Upload 98 files

CONTRIBUTING.md

@@ -0,0 +1,91 @@
+# Contributing
+
+The Open-Sora project welcomes any constructive contribution from the community and the team is more than willing to work on problems you have encountered to make it a better project.
+
+## Development Environment Setup
+
+To contribute to Open-Sora, we would like to first guide you to set up a proper development environment so that you can better implement your code. You can install this library from source with the `editable` flag (`-e`, for development mode) so that your change to the source code will be reflected in runtime without re-installation. 
+
+You can refer to the [Installation Section](./README.md#installation) and replace `pip install -v .` with `pip install -v -e .`.
+
+
+### Code Style
+
+We have some static checks when you commit your code change, please make sure you can pass all the tests and make sure the coding style meets our requirements. We use pre-commit hook to make sure the code is aligned with the writing standard. To set up the code style checking, you need to follow the steps below.
+
+```shell
+# these commands are executed under the Open-Sora directory
+pip install pre-commit
+pre-commit install
+```
+
+Code format checking will be automatically executed when you commit your changes.
+
+
+## Contribution Guide
+
+You need to follow these steps below to make contribution to the main repository via pull request. You can learn about the details of pull request [here](https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/proposing-changes-to-your-work-with-pull-requests/about-pull-requests).
+
+### 1. Fork the Official Repository
+
+Firstly, you need to visit the [Open-Sora repository](https://github.com/hpcaitech/Open-Sora) and fork into your own account. The `fork` button is at the right top corner of the web page alongside with buttons such as `watch` and `star`.
+
+Now, you can clone your own forked repository into your local environment.
+
+```shell
+git clone https://github.com/<YOUR-USERNAME>/Open-Sora.git
+```
+
+### 2. Configure Git
+
+You need to set the official repository as your upstream so that you can synchronize with the latest update in the official repository. You can learn about upstream [here](https://www.atlassian.com/git/tutorials/git-forks-and-upstreams).
+
+Then add the original repository as upstream
+
+```shell
+cd Open-Sora
+git remote add upstream https://github.com/hpcaitech/Open-Sora.git
+```
+
+you can use the following command to verify that the remote is set. You should see both `origin` and `upstream` in the output.
+
+```shell
+git remote -v
+```
+
+### 3. Synchronize with Official Repository
+
+Before you make changes to the codebase, it is always good to fetch the latest updates in the official repository. In order to do so, you can use the commands below.
+
+```shell
+git fetch upstream
+git checkout main
+git merge upstream/main
+git push origin main
+```
+
+### 5. Create a New Branch
+
+You should not make changes to the `main` branch of your forked repository as this might make upstream synchronization difficult. You can create a new branch with the appropriate name. General branch name format should start with `hotfix/` and `feature/`. `hotfix` is for bug fix and `feature` is for addition of a new feature.
+
+
+```shell
+git checkout -b <NEW-BRANCH-NAME>
+```
+
+### 6. Implementation and Code Commit
+
+Now you can implement your code change in the source code. Remember that you installed the system in development, thus you do not need to uninstall and install to make the code take effect. The code change will be reflected in every new PyThon execution.
+You can commit and push the changes to your local repository. The changes should be kept logical, modular and atomic.
+
+```shell
+git add -A
+git commit -m "<COMMIT-MESSAGE>"
+git push -u origin <NEW-BRANCH-NAME>
+```
+
+### 7. Open a Pull Request
+
+You can now create a pull request on the GitHub webpage of your repository. The source branch is `<NEW-BRANCH-NAME>` of your repository and the target branch should be `main` of `hpcaitech/Open-Sora`. After creating this pull request, you should be able to see it [here](https://github.com/hpcaitech/Open-Sora/pulls).
+
+The Open-Sora team will review your code change and merge your code if applicable.

