moshi.
[Moshi][moshi] is a speech-text foundation model and full-duplex spoken dialogue framework.
It uses [Mimi][moshi], a state-of-the-art streaming neural audio codec. Mimi operates at 12.5 Hz, and compress
audio down to 1.1 kbps, in a fully streaming manner (latency of 80ms, the frame size), yet performs better than existing, non-streaming, codec.
This is the MLX implementation for Moshi. For Mimi, this uses our Rust based implementation through the Python binding provided in `rustymimi`, available in the [rust/](https://github.com/kyutai-labs/moshi/tree/main/rust) folder of our main repository.
## Requirements
You will need at least Python 3.10.
```bash
pip install moshi_mlx # moshi MLX, from PyPI
# Or the bleeding edge versions for Moshi and Moshi-MLX.
pip install -e "git+https://git@github.com/kyutai-labs/moshi#egg=moshi_mlx&subdirectory=moshi_mlx"
```
We have tested the MLX version with MacBook Pro M3.
## Usage
Then the model can be run with:
```bash
python -m moshi_mlx.local -q 4 # weights quantized to 4 bits
python -m moshi_mlx.local -q 8 # weights quantized to 8 bits
```
This uses a command line interface, which is bare bone. It doesn't do any echo cancellation,
nor does it try to compensate for a growing lag by skipping frames.
Alternatively you can use `python -m moshi_mlx.local_web` to use
the web UI, connection is via http on [localhost:8998](http://localhost:8998).
## License
The present code is provided under the MIT license.
## Citation
If you use either Mimi or Moshi, please cite the following paper,
```
@article{defossez2024moshi,
title={Moshi: a speech-text foundation model for real-time dialogue},
author={Alexandre Défossez and Laurent Mazaré and Manu Orsini and Amélie Royer and Patrick Pérez and Hervé Jégou and Edouard Grave and Neil Zeghidour},
journal={arXiv:TBC},
year={2024},
}
```
[moshi]: https://arxiv.org/
[main_repo]: https://github.com/kyutai-labs/moshi

0
moshi_mlx/moshi_mlx/py.typed Normal file
View file

4
moshi_mlx/setup.cfg
View file

@ -4,3 +4,7 @@ max-line-length = 120
[flake8]
max-line-length = 120
ignore = E203,E704
exclude =
dist
build