LICENSE

@@ -0,0 +1,681 @@
+                                 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.
+
+   =========================================================================
+   This project is inspired by the listed projects and is subject to the following licenses:
+
+   1. Latte (https://github.com/Vchitect/Latte/blob/main/LICENSE)
+
+   Copyright 2024 Latte
+
+   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.
+
+   2. PixArt-alpha (https://github.com/PixArt-alpha/PixArt-alpha/blob/master/LICENSE)
+
+   Copyright (C) 2024 PixArt-alpha/PixArt-alpha
+
+   This program is free software: you can redistribute it and/or modify
+   it under the terms of the GNU Affero General Public License as published
+   by the Free Software Foundation, either version 3 of the License, or
+   (at your option) any later version.
+
+   This program is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+   GNU Affero General Public License for more details.
+
+   You should have received a copy of the GNU Affero General Public License
+   along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+   3. dpm-solver (https://github.com/LuChengTHU/dpm-solver/blob/main/LICENSE)
+
+   MIT License
+
+   Copyright (c) 2022 Cheng Lu
+
+   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. DiT (https://github.com/facebookresearch/DiT/blob/main/LICENSE.txt)
+
+   Attribution-NonCommercial 4.0 International
+
+   =======================================================================
+
+   Creative Commons Corporation ("Creative Commons") is not a law firm and
+   does not provide legal services or legal advice. Distribution of
+   Creative Commons public licenses does not create a lawyer-client or
+   other relationship. Creative Commons makes its licenses and related
+   information available on an "as-is" basis. Creative Commons gives no
+   warranties regarding its licenses, any material licensed under their
+   terms and conditions, or any related information. Creative Commons
+   disclaims all liability for damages resulting from their use to the
+   fullest extent possible.
+
+   Using Creative Commons Public Licenses
+
+   Creative Commons public licenses provide a standard set of terms and
+   conditions that creators and other rights holders may use to share
+   original works of authorship and other material subject to copyright
+   and certain other rights specified in the public license below. The
+   following considerations are for informational purposes only, are not
+   exhaustive, and do not form part of our licenses.
+
+      Considerations for licensors: Our public licenses are
+      intended for use by those authorized to give the public
+      permission to use material in ways otherwise restricted by
+      copyright and certain other rights. Our licenses are
+      irrevocable. Licensors should read and understand the terms
+      and conditions of the license they choose before applying it.
+      Licensors should also secure all rights necessary before
+      applying our licenses so that the public can reuse the
+      material as expected. Licensors should clearly mark any
+      material not subject to the license. This includes other CC-
+      licensed material, or material used under an exception or
+      limitation to copyright. More considerations for licensors:
+      wiki.creativecommons.org/Considerations_for_licensors
+
+      Considerations for the public: By using one of our public
+      licenses, a licensor grants the public permission to use the
+      licensed material under specified terms and conditions. If
+      the licensor's permission is not necessary for any reason--for
+      example, because of any applicable exception or limitation to
+      copyright--then that use is not regulated by the license. Our
+      licenses grant only permissions under copyright and certain
+      other rights that a licensor has authority to grant. Use of
+      the licensed material may still be restricted for other
+      reasons, including because others have copyright or other
+      rights in the material. A licensor may make special requests,
+      such as asking that all changes be marked or described.
+      Although not required by our licenses, you are encouraged to
+      respect those requests where reasonable. More_considerations
+      for the public: 
+      wiki.creativecommons.org/Considerations_for_licensees
+
+   =======================================================================
+
+   Creative Commons Attribution-NonCommercial 4.0 International Public
+   License
+
+   By exercising the Licensed Rights (defined below), You accept and agree
+   to be bound by the terms and conditions of this Creative Commons
+   Attribution-NonCommercial 4.0 International Public License ("Public
+   License"). To the extent this Public License may be interpreted as a
+   contract, You are granted the Licensed Rights in consideration of Your
+   acceptance of these terms and conditions, and the Licensor grants You
+   such rights in consideration of benefits the Licensor receives from
+   making the Licensed Material available under these terms and
+   conditions.
+
+   Section 1 -- Definitions.
+
+   a. Adapted Material means material subject to Copyright and Similar
+      Rights that is derived from or based upon the Licensed Material
+      and in which the Licensed Material is translated, altered,
+      arranged, transformed, or otherwise modified in a manner requiring
+      permission under the Copyright and Similar Rights held by the
+      Licensor. For purposes of this Public License, where the Licensed
+      Material is a musical work, performance, or sound recording,
+      Adapted Material is always produced where the Licensed Material is
+      synched in timed relation with a moving image.
+
+   b. Adapter's License means the license You apply to Your Copyright
+      and Similar Rights in Your contributions to Adapted Material in
+      accordance with the terms and conditions of this Public License.
+
+   c. Copyright and Similar Rights means copyright and/or similar rights
+      closely related to copyright including, without limitation,
+      performance, broadcast, sound recording, and Sui Generis Database
+      Rights, without regard to how the rights are labeled or
+      categorized. For purposes of this Public License, the rights
+      specified in Section 2(b)(1)-(2) are not Copyright and Similar
+      Rights.
+   d. Effective Technological Measures means those measures that, in the
+      absence of proper authority, may not be circumvented under laws
+      fulfilling obligations under Article 11 of the WIPO Copyright
+      Treaty adopted on December 20, 1996, and/or similar international
+      agreements.
+
+   e. Exceptions and Limitations means fair use, fair dealing, and/or
+      any other exception or limitation to Copyright and Similar Rights
+      that applies to Your use of the Licensed Material.
+
+   f. Licensed Material means the artistic or literary work, database,
+      or other material to which the Licensor applied this Public
+      License.
+
+   g. Licensed Rights means the rights granted to You subject to the
+      terms and conditions of this Public License, which are limited to
+      all Copyright and Similar Rights that apply to Your use of the
+      Licensed Material and that the Licensor has authority to license.
+
+   h. Licensor means the individual(s) or entity(ies) granting rights
+      under this Public License.
+
+   i. NonCommercial means not primarily intended for or directed towards
+      commercial advantage or monetary compensation. For purposes of
+      this Public License, the exchange of the Licensed Material for
+      other material subject to Copyright and Similar Rights by digital
+      file-sharing or similar means is NonCommercial provided there is
+      no payment of monetary compensation in connection with the
+      exchange.
+
+   j. Share means to provide material to the public by any means or
+      process that requires permission under the Licensed Rights, such
+      as reproduction, public display, public performance, distribution,
+      dissemination, communication, or importation, and to make material
+      available to the public including in ways that members of the
+      public may access the material from a place and at a time
+      individually chosen by them.
+
+   k. Sui Generis Database Rights means rights other than copyright
+      resulting from Directive 96/9/EC of the European Parliament and of
+      the Council of 11 March 1996 on the legal protection of databases,
+      as amended and/or succeeded, as well as other essentially
+      equivalent rights anywhere in the world.
+
+   l. You means the individual or entity exercising the Licensed Rights
+      under this Public License. Your has a corresponding meaning.
+
+   Section 2 -- Scope.
+
+   a. License grant.
+
+         1. Subject to the terms and conditions of this Public License,
+            the Licensor hereby grants You a worldwide, royalty-free,
+            non-sublicensable, non-exclusive, irrevocable license to
+            exercise the Licensed Rights in the Licensed Material to:
+
+               a. reproduce and Share the Licensed Material, in whole or
+                  in part, for NonCommercial purposes only; and
+
+               b. produce, reproduce, and Share Adapted Material for
+                  NonCommercial purposes only.
+
+         2. Exceptions and Limitations. For the avoidance of doubt, where
+            Exceptions and Limitations apply to Your use, this Public
+            License does not apply, and You do not need to comply with
+            its terms and conditions.
+
+         3. Term. The term of this Public License is specified in Section
+            6(a).
+
+         4. Media and formats; technical modifications allowed. The
+            Licensor authorizes You to exercise the Licensed Rights in
+            all media and formats whether now known or hereafter created,
+            and to make technical modifications necessary to do so. The
+            Licensor waives and/or agrees not to assert any right or
+            authority to forbid You from making technical modifications
+            necessary to exercise the Licensed Rights, including
+            technical modifications necessary to circumvent Effective
+            Technological Measures. For purposes of this Public License,
+            simply making modifications authorized by this Section 2(a)
+            (4) never produces Adapted Material.
+
+         5. Downstream recipients.
+
+               a. Offer from the Licensor -- Licensed Material. Every
+                  recipient of the Licensed Material automatically
+                  receives an offer from the Licensor to exercise the
+                  Licensed Rights under the terms and conditions of this
+                  Public License.
+
+               b. No downstream restrictions. You may not offer or impose
+                  any additional or different terms or conditions on, or
+                  apply any Effective Technological Measures to, the
+                  Licensed Material if doing so restricts exercise of the
+                  Licensed Rights by any recipient of the Licensed
+                  Material.
+
+         6. No endorsement. Nothing in this Public License constitutes or
+            may be construed as permission to assert or imply that You
+            are, or that Your use of the Licensed Material is, connected
+            with, or sponsored, endorsed, or granted official status by,
+            the Licensor or others designated to receive attribution as
+            provided in Section 3(a)(1)(A)(i).
+
+   b. Other rights.
+
+         1. Moral rights, such as the right of integrity, are not
+            licensed under this Public License, nor are publicity,
+            privacy, and/or other similar personality rights; however, to
+            the extent possible, the Licensor waives and/or agrees not to
+            assert any such rights held by the Licensor to the limited
+            extent necessary to allow You to exercise the Licensed
+            Rights, but not otherwise.
+
+         2. Patent and trademark rights are not licensed under this
+            Public License.
+
+         3. To the extent possible, the Licensor waives any right to
+            collect royalties from You for the exercise of the Licensed
+            Rights, whether directly or through a collecting society
+            under any voluntary or waivable statutory or compulsory
+            licensing scheme. In all other cases the Licensor expressly
+            reserves any right to collect such royalties, including when
+            the Licensed Material is used other than for NonCommercial
+            purposes.
+
+   Section 3 -- License Conditions.
+
+   Your exercise of the Licensed Rights is expressly made subject to the
+   following conditions.
+
+   a. Attribution.
+
+         1. If You Share the Licensed Material (including in modified
+            form), You must:
+
+               a. retain the following if it is supplied by the Licensor
+                  with the Licensed Material:
+
+                  i. identification of the creator(s) of the Licensed
+                     Material and any others designated to receive
+                     attribution, in any reasonable manner requested by
+                     the Licensor (including by pseudonym if
+                     designated);
+
+                  ii. a copyright notice;
+
+                  iii. a notice that refers to this Public License;
+
+                  iv. a notice that refers to the disclaimer of
+                     warranties;
+
+                  v. a URI or hyperlink to the Licensed Material to the
+                     extent reasonably practicable;
+
+               b. indicate if You modified the Licensed Material and
+                  retain an indication of any previous modifications; and
+
+               c. indicate the Licensed Material is licensed under this
+                  Public License, and include the text of, or the URI or
+                  hyperlink to, this Public License.
+
+         2. You may satisfy the conditions in Section 3(a)(1) in any
+            reasonable manner based on the medium, means, and context in
+            which You Share the Licensed Material. For example, it may be
+            reasonable to satisfy the conditions by providing a URI or
+            hyperlink to a resource that includes the required
+            information.
+
+         3. If requested by the Licensor, You must remove any of the
+            information required by Section 3(a)(1)(A) to the extent
+            reasonably practicable.
+
+         4. If You Share Adapted Material You produce, the Adapter's
+            License You apply must not prevent recipients of the Adapted
+            Material from complying with this Public License.
+
+   Section 4 -- Sui Generis Database Rights.
+
+   Where the Licensed Rights include Sui Generis Database Rights that
+   apply to Your use of the Licensed Material:
+
+   a. for the avoidance of doubt, Section 2(a)(1) grants You the right
+      to extract, reuse, reproduce, and Share all or a substantial
+      portion of the contents of the database for NonCommercial purposes
+      only;
+
+   b. if You include all or a substantial portion of the database
+      contents in a database in which You have Sui Generis Database
+      Rights, then the database in which You have Sui Generis Database
+      Rights (but not its individual contents) is Adapted Material; and
+
+   c. You must comply with the conditions in Section 3(a) if You Share
+      all or a substantial portion of the contents of the database.
+
+   For the avoidance of doubt, this Section 4 supplements and does not
+   replace Your obligations under this Public License where the Licensed
+   Rights include other Copyright and Similar Rights.
+
+   Section 5 -- Disclaimer of Warranties and Limitation of Liability.
+
+   a. UNLESS OTHERWISE SEPARATELY UNDERTAKEN BY THE LICENSOR, TO THE
+      EXTENT POSSIBLE, THE LICENSOR OFFERS THE LICENSED MATERIAL AS-IS
+      AND AS-AVAILABLE, AND MAKES NO REPRESENTATIONS OR WARRANTIES OF
+      ANY KIND CONCERNING THE LICENSED MATERIAL, WHETHER EXPRESS,
+      IMPLIED, STATUTORY, OR OTHER. THIS INCLUDES, WITHOUT LIMITATION,
+      WARRANTIES OF TITLE, MERCHANTABILITY, FITNESS FOR A PARTICULAR
+      PURPOSE, NON-INFRINGEMENT, ABSENCE OF LATENT OR OTHER DEFECTS,
+      ACCURACY, OR THE PRESENCE OR ABSENCE OF ERRORS, WHETHER OR NOT
+      KNOWN OR DISCOVERABLE. WHERE DISCLAIMERS OF WARRANTIES ARE NOT
+      ALLOWED IN FULL OR IN PART, THIS DISCLAIMER MAY NOT APPLY TO YOU.
+
+   b. TO THE EXTENT POSSIBLE, IN NO EVENT WILL THE LICENSOR BE LIABLE
+      TO YOU ON ANY LEGAL THEORY (INCLUDING, WITHOUT LIMITATION,
+      NEGLIGENCE) OR OTHERWISE FOR ANY DIRECT, SPECIAL, INDIRECT,
+      INCIDENTAL, CONSEQUENTIAL, PUNITIVE, EXEMPLARY, OR OTHER LOSSES,
+      COSTS, EXPENSES, OR DAMAGES ARISING OUT OF THIS PUBLIC LICENSE OR
+      USE OF THE LICENSED MATERIAL, EVEN IF THE LICENSOR HAS BEEN
+      ADVISED OF THE POSSIBILITY OF SUCH LOSSES, COSTS, EXPENSES, OR
+      DAMAGES. WHERE A LIMITATION OF LIABILITY IS NOT ALLOWED IN FULL OR
+      IN PART, THIS LIMITATION MAY NOT APPLY TO YOU.
+
+   c. The disclaimer of warranties and limitation of liability provided
+      above shall be interpreted in a manner that, to the extent
+      possible, most closely approximates an absolute disclaimer and
+      waiver of all liability.
+
+   Section 6 -- Term and Termination.
+
+   a. This Public License applies for the term of the Copyright and
+      Similar Rights licensed here. However, if You fail to comply with
+      this Public License, then Your rights under this Public License
+      terminate automatically.
+
+   b. Where Your right to use the Licensed Material has terminated under
+      Section 6(a), it reinstates:
+
+         1. automatically as of the date the violation is cured, provided
+            it is cured within 30 days of Your discovery of the
+            violation; or
+
+         2. upon express reinstatement by the Licensor.
+
+      For the avoidance of doubt, this Section 6(b) does not affect any
+      right the Licensor may have to seek remedies for Your violations
+      of this Public License.
+
+   c. For the avoidance of doubt, the Licensor may also offer the
+      Licensed Material under separate terms or conditions or stop
+      distributing the Licensed Material at any time; however, doing so
+      will not terminate this Public License.
+
+   d. Sections 1, 5, 6, 7, and 8 survive termination of this Public
+      License.
+
+   Section 7 -- Other Terms and Conditions.
+
+   a. The Licensor shall not be bound by any additional or different
+      terms or conditions communicated by You unless expressly agreed.
+
+   b. Any arrangements, understandings, or agreements regarding the
+      Licensed Material not stated herein are separate from and
+      independent of the terms and conditions of this Public License.
+
+   Section 8 -- Interpretation.
+
+   a. For the avoidance of doubt, this Public License does not, and
+      shall not be interpreted to, reduce, limit, restrict, or impose
+      conditions on any use of the Licensed Material that could lawfully
+      be made without permission under this Public License.
+
+   b. To the extent possible, if any provision of this Public License is
+      deemed unenforceable, it shall be automatically reformed to the
+      minimum extent necessary to make it enforceable. If the provision
+      cannot be reformed, it shall be severed from this Public License
+      without affecting the enforceability of the remaining terms and
+      conditions.
+
+   c. No term or condition of this Public License will be waived and no
+      failure to comply consented to unless expressly agreed to by the
+      Licensor.
+
+   d. Nothing in this Public License constitutes or may be interpreted
+      as a limitation upon, or waiver of, any privileges and immunities
+      that apply to the Licensor or You, including from the legal
+      processes of any jurisdiction or authority.
+
+   =======================================================================
+
+   Creative Commons is not a party to its public
+   licenses. Notwithstanding, Creative Commons may elect to apply one of
+   its public licenses to material it publishes and in those instances
+   will be considered the “Licensor.” The text of the Creative Commons
+   public licenses is dedicated to the public domain under the CC0 Public
+   Domain Dedication. Except for the limited purpose of indicating that
+   material is shared under a Creative Commons public license or as
+   otherwise permitted by the Creative Commons policies published at
+   creativecommons.org/policies, Creative Commons does not authorize the
+   use of the trademark "Creative Commons" or any other trademark or logo
+   of Creative Commons without its prior written consent including,
+   without limitation, in connection with any unauthorized modifications
+   to any of its public licenses or any other arrangements,
+   understandings, or agreements concerning use of licensed material. For
+   the avoidance of doubt, this paragraph does not form part of the
+   public licenses.
+
+   Creative Commons may be contacted at creativecommons.org.
+
+   5. OpenDiT (https://github.com/NUS-HPC-AI-Lab/OpenDiT/blob/master/LICENSE)
+
+   Copyright OpenDiT
+
+   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.
+
+

Open-Sora/.gitattributes

@@ -0,0 +1,35 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text

Open-Sora/README.md

@@ -0,0 +1,13 @@
+---
+title: Open Sora
+emoji: 📚
+colorFrom: yellow
+colorTo: indigo
+sdk: gradio
+sdk_version: 4.21.0
+app_file: app.py
+pinned: false
+license: apache-2.0
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Open-Sora/requirements.txt

@@ -0,0 +1,12 @@
+torch
+torchvision
+
+packaging
+ninja
+flash-attn --no-build-isolation
+
+-v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" git+https://github.com/NVIDIA/apex.git
+
+-U xformers --index-url https://download.pytorch.org/whl/cu121
+
+-e git+https://github.com/hpcaitech/Open-Sora.git#egg=open-sora

configs/dit/inference/16x256x256.py

@@ -0,0 +1,31 @@
+num_frames = 16
+fps = 8
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="DiT-XL/2",
+    condition="text",
+    from_pretrained="PRETRAINED_MODEL",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="clip",
+    from_pretrained="openai/clip-vit-base-patch32",
+    model_max_length=77,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=4.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/ucf101_labels.txt"
+save_dir = "./outputs/samples/"

configs/dit/inference/1x256x256-class.py

@@ -0,0 +1,31 @@
+num_frames = 1
+fps = 1
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="DiT-XL/2",
+    no_temporal_pos_emb=True,
+    condition="label_1000",
+    from_pretrained="DiT-XL-2-256x256.pt",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="classes",
+    num_classes=1000,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=4.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/imagenet_id.txt"
+save_dir = "./outputs/samples/"

configs/dit/inference/1x256x256.py

@@ -0,0 +1,32 @@
+num_frames = 1
+fps = 1
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="DiT-XL/2",
+    no_temporal_pos_emb=True,
+    condition="text",
+    from_pretrained="PRETRAINED_MODEL",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="clip",
+    from_pretrained="openai/clip-vit-base-patch32",
+    model_max_length=77,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=4.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/imagenet_labels.txt"
+save_dir = "./outputs/samples/"

configs/dit/train/16x256x256.py

@@ -0,0 +1,50 @@
+num_frames = 16
+frame_interval = 3
+image_size = (256, 256)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = False
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="DiT-XL/2",
+    from_pretrained="DiT-XL-2-256x256.pt",
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="clip",
+    from_pretrained="openai/clip-vit-base-patch32",
+    model_max_length=77,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 1000
+load = None
+
+batch_size = 8
+lr = 2e-5
+grad_clip = 1.0

configs/dit/train/1x256x256.py

@@ -0,0 +1,50 @@
+num_frames = 1
+frame_interval = 1
+image_size = (256, 256)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = True
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = False
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="DiT-XL/2",
+    no_temporal_pos_emb=True,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="clip",
+    from_pretrained="openai/clip-vit-base-patch32",
+    model_max_length=77,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 1000
+load = None
+
+batch_size = 128
+lr = 1e-4  # according to DiT repo
+grad_clip = 1.0

configs/latte/inference/16x256x256-class.py

@@ -0,0 +1,30 @@
+num_frames = 16
+fps = 8
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="Latte-XL/2",
+    condition="label_101",
+    from_pretrained="Latte-XL-2-256x256-ucf101.pt",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="classes",
+    num_classes=101,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=4.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/ucf101_id.txt"
+save_dir = "./outputs/samples/"

configs/latte/inference/16x256x256.py

@@ -0,0 +1,31 @@
+num_frames = 16
+fps = 8
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="Latte-XL/2",
+    condition="text",
+    from_pretrained="PRETRAINED_MODEL",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="clip",
+    from_pretrained="openai/clip-vit-base-patch32",
+    model_max_length=77,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=4.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/ucf101_labels.txt"
+save_dir = "./outputs/samples/"

configs/latte/train/16x256x256.py

@@ -0,0 +1,49 @@
+num_frames = 16
+frame_interval = 3
+image_size = (256, 256)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = True
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="Latte-XL/2",
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="clip",
+    from_pretrained="openai/clip-vit-base-patch32",
+    model_max_length=77,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 1000
+load = None
+
+batch_size = 8
+lr = 2e-5
+grad_clip = 1.0

configs/opensora/inference/16x256x256.py

@@ -0,0 +1,34 @@
+num_frames = 16
+fps = 24 // 3
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=0.5,
+    time_scale=1.0,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+    from_pretrained="PRETRAINED_MODEL",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+)
+scheduler = dict(
+    type="iddpm",
+    num_sampling_steps=100,
+    cfg_scale=7.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/t2v_samples.txt"
+save_dir = "./outputs/samples/"

configs/opensora/inference/16x512x512.py

@@ -0,0 +1,35 @@
+num_frames = 16
+fps = 24 // 3
+image_size = (512, 512)
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=1.0,
+    time_scale=1.0,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+    from_pretrained="PRETRAINED_MODEL"
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+    micro_batch_size=128,
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+)
+scheduler = dict(
+    type="iddpm",
+    num_sampling_steps=100,
+    cfg_scale=7.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/t2v_samples.txt"
+save_dir = "./outputs/samples/"

configs/opensora/inference/64x512x512.py

@@ -0,0 +1,35 @@
+num_frames = 64
+fps = 24 // 2
+image_size = (512, 512)
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=1.0,
+    time_scale=2 / 3,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+    from_pretrained="PRETRAINED_MODEL",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+    micro_batch_size=128,
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+)
+scheduler = dict(
+    type="iddpm",
+    num_sampling_steps=100,
+    cfg_scale=7.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 1
+seed = 42
+prompt_path = "./assets/texts/t2v_samples.txt"
+save_dir = "./outputs/samples/"

configs/opensora/train/16x256x256.py

@@ -0,0 +1,53 @@
+num_frames = 16
+frame_interval = 3
+image_size = (256, 256)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = True
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=0.5,
+    time_scale=1.0,
+    from_pretrained="PixArt-XL-2-512x512.pth",
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 1000
+load = None
+
+batch_size = 8
+lr = 2e-5
+grad_clip = 1.0

configs/opensora/train/16x512x512.py

@@ -0,0 +1,54 @@
+num_frames = 16
+frame_interval = 3
+image_size = (512, 512)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = False
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=1.0,
+    time_scale=1.0,
+    from_pretrained=None,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+    micro_batch_size=128,
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 500
+load = None
+
+batch_size = 8
+lr = 2e-5
+grad_clip = 1.0

configs/opensora/train/360x512x512.py

@@ -0,0 +1,55 @@
+num_frames = 360
+frame_interval = 1
+image_size = (512, 512)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = True
+plugin = "zero2-seq"
+sp_size = 2
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=1.0,
+    time_scale=2 / 3,
+    from_pretrained=None,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+    enable_sequence_parallelism=True,  # enable sq here
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+    micro_batch_size=128,
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 250
+load = None
+
+batch_size = 1
+lr = 2e-5
+grad_clip = 1.0

configs/opensora/train/64x512x512-sp.py

@@ -0,0 +1,54 @@
+num_frames = 64
+frame_interval = 2
+image_size = (512, 512)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = True
+plugin = "zero2-seq"
+sp_size = 2
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=1.0,
+    time_scale=2 / 3,
+    from_pretrained=None,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+    enable_sequence_parallelism=True,  # enable sq here
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 1000
+load = None
+
+batch_size = 1
+lr = 2e-5
+grad_clip = 1.0

configs/opensora/train/64x512x512.py

@@ -0,0 +1,54 @@
+num_frames = 64
+frame_interval = 2
+image_size = (512, 512)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = True
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=1.0,
+    time_scale=2 / 3,
+    from_pretrained=None,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+    micro_batch_size=64,
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 250
+load = None
+
+batch_size = 4
+lr = 2e-5
+grad_clip = 1.0

configs/pixart/inference/16x256x256.py

@@ -0,0 +1,32 @@
+num_frames = 16
+fps = 8
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="PixArt-XL/2",
+    space_scale=0.5,
+    time_scale=1.0,
+    from_pretrained="outputs/098-F16S3-PixArt-XL-2/epoch7-global_step30000/model_ckpt.pt",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=7.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/t2v_samples.txt"
+save_dir = "./outputs/samples/"

configs/pixart/inference/1x1024MS.py

@@ -0,0 +1,34 @@
+num_frames = 1
+fps = 1
+image_size = (1920, 512)
+multi_resolution = True
+
+# Define model
+model = dict(
+    type="PixArtMS-XL/2",
+    space_scale=2.0,
+    time_scale=1.0,
+    no_temporal_pos_emb=True,
+    from_pretrained="PixArt-XL-2-1024-MS.pth",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=7.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/t2i_samples.txt"
+save_dir = "./outputs/samples/"

configs/pixart/inference/1x256x256.py

@@ -0,0 +1,33 @@
+num_frames = 1
+fps = 1
+image_size = (256, 256)
+
+# Define model
+model = dict(
+    type="PixArt-XL/2",
+    space_scale=1.0,
+    time_scale=1.0,
+    no_temporal_pos_emb=True,
+    from_pretrained="PixArt-XL-2-256x256.pth",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=7.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/t2i_samples.txt"
+save_dir = "./outputs/samples/"

configs/pixart/inference/1x512x512.py

@@ -0,0 +1,33 @@
+num_frames = 1
+fps = 1
+image_size = (512, 512)
+
+# Define model
+model = dict(
+    type="PixArt-XL/2",
+    space_scale=1.0,
+    time_scale=1.0,
+    no_temporal_pos_emb=True,
+    from_pretrained="PixArt-XL-2-512x512.pth",
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+)
+scheduler = dict(
+    type="dpm-solver",
+    num_sampling_steps=20,
+    cfg_scale=7.0,
+)
+dtype = "fp16"
+
+# Others
+batch_size = 2
+seed = 42
+prompt_path = "./assets/texts/t2i_samples.txt"
+save_dir = "./outputs/samples/"

configs/pixart/train/16x256x256.py

@@ -0,0 +1,53 @@
+num_frames = 16
+frame_interval = 3
+image_size = (256, 256)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = False
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="PixArt-XL/2",
+    space_scale=0.5,
+    time_scale=1.0,
+    from_pretrained="PixArt-XL-2-512x512.pth",
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 1000
+load = None
+
+batch_size = 8
+lr = 2e-5
+grad_clip = 1.0

configs/pixart/train/1x512x512.py

@@ -0,0 +1,54 @@
+num_frames = 1
+frame_interval = 1
+image_size = (512, 512)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = True
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = True
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="PixArt-XL/2",
+    space_scale=1.0,
+    time_scale=1.0,
+    no_temporal_pos_emb=True,
+    from_pretrained="PixArt-XL-2-512x512.pth",
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 1000
+load = None
+
+batch_size = 32
+lr = 2e-5
+grad_clip = 1.0

configs/pixart/train/64x512x512.py

@@ -0,0 +1,54 @@
+num_frames = 64
+frame_interval = 2
+image_size = (512, 512)
+
+# Define dataset
+root = None
+data_path = "CSV_PATH"
+use_image_transform = False
+num_workers = 4
+
+# Define acceleration
+dtype = "bf16"
+grad_checkpoint = True
+plugin = "zero2"
+sp_size = 1
+
+# Define model
+model = dict(
+    type="PixArt-XL/2",
+    space_scale=1.0,
+    time_scale=2 / 3,
+    from_pretrained=None,
+    enable_flashattn=True,
+    enable_layernorm_kernel=True,
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+    micro_batch_size=128,
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",
+)
+
+# Others
+seed = 42
+outputs = "outputs"
+wandb = False
+
+epochs = 1000
+log_every = 10
+ckpt_every = 250
+load = None
+
+batch_size = 4
+lr = 2e-5
+grad_clip = 1.0