201
rust/LICENSE Normal file
View file

@ -0,0 +1,201 @@
Apache License
Version 2.0, January 2004
http://www.apache.org/licenses/
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
1. Definitions.
"License" shall mean the terms and conditions for use, reproduction,
and distribution as defined by Sections 1 through 9 of this document.
"Licensor" shall mean the copyright owner or entity authorized by
the copyright owner that is granting the License.
"Legal Entity" shall mean the union of the acting entity and all
other entities that control, are controlled by, or are under common
control with that entity. For the purposes of this definition,
"control" means (i) the power, direct or indirect, to cause the
direction or management of such entity, whether by contract or
otherwise, or (ii) ownership of fifty percent (50%) or more of the
outstanding shares, or (iii) beneficial ownership of such entity.
"You" (or "Your") shall mean an individual or Legal Entity
exercising permissions granted by this License.
"Source" form shall mean the preferred form for making modifications,
including but not limited to software source code, documentation
source, and configuration files.
"Object" form shall mean any form resulting from mechanical
transformation or translation of a Source form, including but
not limited to compiled object code, generated documentation,
and conversions to other media types.
"Work" shall mean the work of authorship, whether in Source or
Object form, made available under the License, as indicated by a
copyright notice that is included in or attached to the work
(an example is provided in the Appendix below).
"Derivative Works" shall mean any work, whether in Source or Object
form, that is based on (or derived from) the Work and for which the
editorial revisions, annotations, elaborations, or other modifications
represent, as a whole, an original work of authorship. For the purposes
of this License, Derivative Works shall not include works that remain
separable from, or merely link (or bind by name) to the interfaces of,
the Work and Derivative Works thereof.
"Contribution" shall mean any work of authorship, including
the original version of the Work and any modifications or additions
to that Work or Derivative Works thereof, that is intentionally
submitted to Licensor for inclusion in the Work by the copyright owner
or by an individual or Legal Entity authorized to submit on behalf of
the copyright owner. For the purposes of this definition, "submitted"
means any form of electronic, verbal, or written communication sent
to the Licensor or its representatives, including but not limited to
communication on electronic mailing lists, source code control systems,
and issue tracking systems that are managed by, or on behalf of, the
Licensor for the purpose of discussing and improving the Work, but
excluding communication that is conspicuously marked or otherwise
designated in writing by the copyright owner as "Not a Contribution."
"Contributor" shall mean Licensor and any individual or Legal Entity
on behalf of whom a Contribution has been received by Licensor and
subsequently incorporated within the Work.
2. Grant of Copyright License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
copyright license to reproduce, prepare Derivative Works of,
publicly display, publicly perform, sublicense, and distribute the
Work and such Derivative Works in Source or Object form.
3. Grant of Patent License. Subject to the terms and conditions of
this License, each Contributor hereby grants to You a perpetual,
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
(except as stated in this section) patent license to make, have made,
use, offer to sell, sell, import, and otherwise transfer the Work,
where such license applies only to those patent claims licensable
by such Contributor that are necessarily infringed by their
Contribution(s) alone or by combination of their Contribution(s)
with the Work to which such Contribution(s) was submitted. If You
institute patent litigation against any entity (including a
cross-claim or counterclaim in a lawsuit) alleging that the Work
or a Contribution incorporated within the Work constitutes direct
or contributory patent infringement, then any patent licenses
granted to You under this License for that Work shall terminate
as of the date such litigation is filed.
4. Redistribution. You may reproduce and distribute copies of the
Work or Derivative Works thereof in any medium, with or without
modifications, and in Source or Object form, provided that You
meet the following conditions:
(a) You must give any other recipients of the Work or
Derivative Works a copy of this License; and
(b) You must cause any modified files to carry prominent notices
stating that You changed the files; and
(c) You must retain, in the Source form of any Derivative Works
that You distribute, all copyright, patent, trademark, and
attribution notices from the Source form of the Work,
excluding those notices that do not pertain to any part of
the Derivative Works; and
(d) If the Work includes a "NOTICE" text file as part of its
distribution, then any Derivative Works that You distribute must
include a readable copy of the attribution notices contained
within such NOTICE file, excluding those notices that do not
pertain to any part of the Derivative Works, in at least one
of the following places: within a NOTICE text file distributed
as part of the Derivative Works; within the Source form or
documentation, if provided along with the Derivative Works; or,
within a display generated by the Derivative Works, if and
wherever such third-party notices normally appear. The contents
of the NOTICE file are for informational purposes only and
do not modify the License. You may add Your own attribution
notices within Derivative Works that You distribute, alongside
or as an addendum to the NOTICE text from the Work, provided
that such additional attribution notices cannot be construed
as modifying the License.
You may add Your own copyright statement to Your modifications and
may provide additional or different license terms and conditions
for use, reproduction, or distribution of Your modifications, or
for any such Derivative Works as a whole, provided Your use,
reproduction, and distribution of the Work otherwise complies with
the conditions stated in this License.
5. Submission of Contributions. Unless You explicitly state otherwise,
any Contribution intentionally submitted for inclusion in the Work
by You to the Licensor shall be under the terms and conditions of
this License, without any additional terms or conditions.
Notwithstanding the above, nothing herein shall supersede or modify
the terms of any separate license agreement you may have executed
with Licensor regarding such Contributions.
6. Trademarks. This License does not grant permission to use the trade
names, trademarks, service marks, or product names of the Licensor,
except as required for reasonable and customary use in describing the
origin of the Work and reproducing the content of the NOTICE file.
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.
9. Accepting Warranty or Additional Liability. While redistributing
the Work or Derivative Works thereof, You may choose to offer,
and charge a fee for, acceptance of support, warranty, indemnity,
or other liability obligations and/or rights consistent with this
License. However, in accepting such obligations, You may act only
on Your own behalf and on Your sole responsibility, not on behalf
of any other Contributor, and only if You agree to indemnify,
defend, and hold each Contributor harmless for any liability
incurred by, or claims asserted against, such Contributor by reason
of your accepting any such warranty or additional liability.
END OF TERMS AND CONDITIONS
APPENDIX: How to apply the Apache License to your work.
To apply the Apache License to your work, attach the following
boilerplate notice, with the fields enclosed by brackets "[]"
replaced with your own identifying information. (Don't include
the brackets!) The text should be enclosed in the appropriate
comment syntax for the file format. We also recommend that a
file or class name and description of purpose be included on the
same "printed page" as the copyright notice for easier
identification within third-party archives.
Copyright [yyyy] [name of copyright owner]
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

62
rust/README.md Normal file
View file

@ -0,0 +1,62 @@
# moshi - rust
See the [top-level README.md](../README.md) for more information.
This provides the Rust backend (both Mimi and Moshi) and client implementation.
The Mimi implementation is available through Python bindings, through the `rustymimi` package.
## Requirements
You will need a recent version of the [Rust toolchain](https://rustup.rs/).
## Rust based Mimi with Python bindings
First, a standalone rust based implementation of Mimi is provided, along with Python bindings.
This is the one used by `moshi_mlx`. It is automatically installed with `moshi_mlx`, but you
can install it separately as
```bash
# Install from pip:
pip install rustymimi==0.1.1
# Alternatively, if you want to compile the package run from the root of the repo.
maturin dev -r -m rust/mimi-pyo3/Cargo.toml
```
## Rust server
In order to run the rust inference server, use the following command from within
the this directory:
```bash
cargo run --features cuda --bin moshi-backend -r -- --config moshi-backend/config.json standalone
```
When using macOS, you can replace `--features cuda` with `--features metal`.
Alternatively you can use `config-q8.json` rather than `config.json` to use the
quantified q8 model.
Once the server has printed 'standalone worker listening', you can use the web
UI. By default the rust version uses https so it will be at
[localhost:8998](https://localhost:8998).
You will get some warnings about the site being unsafe. When using chrome you
can bypass it by selecting "Details" or "Advanced", then "Visit this unsafe
site" or "Proceed to localhost (unsafe)".
## Rust client
We recommend using the web UI as it provides some echo cancellation that helps
the overall model quality. Alternatively we provide some command line interfaces
for the rust and python versions, the protocol is the same as with the web UI so
there is nothing to change on the server side.
### Rust Command Line
From within the `rust` directory, run the following:
```bash
cargo run --bin moshi-cli -r -- tui --host localhost
```
## License
The present code is provided under the Apache license.

12
scripts/run_ci_when_installed.sh Executable file
View file

@ -0,0 +1,12 @@
#!/bin/bash
# This script is used to detect if moshi or moshi_mlx are installed, and run
# their CI only in that case!
package=$1
if python -c "from $package import models"; then
# package is installed, let's run the command
eval $2
else
echo "Package $package not installed, skipping the CI for it."
fi