docs/README_zh.md

@@ -0,0 +1,206 @@
+<p align="center">
+    <img src="../assets/readme/icon.png" width="250"/>
+<p>
+
+<div align="center">
+    <a href="https://github.com/hpcaitech/Open-Sora/stargazers"><img src="https://img.shields.io/github/stars/hpcaitech/Open-Sora?style=social"></a>
+    <a href="https://hpcaitech.github.io/Open-Sora/"><img src="https://img.shields.io/badge/Gallery-View-orange?logo=&amp"></a>
+    <a href="https://discord.gg/shpbperhGs"><img src="https://img.shields.io/badge/Discord-join-blueviolet?logo=discord&amp"></a>
+    <a href="https://join.slack.com/t/colossalaiworkspace/shared_invite/zt-247ipg9fk-KRRYmUl~u2ll2637WRURVA"><img src="https://img.shields.io/badge/Slack-ColossalAI-blueviolet?logo=slack&amp"></a>
+    <a href="https://twitter.com/yangyou1991/status/1769411544083996787?s=61&t=jT0Dsx2d-MS5vS9rNM5e5g"><img src="https://img.shields.io/badge/Twitter-Discuss-blue?logo=twitter&amp"></a>
+    <a href="https://raw.githubusercontent.com/hpcaitech/public_assets/main/colossalai/img/WeChat.png"><img src="https://img.shields.io/badge/微信-小助手加群-green?logo=wechat&amp"></a>
+</div>
+
+## Open-Sora： 完全开源的高效复现类Sora视频生成方案
+**Open-Sora**项目是一项致力于**高效**制作高质量视频，并使所有人都能使用其模型、工具和内容的计划。
+通过采用**开源**原则，Open-Sora 不仅实现了先进视频生成技术的低成本普及，还提供了一个精简且用户友好的方案，简化了视频制作的复杂性。
+通过 Open-Sora，我们希望更多开发者一起探索内容创作领域的创新、创造和包容。
+ [[English]](/README.md)
+
+## 📰 资讯
+
+* **[2024.03.18]** 🔥 我们发布了**Open-Sora 1.0**，这是一个完全开源的视频生成项目。
+* Open-Sora 1.0 支持视频数据预处理、<a href="https://github.com/hpcaitech/ColossalAI"><img src="../assets/readme/colossal_ai.png" width="8%" ></a> 加速训练、推理等全套流程。
+* 我们提供的[模型权重](#model-weights)只需 3 天的训练就能生成 2~5 秒的 512x512 视频。
+* **[2024.03.04]** Open-Sora：开源Sora复现方案，成本降低46%，序列扩充至近百万
+
+## 🎥 最新视频
+
+| **2s 512×512**                                                                                                                                                                 | **2s 512×512**                                                                                                                                                              | **2s 512×512**                                                                                                                                    |
+| ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
+| [<img src="/assets/readme/sample_0.gif" width="">](https://github.com/hpcaitech/Open-Sora/assets/99191637/de1963d3-b43b-4e68-a670-bb821ebb6f80)                                 | [<img src="/assets/readme/sample_1.gif" width="">](https://github.com/hpcaitech/Open-Sora/assets/99191637/13f8338f-3d42-4b71-8142-d234fbd746cc)                              | [<img src="/assets/readme/sample_2.gif" width="">](https://github.com/hpcaitech/Open-Sora/assets/99191637/fa6a65a6-e32a-4d64-9a9e-eabb0ebb8c16)    |
+| A serene night scene in a forested area. [...] The video is a time-lapse, capturing the transition from day to night, with the lake and forest serving as a constant backdrop. | A soaring drone footage captures the majestic beauty of a coastal cliff, [...] The water gently laps at the rock base and the greenery that clings to the top of the cliff. | The majestic beauty of a waterfall cascading down a cliff into a serene lake. [...] The camera angle provides a bird's eye view of the waterfall. |
+| [<img src="/assets/readme/sample_3.gif" width="">](https://github.com/hpcaitech/Open-Sora/assets/99191637/64232f84-1b36-4750-a6c0-3e610fa9aa94) | [<img src="/assets/readme/sample_4.gif" width="">](https://github.com/hpcaitech/Open-Sora/assets/99191637/983a1965-a374-41a7-a76b-c07941a6c1e9) | [<img src="/assets/readme/sample_5.gif" width="">](https://github.com/hpcaitech/Open-Sora/assets/99191637/ec10c879-9767-4c31-865f-2e8d6cf11e65) |
+| A bustling city street at night, filled with the glow of car headlights and the ambient light of streetlights. [...]                                                           | The vibrant beauty of a sunflower field. The sunflowers are arranged in neat rows, creating a sense of order and symmetry. [...]                                            | A serene underwater scene featuring a sea turtle swimming through a coral reef. The turtle, with its greenish-brown shell [...]                   |
+
+视频经过降采样处理为`.gif`格式，以便显示。点击查看原始视频。为便于显示，文字经过修剪，全文请参见 [此处](/assets/texts/t2v_samples.txt)。在我们的[图片库](https://hpcaitech.github.io/Open-Sora/)中查看更多样本。
+
+## 🔆 新功能
+
+* 📍Open-Sora-v1 已发布。[这里](#model-weights)提供了模型权重。只需 400K 视频片段和在单卡 H800 上训200天（类比Stable Video Diffusion 的 152M 样本），我们就能生成 2 秒的 512×512 视频。
+* ✅ 从图像扩散模型到视频扩散模型的三阶段训练。我们提供每个阶段的权重。
+* ✅ 支持训练加速，包括加速变压器、更快的 T5 和 VAE 以及序列并行。在对 64x512x512 视频进行训练时，Open-Sora 可将训练速度提高**55%**。详细信息请参见[加速训练](docs/acceleration.md)。
+* ✅ 我们提供用于数据预处理的视频切割和字幕工具。有关说明请点击[此处](tools/data/README.md)，我们的数据收集计划请点击 [数据集](docs/datasets.md)。
+* ✅ 我们发现来自[VideoGPT](https://wilson1yan.github.io/videogpt/index.html)的 VQ-VAE 质量较低，因此采用了来自[Stability-AI](https://huggingface.co/stabilityai/sd-vae-ft-mse-original) 的更好的 VAE。我们还发现在时间维度上进行修补会降低质量。更多讨论，请参阅我们的 **[报告](docs/report_v1.md)**。
+* ✅ 我们研究了不同的架构，包括 DiT、Latte 和我们提出的 **STDiT**。我们的STDiT在质量和速度之间实现了更好的权衡。更多讨论，请参阅我们的 **[报告](docs/report_v1.md)**。
+* ✅ 支持剪辑和 T5 文本调节。
+* ✅ 通过将图像视为单帧视频，我们的项目支持在图像和视频（如 ImageNet 和 UCF101）上训练 DiT。更多说明请参见 [指令解析](docs/command.md)。
+* ✅ 利用[DiT](https://github.com/facebookresearch/DiT)、[Latte](https://github.com/Vchitect/Latte) 和 [PixArt](https://pixart-alpha.github.io/) 的官方权重支持推理。
+
+<details>
+<summary>查看更多</summary>
+
+* ✅ 重构代码库。请参阅[结构](docs/structure.md)，了解项目结构以及如何使用配置文件。
+
+</details>
+
+### 下一步计划【按优先级排序】
+
+* [ ] 完成数据处理管道（包括密集光流、美学评分、文本图像相似性、重复数据删除等）。更多信息请参见[数据集](/docs/datasets.md)。**[项目进行中］**
+* [ ] 训练视频-VAE。 **[项目进行中]**
+
+<details>
+<summary>查看更多</summary>
+
+* [ ] 支持图像和视频调节。
+* [ ] 评估流程。
+* [ ] 加入更好的调度程序，如 SD3 中的整流程序。
+* [ ] 支持可变长宽比、分辨率和持续时间。
+* [ ] 发布后支持 SD3。
+
+</details>
+
+## 目录
+
+* [安装](#installation)
+* [模型权重](/#model-weights)
+* [推理](/#inference)
+* [数据处理](/#data-processing)
+* [训练](/#training)
+* [贡献](/#contribution)
+* [声明](/#acknowledgement)
+* [引用](/#citation)
+
+## Installation
+
+```bash
+# create a virtual env
+conda create -n opensora python=3.10
+
+# install torch
+# the command below is for CUDA 12.1, choose install commands from 
+# https://pytorch.org/get-started/locally/ based on your own CUDA version
+pip3 install torch torchvision
+
+# install flash attention (optional)
+pip install packaging ninja
+pip install flash-attn --no-build-isolation
+
+# install apex (optional)
+pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" git+https://github.com/NVIDIA/apex.git
+
+# install xformers
+pip3 install -U xformers --index-url https://download.pytorch.org/whl/cu121
+
+# install this project
+git clone https://github.com/hpcaitech/Open-Sora
+cd Open-Sora
+pip install -v .
+```
+
+安装完成后，建议阅读[结构](docs/structure.md)，了解项目结构以及如何使用配置文件。
+
+## 模型权重
+
+| 分辨率  | 数据   | 迭代次数 | 批量大小 | GPU 天数 (H800) | 网址       |
+| ---------- | ------ | ----------- | ---------- | --------------- | ---------- |
+| 16×256×256 | 366K   | 80k         | 8×64       | 117             | [:link:]() |
+| 16×256×256 | 20K HQ | 24k         | 8×64       | 45              | [:link:]() |
+| 16×512×512 | 20K HQ | 20k         | 2×64       | 35              | [:link:]() |
+| 64×512×512 | 50K HQ |             |            |                 | TBD        |
+
+我们模型的权重部分由[PixArt-α](https://github.com/PixArt-alpha/PixArt-alpha) 初始化。参数数量为 724M。有关训练的更多信息，请参阅我们的 **[报告](/docs/report_v1.md)**。有关数据集的更多信息，请参阅[数据](/docs/dataset.md)。HQ 表示高质量。
+:warning: **局限性**：我们的模型是在有限的预算内训练出来的。质量和文本对齐度相对较差。特别是在生成人类时，模型表现很差，无法遵循详细的指令。我们正在努力改进质量和文本对齐。
+
+## 推理
+
+要使用我们提供的权重进行推理，首先要将[T5](https://huggingface.co/DeepFloyd/t5-v1_1-xxl/tree/main)权重下载到pretrained_models/t5_ckpts/t5-v1_1-xxl 中。然后下载模型权重。运行以下命令生成样本。请参阅[此处](docs/structure.md#inference-config-demos)自定义配置。
+
+```bash
+# Sample 16x256x256 (5s/sample)
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/opensora/inference/16x256x256.py --ckpt-path ./path/to/your/ckpt.pth
+
+# Sample 16x512x512 (20s/sample, 100 time steps)
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/opensora/inference/16x512x512.py --ckpt-path ./path/to/your/ckpt.pth
+
+# Sample 64x512x512 (40s/sample, 100 time steps)
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/opensora/inference/64x512x512.py --ckpt-path ./path/to/your/ckpt.pth
+
+# Sample 64x512x512 with sequence parallelism (30s/sample, 100 time steps)
+# sequence parallelism is enabled automatically when nproc_per_node is larger than 1
+torchrun --standalone --nproc_per_node 2 scripts/inference.py configs/opensora/inference/64x512x512.py --ckpt-path ./path/to/your/ckpt.pth
+```
+
+我们在 H800 GPU 上进行了速度测试。如需使用其他模型进行推理，请参阅[此处](docs/commands.md)获取更多说明。
+
+## 数据处理
+
+高质量数据是高质量模型的关键。[这里](/docs/datasets.md)有我们使用过的数据集和数据收集计划。我们提供处理视频数据的工具。目前，我们的数据处理流程包括以下步骤：
+
+1. 下载数据集。[[文件](/tools/datasets/README.md)]
+2. 将视频分割成片段。 [[文件](/tools/scenedetect/README.md)]
+3. 生成视频字幕。 [[文件](/tools/caption/README.md)]
+
+## 训练
+
+要启动训练，首先要将[T5](https://huggingface.co/DeepFloyd/t5-v1_1-xxl/tree/main)权重下载到pretrained_models/t5_ckpts/t5-v1_1-xxl 中。然后运行以下命令在单个节点上启动训练。
+
+```bash
+# 1 GPU, 16x256x256
+torchrun --nnodes=1 --nproc_per_node=1 scripts/train.py configs/opensora/train/16x256x512.py --data-path YOUR_CSV_PATH
+# 8 GPUs, 64x512x512
+torchrun --nnodes=1 --nproc_per_node=8 scripts/train.py configs/opensora/train/64x512x512.py --data-path YOUR_CSV_PATH --ckpt-path YOUR_PRETRAINED_CKPT
+```
+
+要在多个节点上启动训练，请根据[ColossalAI](https://colossalai.org/docs/basics/launch_colossalai/#launch-with-colossal-ai-cli) 准备一个主机文件，并运行以下命令。
+
+```bash
+colossalai run --nproc_per_node 8 --hostfile hostfile scripts/train.py configs/opensora/train/64x512x512.py --data-path YOUR_CSV_PATH --ckpt-path YOUR_PRETRAINED_CKPT
+```
+
+有关其他型号的培训和高级使用方法，请参阅[此处](docs/commands.md)获取更多说明。
+
+## 贡献
+
+如果您希望为该项目做出贡献，可以参考 [贡献指南](./CONTRIBUTING.md).
+
+## 声明
+
+* [DiT](https://github.com/facebookresearch/DiT): Scalable Diffusion Models with Transformers.
+* [OpenDiT](https://github.com/NUS-HPC-AI-Lab/OpenDiT): An acceleration for DiT training. We adopt valuable acceleration strategies for training progress from OpenDiT.
+* [PixArt](https://github.com/PixArt-alpha/PixArt-alpha): An open-source DiT-based text-to-image model.
+* [Latte](https://github.com/Vchitect/Latte): An attempt to efficiently train DiT for video.
+* [StabilityAI VAE](https://huggingface.co/stabilityai/sd-vae-ft-mse-original): A powerful image VAE model.
+* [CLIP](https://github.com/openai/CLIP): A powerful text-image embedding model.
+* [T5](https://github.com/google-research/text-to-text-transfer-transformer): A powerful text encoder.
+* [LLaVA](https://github.com/haotian-liu/LLaVA): A powerful image captioning model based on [Yi-34B](https://huggingface.co/01-ai/Yi-34B).
+
+我们对他们的出色工作和对开源的慷慨贡献表示感谢。
+
+## 引用
+
+```bibtex
+@software{opensora,
+  author = {Zangwei Zheng and Xiangyu Peng and Yang You},
+  title = {Open-Sora: Democratizing Efficient Video Production for All},
+  month = {March},
+  year = {2024},
+  url = {https://github.com/hpcaitech/Open-Sora}
+}
+```
+
+[Zangwei Zheng](https://github.com/zhengzangw) and [Xiangyu Peng](https://github.com/xyupeng) equally contributed to this work during their internship at [HPC-AI Tech](https://hpc-ai.com/).
+
+## Star 走势
+
+[![Star History Chart](https://api.star-history.com/svg?repos=hpcaitech/Open-Sora&type=Date)](https://star-history.com/#hpcaitech/Open-Sora&Date)

docs/acceleration.md

@@ -0,0 +1,57 @@
+# Acceleration
+
+Open-Sora aims to provide a high-speed training framework for diffusion models. We can achieve **55%** training speed acceleration when training on **64 frames 512x512 videos**. Our framework support training **1min 1080p videos**.
+
+## Accelerated Transformer
+
+Open-Sora boosts the training speed by:
+
+- Kernal optimization including [flash attention](https://github.com/Dao-AILab/flash-attention), fused layernorm kernal, and the ones compiled by colossalAI.
+- Hybrid parallelism including ZeRO.
+- Gradient checkpointing for larger batch size.
+
+Our training speed on images is comparable to [OpenDiT](https://github.com/NUS-HPC-AI-Lab/OpenDiT), an project to accelerate DiT training. The training speed is measured on 8 H800 GPUs with batch size 128, image size 256x256.
+
+| Model    | Throughput (img/s/GPU) | Throughput (tokens/s/GPU) |
+| -------- | ---------------------- | ------------------------- |
+| DiT      | 100                    | 26k                       |
+| OpenDiT  | 175                    | 45k                       |
+| OpenSora | 175                    | 45k                       |
+
+## Efficient STDiT
+
+Our STDiT adopts spatial-temporal attention to model the video data. Compared with directly applying full attention on DiT, our STDiT is more efficient as the number of frames increases. Our current framework only supports sequence parallelism for very long sequence.
+
+The training speed is measured on 8 H800 GPUs with acceleration techniques applied, GC means gradient checkpointing. Both with T5 conditioning like PixArt.
+
+| Model            | Setting        | Throughput (sample/s/GPU) | Throughput (tokens/s/GPU) |
+| ---------------- | -------------- | ------------------------- | ------------------------- |
+| DiT              | 16x256  (4k)   | 7.20                      | 29k                       |
+| STDiT            | 16x256  (4k)   | 7.00                      | 28k                       |
+| DiT              | 16x512  (16k)  | 0.85                      | 14k                       |
+| STDiT            | 16x512  (16k)  | 1.45                      | 23k                       |
+| DiT (GC)         | 64x512  (65k)  | 0.08                      | 5k                        |
+| STDiT (GC)       | 64x512  (65k)  | 0.40                      | 25k                       |
+| STDiT (GC, sp=2) | 360x512 (370k) | 0.10                      | 18k                       |
+
+With a 4x downsampling in the temporal dimension with Video-VAE, an 24fps video has 450 frames. The gap between the speed of STDiT (28k tokens/s) and DiT on images (up to 45k tokens/s) mainly comes from the T5 and VAE encoding, and temperal attention.
+
+## Accelerated Encoder (T5, VAE)
+
+During training, texts are encoded by T5, and videos are encoded by VAE. Typically there are two ways to accelerate the training:
+
+1. Preprocess text and video data in advance and save them to disk.
+2. Encode text and video data during training, and accelerate the encoding process.
+
+For option 1, 120 tokens for one sample require 1M disk space, and a 64x64x64 latent requires 4M. Considering a training dataset with 10M video clips, the total disk space required is 50TB. Our storage system is not ready at this time for this scale of data.
+
+For option 2, we boost T5 speed and memory requirement. According to [OpenDiT](https://github.com/NUS-HPC-AI-Lab/OpenDiT), we find VAE consumes a large number of GPU memory. Thus we split batch size into smaller ones for VAE encoding. With both techniques, we can greatly accelerated the training speed.
+
+The training speed is measured on 8 H800 GPUs with STDiT.
+
+| Acceleration | Setting       | Throughput (img/s/GPU) | Throughput (tokens/s/GPU) |
+| ------------ | ------------- | ---------------------- | ------------------------- |
+| Baseline     | 16x256  (4k)  | 6.16                   | 25k                       |
+| w. faster T5 | 16x256  (4k)  | 7.00                   | 29k                       |
+| Baseline     | 64x512  (65k) | 0.94                   | 15k                       |
+| w. both      | 64x512  (65k) | 1.45                   | 23k                       |

docs/commands.md

@@ -0,0 +1,91 @@
+# Commands
+
+## Inference
+
+You can modify corresponding config files to change the inference settings. See more details [here](/docs/structure.md#inference-config-demos).
+
+### Inference with DiT pretrained on ImageNet
+
+The following command automatically downloads the pretrained weights on ImageNet and runs inference.
+
+```bash
+python scripts/inference.py configs/dit/inference/1x256x256-class.py --ckpt-path DiT-XL-2-256x256.pt
+```
+
+### Inference with Latte pretrained on UCF101
+
+The following command automatically downloads the pretrained weights on UCF101 and runs inference.
+
+```bash
+python scripts/inference.py configs/latte/inference/16x256x256-class.py --ckpt-path Latte-XL-2-256x256-ucf101.pt
+```
+
+### Inference with PixArt-α pretrained weights
+
+Download T5 into `./pretrained_models` and run the following command.
+
+```bash
+# 256x256
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/pixart/inference/1x256x256.py --ckpt-path PixArt-XL-2-256x256.pth
+
+# 512x512
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/pixart/inference/1x512x512.py --ckpt-path PixArt-XL-2-512x512.pth
+
+# 1024 multi-scale
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/pixart/inference/1x1024MS.py --ckpt-path PixArt-XL-2-1024MS.pth
+```
+
+### Inference with checkpoints saved during training
+
+During training, an experiment logging folder is created in `outputs` directory. Under each checpoint folder, e.g. `epoch12-global_step2000`, there is a `ema.pt` and the shared `model` folder. Run the following command to perform inference.
+
+```bash
+# inference with ema model
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/opensora/inference/16x256x256.py --ckpt-path outputs/001-STDiT-XL-2/epoch12-global_step2000/ema.pt
+
+# inference with model
+torchrun --standalone --nproc_per_node 1 scripts/inference.py configs/opensora/inference/16x256x256.py --ckpt-path outputs/001-STDiT-XL-2/epoch12-global_step2000
+
+# inference with sequence parallelism
+# sequence parallelism is enabled automatically when nproc_per_node is larger than 1
+torchrun --standalone --nproc_per_node 2 scripts/inference.py configs/opensora/inference/16x256x256.py --ckpt-path outputs/001-STDiT-XL-2/epoch12-global_step2000
+```
+
+The second command will automatically generate a `model_ckpt.pt` file in the checkpoint folder.
+
+### Inference Hyperparameters
+
+1. DPM-solver is good at fast inference for images. However, the video result is not satisfactory. You can use it for fast demo purpose.
+
+```python
+type="dmp-solver"
+num_sampling_steps=20
+```
+
+1. You can use [SVD](https://huggingface.co/stabilityai/stable-video-diffusion-img2vid-xt)'s finetuned VAE decoder on videos for inference (consumes more memory). However, we do not see significant improvement in the video result. To use it, download [the pretrained weights](https://huggingface.co/maxin-cn/Latte/tree/main/t2v_required_models/vae_temporal_decoder) into `./pretrained_models/vae_temporal_decoder` and modify the config file as follows.
+
+```python
+vae = dict(
+    type="VideoAutoencoderKLTemporalDecoder",
+    from_pretrained="pretrained_models/vae_temporal_decoder",
+)
+
+## Training
+
+To resume training, run the following command. ``--load`` different from ``--ckpt-path`` as it loads the optimizer and dataloader states.
+
+```bash
+torchrun --nnodes=1 --nproc_per_node=8 scripts/train.py configs/opensora/train/64x512x512.py --data-path YOUR_CSV_PATH --load YOUR_PRETRAINED_CKPT
+```
+
+To enable wandb logging, add `--wandb` to the command.
+
+```bash
+WANDB_API_KEY=YOUR_WANDB_API_KEY torchrun --nnodes=1 --nproc_per_node=8 scripts/train.py configs/opensora/train/64x512x512.py --data-path YOUR_CSV_PATH --wandb True
+```
+
+You can modify corresponding config files to change the training settings. See more details [here](/docs/structure.md#training-config-demos).
+
+### Training Hyperparameters
+
+1. `dtype` is the data type for training. Only `fp16` and `bf16` are supported. ColossalAI automatically enables the mixed precision training for `fp16` and `bf16`. During training, we find `bf16` more stable.

docs/datasets.md

@@ -0,0 +1,28 @@
+# Datasets
+
+## Datasets used for now
+
+### HD-VG-130M
+
+[HD-VG-130M](https://github.com/daooshee/HD-VG-130M?tab=readme-ov-file) comprises 130M text-video pairs. The caption is generated by BLIP-2. We find the cut and the text quality are relatively poor. It contains 20 splits. For OpenSora 1.0, we use the first split. We plan to use the whole dataset and re-process it.
+
+### Inter4k
+
+[Inter4k](https://github.com/alexandrosstergiou/Inter4K) is a dataset containing 1k video clips with 4K resolution. The dataset is proposed for super-resolution tasks. We use the dataset for HQ training. The videos are processed as mentioned [here](/README.md#data-processing).
+
+### Pexels.com
+
+[Pexels.com](https://www.pexels.com/) is a website that provides free stock photos and videos. We collect 19K video clips from this website for HQ training. The videos are processed as mentioned [here](/README.md#data-processing).
+
+## Datasets watching list
+
+We are also watching the following datasets and considering using them in the future, which depends on our disk space and the quality of the dataset.
+
+| Name              | Size         | Description                   |
+| ----------------- | ------------ | ----------------------------- |
+| Panda-70M         | 70M videos   | High quality video-text pairs |
+| WebVid-10M        | 10M videos   | Low quality                   |
+| InternVid-10M-FLT | 10M videos   |                               |
+| EGO4D             | 3670 hours   |                               |
+| OpenDV-YouTube    | 1700 hours   |                               |
+| VidProM           | 6.69M videos |                               |

docs/report_v1.md

@@ -0,0 +1,47 @@
+# Open-Sora v1 Report
+
+OpenAI's Sora is amazing at generating one minutes high quality videos. However, it reveals almost no information about its details. To make AI more "open", we are dedicated to build an open-source version of Sora. This report describes our first attempt to train a transformer-based video diffusion model.
+
+## Efficiency in choosing the architecture
+
+To lower the computational cost, we want to utilize existing VAE models. Sora uses spatial-temporal VAE to reduce the temporal dimensions. However, we found that there is no open-source high-quality spatial-temporal VAE model. [MAGVIT](https://github.com/google-research/magvit)'s 4x4x4 VAE is not open-sourced, while [VideoGPT](https://wilson1yan.github.io/videogpt/index.html)'s 2x4x4 VAE has a low quality in our experiments. Thus, we decided to use a 2D VAE (from [Stability-AI](https://huggingface.co/stabilityai/sd-vae-ft-mse-original)) in our first version.
+
+The video training involves a large amount of tokens. Considering 24fps 1min videos, we have 1440 frames. With VAE downsampling 4x and patch size downsampling 2x, we have 1440x1024≈1.5M tokens. Full attention on 1.5M tokens leads to a huge computational cost. Thus, we use spatial-temporal attention to reduce the cost following [Latte](https://github.com/Vchitect/Latte).
+
+As shown in the figure, we insert a temporal attention right after each spatial attention in STDiT (ST stands for spatial-temporal). This is similar to variant 3 in Latte's paper. However, we do not control a similar number of parameters for these variants. While Latte's paper claims their variant is better than variant 3, our experiments on 16x256x256 videos show that with same number of iterations, the performance ranks as: DiT (full) > STDiT (Sequential) > STDiT (Parallel) ≈ Latte. Thus, we choose STDiT (Sequential) out of efficiency. Speed benchmark is provided [here](/docs/acceleration.md#efficient-stdit).
+
+![Architecture Comparison](https://i0.imgs.ovh/2024/03/15/eLk9D.png)
+
+To focus on video generation, we hope to train the model based on a powerful image generation model. [PixArt-α](https://github.com/PixArt-alpha/PixArt-alpha) is an efficiently trained high-quality image generation model with T5-conditioned DiT structure. We initialize our model with PixArt-α and initialize the projection layer of inserted temporal attention with zero. This initialization preserves model's ability of image generation at beginning, while Latte's architecture cannot. The inserted attention increases the number of parameter from 580M to 724M.
+
+![Architecture](https://i0.imgs.ovh/2024/03/16/erC1d.png)
+
+Drawing from the success of PixArt-α and Stable Video Diffusion, we also adopt a progressive training strategy: 16x256x256 on 366K pretraining datasets, and then 16x256x256, 16x512x512, and 64x512x512 on 20K datasets. With scaled position embedding, this strategy greatly reduces the computational cost.
+
+We also try to use a 3D patch embedder in DiT. However, with 2x downsampling on temporal dimension, the generated videos have a low quality. Thus, we leave the downsampling to temporal VAE in our next version. For now, we sample at every 3 frames with 16 frames training and every 2 frames with 64 frames training.
+
+## Data is the key to high quality
+
+We find that the number and quality of data have a great impact on the quality of generated videos, even larger than the model architecture and training strategy. At this time, we only prepared the first split (366K video clips) from [HD-VG-130M](https://github.com/daooshee/HD-VG-130M). The quality of these videos varies greatly, and the captions are not that accurate. Thus, we further collect 20k relatively high quality videos from [Pexels](https://www.pexels.com/), which provides free license videos. We label the video with LLaVA, an image captioning model, with three frames and a designed prompt. With designed prompt, LLaVA can generate good quality of captions.
+
+![Caption](https://i0.imgs.ovh/2024/03/16/eXdvC.png)
+
+As we lay more emphasis on the quality of data, we prepare to collect more data and build a video preprocessing pipeline in our next version.
+
+## Training Details
+
+With a limited training budgets, we made only a few exploration. We find learning rate 1e-4 is too large and scales down to 2e-5. When training with a large batch size, we find `fp16` less stable than `bf16` and may lead to generation failure. Thus, we switch to `bf16` for training on 64x512x512. For other hyper-parameters, we follow previous works.
+
+## Loss curves
+
+16x256x256 Pretraining Loss Curve
+
+![16x256x256 Pretraining Loss Curve](https://i0.imgs.ovh/2024/03/16/erXQj.png)
+
+16x256x256 HQ Training Loss Curve
+
+![16x256x256 HQ Training Loss Curve](https://i0.imgs.ovh/2024/03/16/ernXv.png)
+
+16x512x512 HQ Training Loss Curve
+
+![16x512x512 HQ Training Loss Curve](https://i0.imgs.ovh/2024/03/16/erHBe.png)

docs/structure.md

@@ -0,0 +1,178 @@
+# Repo & Config Structure
+
+## Repo Structure
+
+```plaintext
+Open-Sora
+├── README.md
+├── docs
+│   ├── acceleration.md            -> Acceleration & Speed benchmark
+│   ├── command.md                 -> Commands for training & inference
+│   ├── datasets.md                -> Datasets used in this project
+│   ├── structure.md               -> This file
+│   └── report_v1.md               -> Report for Open-Sora v1
+├── scripts
+│   ├── train.py                   -> diffusion training script
+│   └── inference.py               -> Report for Open-Sora v1
+├── configs                        -> Configs for training & inference
+├── opensora
+│   ├── __init__.py
+│   ├── registry.py                -> Registry helper
+│   ├── acceleration               -> Acceleration related code
+│   ├── dataset                    -> Dataset related code
+│   ├── models
+│   │   ├── layers                 -> Common layers
+│   │   ├── vae                    -> VAE as image encoder
+│   │   ├── text_encoder           -> Text encoder
+│   │   │   ├── classes.py         -> Class id encoder (inference only)
+│   │   │   ├── clip.py            -> CLIP encoder
+│   │   │   └── t5.py              -> T5 encoder
+│   │   ├── dit
+│   │   ├── latte
+│   │   ├── pixart
+│   │   └── stdit                  -> Our STDiT related code
+│   ├── schedulers                 -> Diffusion shedulers
+│   │   ├── iddpm                  -> IDDPM for training and inference
+│   │   └── dpms                   -> DPM-Solver for fast inference
+│   └── utils
+└── tools                          -> Tools for data processing and more
+```
+
+## Configs
+
+Our config files follows [MMEgine](https://github.com/open-mmlab/mmengine). MMEngine will reads the config file (a `.py` file) and parse it into a dictionary-like object.
+
+```plaintext
+Open-Sora
+└── configs                        -> Configs for training & inference
+    ├── opensora                   -> STDiT related configs
+    │   ├── inference
+    │   │   ├── 16x256x256.py      -> Sample videos 16 frames 256x256
+    │   │   ├── 16x512x512.py      -> Sample videos 16 frames 512x512
+    │   │   └── 64x512x512.py      -> Sample videos 64 frames 512x512
+    │   └── train
+    │       ├── 16x256x256.py      -> Train on videos 16 frames 256x256
+    │       ├── 16x256x256.py      -> Train on videos 16 frames 256x256
+    │       └── 64x512x512.py      -> Train on videos 64 frames 512x512
+    ├── dit                        -> DiT related configs
+    │   ├── inference
+    │   │   ├── 1x256x256-class.py -> Sample images with ckpts from DiT
+    │   │   ├── 1x256x256.py       -> Sample images with clip condition
+    │   │   └── 16x256x256.py      -> Sample videos
+    │   └── train
+    │       ├── 1x256x256.py       -> Train on images with clip condition
+    │       └── 16x256x256.py      -> Train on videos
+    ├── latte                      -> Latte related configs
+    └── pixart                     -> PixArt related configs
+```
+
+## Inference config demos
+
+To change the inference settings, you can directly modify the corresponding config file. Or you can pass arguments to overwrite the config file ([config_utils.py](/opensora/utils/config_utils.py)). To change sampling prompts, you should modify the `.txt` file passed to the `--prompt_path` argument.
+
+```plaintext
+--prompt_path ./assets/texts/t2v_samples.txt  -> prompt_path
+--ckpt-path ./path/to/your/ckpt.pth           -> model["from_pretrained"]
+```
+
+The explanation of each field is provided below.
+
+```python
+# Define sampling size
+num_frames = 64               # number of frames
+fps = 24 // 2                 # frames per second (divided by 2 for frame_interval=2)
+image_size = (512, 512)       # image size (height, width)
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",        # Select model type (STDiT-XL/2, DiT-XL/2, etc.)
+    space_scale=1.0,          # (Optional) Space positional encoding scale (new height / old height)
+    time_scale=2 / 3,         # (Optional) Time positional encoding scale (new frame_interval / old frame_interval)
+    enable_flashattn=True,    # (Optional) Speed up training and inference with flash attention
+    enable_layernorm_kernel=True, # (Optional) Speed up training and inference with fused kernel
+    from_pretrained="PRETRAINED_MODEL",  # (Optional) Load from pretrained model
+    no_temporal_pos_emb=True,  # (Optional) Disable temporal positional encoding (for image)
+)
+vae = dict(
+    type="VideoAutoencoderKL", # Select VAE type
+    from_pretrained="stabilityai/sd-vae-ft-ema", # Load from pretrained VAE
+    micro_batch_size=128,      # VAE with micro batch size to save memory
+)
+text_encoder = dict(
+    type="t5",                 # Select text encoder type (t5, clip)
+    from_pretrained="./pretrained_models/t5_ckpts", # Load from pretrained text encoder
+    model_max_length=120,      # Maximum length of input text
+)
+scheduler = dict(
+    type="iddpm",              # Select scheduler type (iddpm, dpm-solver)
+    num_sampling_steps=100,    # Number of sampling steps
+    cfg_scale=7.0,             # hyper-parameter for classifier-free diffusion
+)
+dtype = "fp16"                 # Computation type (fp16, fp32, bf16)
+
+# Other settings
+batch_size = 1                 # batch size
+seed = 42                      # random seed
+prompt_path = "./assets/texts/t2v_samples.txt"  # path to prompt file
+save_dir = "./samples"         # path to save samples
+```
+
+## Training config demos
+
+```python
+# Define sampling size
+num_frames = 64
+frame_interval = 2             # sample every 2 frames
+image_size = (512, 512)
+
+# Define dataset
+root = None                    # root path to the dataset
+data_path = "CSV_PATH"         # path to the csv file
+use_image_transform = False    # True if training on images
+num_workers = 4                # number of workers for dataloader
+
+# Define acceleration
+dtype = "bf16"                 # Computation type (fp16, bf16)
+grad_checkpoint = True         # Use gradient checkpointing
+plugin = "zero2"               # Plugin for distributed training (zero2, zero2-seq)
+sp_size = 1                    # Sequence parallelism size (1 for no sequence parallelism)
+
+# Define model
+model = dict(
+    type="STDiT-XL/2",
+    space_scale=1.0,
+    time_scale=2 / 3,
+    from_pretrained="YOUR_PRETRAINED_MODEL",
+    enable_flashattn=True,        # Enable flash attention
+    enable_layernorm_kernel=True, # Enable layernorm kernel
+)
+vae = dict(
+    type="VideoAutoencoderKL",
+    from_pretrained="stabilityai/sd-vae-ft-ema",
+    micro_batch_size=128,
+)
+text_encoder = dict(
+    type="t5",
+    from_pretrained="./pretrained_models/t5_ckpts",
+    model_max_length=120,
+    shardformer=True,           # Enable shardformer for T5 acceleration
+)
+scheduler = dict(
+    type="iddpm",
+    timestep_respacing="",      # Default 1000 timesteps
+)
+
+# Others
+seed = 42
+outputs = "outputs"             # path to save checkpoints
+wandb = False                   # Use wandb for logging
+
+epochs = 1000                   # number of epochs (just large enough, kill when satisfied)
+log_every = 10
+ckpt_every = 250
+load = None                     # path to resume training
+
+batch_size = 4
+lr = 2e-5
+grad_clip = 1.0                 # gradient clipping
+```

opensora/__init__.py

@@ -0,0 +1,4 @@
+from .acceleration import *
+from .datasets import *
+from .models import *
+from .registry import *

opensora/acceleration/checkpoint.py

@@ -0,0 +1,24 @@
+from collections.abc import Iterable
+
+import torch.nn as nn
+from torch.utils.checkpoint import checkpoint, checkpoint_sequential
+
+
+def set_grad_checkpoint(model, use_fp32_attention=False, gc_step=1):
+    assert isinstance(model, nn.Module)
+
+    def set_attr(module):
+        module.grad_checkpointing = True
+        module.fp32_attention = use_fp32_attention
+        module.grad_checkpointing_step = gc_step
+
+    model.apply(set_attr)
+
+
+def auto_grad_checkpoint(module, *args, **kwargs):
+    if getattr(module, "grad_checkpointing", False):
+        if not isinstance(module, Iterable):
+            return checkpoint(module, *args, **kwargs)
+        gc_step = module[0].grad_checkpointing_step
+        return checkpoint_sequential(module, gc_step, *args, **kwargs)
+    return module(*args, **kwargs)

opensora/acceleration/communications.py

@@ -0,0 +1,188 @@
+import torch
+import torch.distributed as dist
+
+
+# ====================
+# All-To-All
+# ====================
+def _all_to_all(
+    input_: torch.Tensor,
+    world_size: int,
+    group: dist.ProcessGroup,
+    scatter_dim: int,
+    gather_dim: int,
+):
+    input_list = [t.contiguous() for t in torch.tensor_split(input_, world_size, scatter_dim)]
+    output_list = [torch.empty_like(input_list[0]) for _ in range(world_size)]
+    dist.all_to_all(output_list, input_list, group=group)
+    return torch.cat(output_list, dim=gather_dim).contiguous()
+
+
+class _AllToAll(torch.autograd.Function):
+    """All-to-all communication.
+
+    Args:
+        input_: input matrix
+        process_group: communication group
+        scatter_dim: scatter dimension
+        gather_dim: gather dimension
+    """
+
+    @staticmethod
+    def forward(ctx, input_, process_group, scatter_dim, gather_dim):
+        ctx.process_group = process_group
+        ctx.scatter_dim = scatter_dim
+        ctx.gather_dim = gather_dim
+        ctx.world_size = dist.get_world_size(process_group)
+        output = _all_to_all(input_, ctx.world_size, process_group, scatter_dim, gather_dim)
+        return output
+
+    @staticmethod
+    def backward(ctx, grad_output):
+        grad_output = _all_to_all(
+            grad_output,
+            ctx.world_size,
+            ctx.process_group,
+            ctx.gather_dim,
+            ctx.scatter_dim,
+        )
+        return (
+            grad_output,
+            None,
+            None,
+            None,
+        )
+
+
+def all_to_all(
+    input_: torch.Tensor,
+    process_group: dist.ProcessGroup,
+    scatter_dim: int = 2,
+    gather_dim: int = 1,
+):
+    return _AllToAll.apply(input_, process_group, scatter_dim, gather_dim)
+
+
+def _gather(
+    input_: torch.Tensor,
+    world_size: int,
+    group: dist.ProcessGroup,
+    gather_dim: int,
+):
+    if gather_list is None:
+        gather_list = [torch.empty_like(input_) for _ in range(world_size)]
+    dist.gather(input_, gather_list, group=group, gather_dim=gather_dim)
+    return gather_list
+
+
+# ====================
+# Gather-Split
+# ====================
+
+
+def _split(input_, pg: dist.ProcessGroup, dim=-1):
+    # skip if only one rank involved
+    world_size = dist.get_world_size(pg)
+    rank = dist.get_rank(pg)
+    if world_size == 1:
+        return input_
+
+    # Split along last dimension.
+    dim_size = input_.size(dim)
+    assert dim_size % world_size == 0, (
+        f"The dimension to split ({dim_size}) is not a multiple of world size ({world_size}), "
+        f"cannot split tensor evenly"
+    )
+
+    tensor_list = torch.split(input_, dim_size // world_size, dim=dim)
+    output = tensor_list[rank].contiguous()
+
+    return output
+
+
+def _gather(input_, pg: dist.ProcessGroup, dim=-1):
+    # skip if only one rank involved
+    input_ = input_.contiguous()
+    world_size = dist.get_world_size(pg)
+    dist.get_rank(pg)
+
+    if world_size == 1:
+        return input_
+
+    # all gather
+    tensor_list = [torch.empty_like(input_) for _ in range(world_size)]
+    assert input_.device.type == "cuda"
+    torch.distributed.all_gather(tensor_list, input_, group=pg)
+
+    # concat
+    output = torch.cat(tensor_list, dim=dim).contiguous()
+
+    return output
+
+
+class _GatherForwardSplitBackward(torch.autograd.Function):
+    """Gather the input from model parallel region and concatenate.
+
+    Args:
+        input_: input matrix.
+        process_group: parallel mode.
+        dim: dimension
+    """
+
+    @staticmethod
+    def symbolic(graph, input_):
+        return _gather(input_)
+
+    @staticmethod
+    def forward(ctx, input_, process_group, dim, grad_scale):
+        ctx.mode = process_group
+        ctx.dim = dim
+        ctx.grad_scale = grad_scale
+        return _gather(input_, process_group, dim)
+
+    @staticmethod
+    def backward(ctx, grad_output):
+        if ctx.grad_scale == "up":
+            grad_output = grad_output * dist.get_world_size(ctx.mode)
+        elif ctx.grad_scale == "down":
+            grad_output = grad_output / dist.get_world_size(ctx.mode)
+
+        return _split(grad_output, ctx.mode, ctx.dim), None, None, None
+
+
+class _SplitForwardGatherBackward(torch.autograd.Function):
+    """
+    Split the input and keep only the corresponding chuck to the rank.
+
+    Args:
+        input_: input matrix.
+        process_group: parallel mode.
+        dim: dimension
+    """
+
+    @staticmethod
+    def symbolic(graph, input_):
+        return _split(input_)
+
+    @staticmethod
+    def forward(ctx, input_, process_group, dim, grad_scale):
+        ctx.mode = process_group
+        ctx.dim = dim
+        ctx.grad_scale = grad_scale
+        return _split(input_, process_group, dim)
+
+    @staticmethod
+    def backward(ctx, grad_output):
+        if ctx.grad_scale == "up":
+            grad_output = grad_output * dist.get_world_size(ctx.mode)
+        elif ctx.grad_scale == "down":
+            grad_output = grad_output / dist.get_world_size(ctx.mode)
+        return _gather(grad_output, ctx.mode, ctx.dim), None, None, None
+
+
+def split_forward_gather_backward(input_, process_group, dim, grad_scale=1.0):
+    return _SplitForwardGatherBackward.apply(input_, process_group, dim, grad_scale)
+
+
+def gather_forward_split_backward(input_, process_group, dim, grad_scale=None):
+    return _GatherForwardSplitBackward.apply(input_, process_group, dim, grad_scale)

opensora/acceleration/parallel_states.py

@@ -0,0 +1,19 @@
+import torch.distributed as dist
+
+_GLOBAL_PARALLEL_GROUPS = dict()
+
+
+def set_data_parallel_group(group: dist.ProcessGroup):
+    _GLOBAL_PARALLEL_GROUPS["data"] = group
+
+
+def get_data_parallel_group():
+    return _GLOBAL_PARALLEL_GROUPS.get("data", None)
+
+
+def set_sequence_parallel_group(group: dist.ProcessGroup):
+    _GLOBAL_PARALLEL_GROUPS["sequence"] = group
+
+
+def get_sequence_parallel_group():
+    return _GLOBAL_PARALLEL_GROUPS.get("sequence", None)

opensora/acceleration/plugin.py

@@ -0,0 +1,100 @@
+import random
+from typing import Optional
+
+import numpy as np
+import torch
+from colossalai.booster.plugin import LowLevelZeroPlugin
+from colossalai.cluster import ProcessGroupMesh
+from torch.utils.data import DataLoader
+from torch.utils.data.distributed import DistributedSampler
+
+DP_AXIS, SP_AXIS = 0, 1
+
+
+class ZeroSeqParallelPlugin(LowLevelZeroPlugin):
+    def __init__(
+        self,
+        sp_size: int = 1,
+        stage: int = 2,
+        precision: str = "fp16",
+        initial_scale: float = 2**32,
+        min_scale: float = 1,
+        growth_factor: float = 2,
+        backoff_factor: float = 0.5,
+        growth_interval: int = 1000,
+        hysteresis: int = 2,
+        max_scale: float = 2**32,
+        max_norm: float = 0.0,
+        norm_type: float = 2.0,
+        reduce_bucket_size_in_m: int = 12,
+        communication_dtype: Optional[torch.dtype] = None,
+        overlap_communication: bool = True,
+        cpu_offload: bool = False,
+        master_weights: bool = True,
+        verbose: bool = False,
+    ) -> None:
+        super().__init__(
+            stage=stage,
+            precision=precision,
+            initial_scale=initial_scale,
+            min_scale=min_scale,
+            growth_factor=growth_factor,
+            backoff_factor=backoff_factor,
+            growth_interval=growth_interval,
+            hysteresis=hysteresis,
+            max_scale=max_scale,
+            max_norm=max_norm,
+            norm_type=norm_type,
+            reduce_bucket_size_in_m=reduce_bucket_size_in_m,
+            communication_dtype=communication_dtype,
+            overlap_communication=overlap_communication,
+            cpu_offload=cpu_offload,
+            master_weights=master_weights,
+            verbose=verbose,
+        )
+        self.sp_size = sp_size
+        assert self.world_size % sp_size == 0, "world_size must be divisible by sp_size"
+        self.dp_size = self.world_size // sp_size
+        self.pg_mesh = ProcessGroupMesh(self.dp_size, self.sp_size)
+        self.dp_group = self.pg_mesh.get_group_along_axis(DP_AXIS)
+        self.sp_group = self.pg_mesh.get_group_along_axis(SP_AXIS)
+        self.dp_rank = self.pg_mesh.coordinate(DP_AXIS)
+        self.sp_rank = self.pg_mesh.coordinate(SP_AXIS)
+
+    def __del__(self):
+        """Destroy the prcess groups in ProcessGroupMesh"""
+        self.pg_mesh.destroy_mesh_process_groups()
+
+    def prepare_dataloader(
+        self,
+        dataset,
+        batch_size,
+        shuffle=False,
+        seed=1024,
+        drop_last=False,
+        pin_memory=False,
+        num_workers=0,
+        distributed_sampler_cls=None,
+        **kwargs,
+    ):
+        _kwargs = kwargs.copy()
+        distributed_sampler_cls = distributed_sampler_cls or DistributedSampler
+        sampler = distributed_sampler_cls(dataset, num_replicas=self.dp_size, rank=self.dp_rank, shuffle=shuffle)
+
+        # Deterministic dataloader
+        def seed_worker(worker_id):
+            worker_seed = seed
+            np.random.seed(worker_seed)
+            torch.manual_seed(worker_seed)
+            random.seed(worker_seed)
+
+        return DataLoader(
+            dataset,
+            batch_size=batch_size,
+            sampler=sampler,
+            worker_init_fn=seed_worker,
+            drop_last=drop_last,
+            pin_memory=pin_memory,
+            num_workers=num_workers,
+            **_kwargs,
+        )

opensora/acceleration/shardformer/modeling/t5.py

@@ -0,0 +1,39 @@
+import torch
+import torch.nn as nn
+
+
+class T5LayerNorm(nn.Module):
+    def __init__(self, hidden_size, eps=1e-6):
+        """
+        Construct a layernorm module in the T5 style. No bias and no subtraction of mean.
+        """
+        super().__init__()
+        self.weight = nn.Parameter(torch.ones(hidden_size))
+        self.variance_epsilon = eps
+
+    def forward(self, hidden_states):
+        # T5 uses a layer_norm which only scales and doesn't shift, which is also known as Root Mean
+        # Square Layer Normalization https://arxiv.org/abs/1910.07467 thus varience is calculated
+        # w/o mean and there is no bias. Additionally we want to make sure that the accumulation for
+        # half-precision inputs is done in fp32
+
+        variance = hidden_states.to(torch.float32).pow(2).mean(-1, keepdim=True)
+        hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
+
+        # convert into half-precision if necessary
+        if self.weight.dtype in [torch.float16, torch.bfloat16]:
+            hidden_states = hidden_states.to(self.weight.dtype)
+
+        return self.weight * hidden_states
+
+    @staticmethod
+    def from_native_module(module, *args, **kwargs):
+        assert module.__class__.__name__ == "FusedRMSNorm", (
+            "Recovering T5LayerNorm requires the original layer to be apex's Fused RMS Norm."
+            "Apex's fused norm is automatically used by Hugging Face Transformers https://github.com/huggingface/transformers/blob/main/src/transformers/models/t5/modeling_t5.py#L265C5-L265C48"
+        )
+
+        layer_norm = T5LayerNorm(module.normalized_shape, eps=module.eps)
+        layer_norm.weight.data.copy_(module.weight.data)
+        layer_norm = layer_norm.to(module.weight.device)
+        return layer_norm

opensora/acceleration/shardformer/policy/t5_encoder.py

@@ -0,0 +1,67 @@
+from colossalai.shardformer.modeling.jit import get_jit_fused_dropout_add_func
+from colossalai.shardformer.modeling.t5 import get_jit_fused_T5_layer_ff_forward, get_T5_layer_self_attention_forward
+from colossalai.shardformer.policies.base_policy import Policy, SubModuleReplacementDescription
+
+
+class T5EncoderPolicy(Policy):
+    def config_sanity_check(self):
+        assert not self.shard_config.enable_tensor_parallelism
+        assert not self.shard_config.enable_flash_attention
+
+    def preprocess(self):
+        return self.model
+
+    def module_policy(self):
+        from transformers.models.t5.modeling_t5 import T5LayerFF, T5LayerSelfAttention, T5Stack
+
+        policy = {}
+
+        # check whether apex is installed
+        try:
+            from opensora.acceleration.shardformer.modeling.t5 import T5LayerNorm
+
+            # recover hf from fused rms norm to T5 norm which is faster
+            self.append_or_create_submodule_replacement(
+                description=SubModuleReplacementDescription(
+                    suffix="layer_norm",
+                    target_module=T5LayerNorm,
+                ),
+                policy=policy,
+                target_key=T5LayerFF,
+            )
+            self.append_or_create_submodule_replacement(
+                description=SubModuleReplacementDescription(suffix="layer_norm", target_module=T5LayerNorm),
+                policy=policy,
+                target_key=T5LayerSelfAttention,
+            )
+            self.append_or_create_submodule_replacement(
+                description=SubModuleReplacementDescription(suffix="final_layer_norm", target_module=T5LayerNorm),
+                policy=policy,
+                target_key=T5Stack,
+            )
+        except (ImportError, ModuleNotFoundError):
+            pass
+
+        # use jit operator
+        if self.shard_config.enable_jit_fused:
+            self.append_or_create_method_replacement(
+                description={
+                    "forward": get_jit_fused_T5_layer_ff_forward(),
+                    "dropout_add": get_jit_fused_dropout_add_func(),
+                },
+                policy=policy,
+                target_key=T5LayerFF,
+            )
+            self.append_or_create_method_replacement(
+                description={
+                    "forward": get_T5_layer_self_attention_forward(),
+                    "dropout_add": get_jit_fused_dropout_add_func(),
+                },
+                policy=policy,
+                target_key=T5LayerSelfAttention,
+            )
+
+        return policy
+
+    def postprocess(self):
+        return self.model

opensora/datasets/__init__.py

@@ -0,0 +1,2 @@
+from .datasets import DatasetFromCSV, get_transforms_image, get_transforms_video
+from .utils import prepare_dataloader, save_sample

opensora/datasets/datasets.py

@@ -0,0 +1,114 @@
+import csv
+import os
+
+import numpy as np
+import torch
+import torchvision
+import torchvision.transforms as transforms
+from torchvision.datasets.folder import IMG_EXTENSIONS, pil_loader
+
+from . import video_transforms
+from .utils import center_crop_arr
+
+
+def get_transforms_video(resolution=256):
+    transform_video = transforms.Compose(
+        [
+            video_transforms.ToTensorVideo(),  # TCHW
+            video_transforms.RandomHorizontalFlipVideo(),
+            video_transforms.UCFCenterCropVideo(resolution),
+            transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
+        ]
+    )
+    return transform_video
+
+
+def get_transforms_image(image_size=256):
+    transform = transforms.Compose(
+        [
+            transforms.Lambda(lambda pil_image: center_crop_arr(pil_image, image_size)),
+            transforms.RandomHorizontalFlip(),
+            transforms.ToTensor(),
+            transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
+        ]
+    )
+    return transform
+
+
+class DatasetFromCSV(torch.utils.data.Dataset):
+    """load video according to the csv file.
+
+    Args:
+        target_video_len (int): the number of video frames will be load.
+        align_transform (callable): Align different videos in a specified size.
+        temporal_sample (callable): Sample the target length of a video.
+    """
+
+    def __init__(
+        self,
+        csv_path,
+        num_frames=16,
+        frame_interval=1,
+        transform=None,
+        root=None,
+    ):
+        self.csv_path = csv_path
+        with open(csv_path, "r") as f:
+            reader = csv.reader(f)
+            self.samples = list(reader)
+
+        ext = self.samples[0][0].split(".")[-1]
+        if ext.lower() in ("mp4", "avi", "mov", "mkv"):
+            self.is_video = True
+        else:
+            assert f".{ext.lower()}" in IMG_EXTENSIONS, f"Unsupported file format: {ext}"
+            self.is_video = False
+
+        self.transform = transform
+
+        self.num_frames = num_frames
+        self.frame_interval = frame_interval
+        self.temporal_sample = video_transforms.TemporalRandomCrop(num_frames * frame_interval)
+        self.root = root
+
+    def getitem(self, index):
+        sample = self.samples[index]
+        path = sample[0]
+        if self.root:
+            path = os.path.join(self.root, path)
+        text = sample[1]
+
+        if self.is_video:
+            vframes, aframes, info = torchvision.io.read_video(filename=path, pts_unit="sec", output_format="TCHW")
+            total_frames = len(vframes)
+
+            # Sampling video frames
+            start_frame_ind, end_frame_ind = self.temporal_sample(total_frames)
+            assert (
+                end_frame_ind - start_frame_ind >= self.num_frames
+            ), f"{path} with index {index} has not enough frames."
+            frame_indice = np.linspace(start_frame_ind, end_frame_ind - 1, self.num_frames, dtype=int)
+
+            video = vframes[frame_indice]
+            video = self.transform(video)  # T C H W
+        else:
+            image = pil_loader(path)
+            image = self.transform(image)
+            video = image.unsqueeze(0).repeat(self.num_frames, 1, 1, 1)
+
+        # TCHW -> CTHW
+        video = video.permute(1, 0, 2, 3)
+
+        return {"video": video, "text": text}
+
+    def __getitem__(self, index):
+        for _ in range(10):
+            try:
+                return self.getitem(index)
+            except Exception as e:
+                print(e)
+                index = np.random.randint(len(self))
+        raise RuntimeError("Too many bad data.")
+
+    def __len__(self):
+        return len(self.samples)

opensora/datasets/utils.py

@@ -0,0 +1,135 @@
+import random
+from typing import Iterator, Optional
+
+import numpy as np
+import torch
+from PIL import Image
+from torch.distributed import ProcessGroup
+from torch.distributed.distributed_c10d import _get_default_group
+from torch.utils.data import DataLoader, Dataset
+from torch.utils.data.distributed import DistributedSampler
+from torchvision.io import write_video
+from torchvision.utils import save_image
+
+
+def save_sample(x, fps=8, save_path=None, normalize=True, value_range=(-1, 1)):
+    """
+    Args:
+        x (Tensor): shape [C, T, H, W]
+    """
+    assert x.ndim == 4
+
+    if x.shape[1] == 1:  # T = 1: save as image
+        save_path += ".png"
+        x = x.squeeze(1)
+        save_image([x], save_path, normalize=normalize, value_range=value_range)
+    else:
+        save_path += ".mp4"
+        if normalize:
+            low, high = value_range
+            x.clamp_(min=low, max=high)
+            x.sub_(low).div_(max(high - low, 1e-5))
+
+        x = x.mul(255).add_(0.5).clamp_(0, 255).permute(1, 2, 3, 0).to("cpu", torch.uint8)
+        write_video(save_path, x, fps=fps, video_codec="h264")
+    print(f"Saved to {save_path}")
+
+
+class StatefulDistributedSampler(DistributedSampler):
+    def __init__(
+        self,
+        dataset: Dataset,
+        num_replicas: Optional[int] = None,
+        rank: Optional[int] = None,
+        shuffle: bool = True,
+        seed: int = 0,
+        drop_last: bool = False,
+    ) -> None:
+        super().__init__(dataset, num_replicas, rank, shuffle, seed, drop_last)
+        self.start_index: int = 0
+
+    def __iter__(self) -> Iterator:
+        iterator = super().__iter__()
+        indices = list(iterator)
+        indices = indices[self.start_index :]
+        return iter(indices)
+
+    def __len__(self) -> int:
+        return self.num_samples - self.start_index
+
+    def set_start_index(self, start_index: int) -> None:
+        self.start_index = start_index
+
+
+def prepare_dataloader(
+    dataset,
+    batch_size,
+    shuffle=False,
+    seed=1024,
+    drop_last=False,
+    pin_memory=False,
+    num_workers=0,
+    process_group: Optional[ProcessGroup] = None,
+    **kwargs,
+):
+    r"""
+    Prepare a dataloader for distributed training. The dataloader will be wrapped by
+    `torch.utils.data.DataLoader` and `StatefulDistributedSampler`.
+
+
+    Args:
+        dataset (`torch.utils.data.Dataset`): The dataset to be loaded.
+        shuffle (bool, optional): Whether to shuffle the dataset. Defaults to False.
+        seed (int, optional): Random worker seed for sampling, defaults to 1024.
+        add_sampler: Whether to add ``DistributedDataParallelSampler`` to the dataset. Defaults to True.
+        drop_last (bool, optional): Set to True to drop the last incomplete batch, if the dataset size
+            is not divisible by the batch size. If False and the size of dataset is not divisible by
+            the batch size, then the last batch will be smaller, defaults to False.
+        pin_memory (bool, optional): Whether to pin memory address in CPU memory. Defaults to False.
+        num_workers (int, optional): Number of worker threads for this dataloader. Defaults to 0.
+        kwargs (dict): optional parameters for ``torch.utils.data.DataLoader``, more details could be found in
+                `DataLoader <https://pytorch.org/docs/stable/_modules/torch/utils/data/dataloader.html#DataLoader>`_.
+
+    Returns:
+        :class:`torch.utils.data.DataLoader`: A DataLoader used for training or testing.
+    """
+    _kwargs = kwargs.copy()
+    process_group = process_group or _get_default_group()
+    sampler = StatefulDistributedSampler(
+        dataset, num_replicas=process_group.size(), rank=process_group.rank(), shuffle=shuffle
+    )
+
+    # Deterministic dataloader
+    def seed_worker(worker_id):
+        worker_seed = seed
+        np.random.seed(worker_seed)
+        torch.manual_seed(worker_seed)
+        random.seed(worker_seed)
+
+    return DataLoader(
+        dataset,
+        batch_size=batch_size,
+        sampler=sampler,
+        worker_init_fn=seed_worker,
+        drop_last=drop_last,
+        pin_memory=pin_memory,
+        num_workers=num_workers,
+        **_kwargs,
+    )
+
+
+def center_crop_arr(pil_image, image_size):
+    """
+    Center cropping implementation from ADM.
+    https://github.com/openai/guided-diffusion/blob/8fb3ad9197f16bbc40620447b2742e13458d2831/guided_diffusion/image_datasets.py#L126
+    """
+    while min(*pil_image.size) >= 2 * image_size:
+        pil_image = pil_image.resize(tuple(x // 2 for x in pil_image.size), resample=Image.BOX)
+
+    scale = image_size / min(*pil_image.size)
+    pil_image = pil_image.resize(tuple(round(x * scale) for x in pil_image.size), resample=Image.BICUBIC)
+
+    arr = np.array(pil_image)
+    crop_y = (arr.shape[0] - image_size) // 2
+    crop_x = (arr.shape[1] - image_size) // 2
+    return Image.fromarray(arr[crop_y : crop_y + image_size, crop_x : crop_x + image_size])

opensora/datasets/video_transforms.py

@@ -0,0 +1,501 @@
+# Copyright 2024 Vchitect/Latte
+
+# 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.# Modified from Latte
+
+# - This file is adapted from https://github.com/Vchitect/Latte/blob/main/datasets/video_transforms.py
+
+
+import numbers
+import random
+
+import torch
+
+
+def _is_tensor_video_clip(clip):
+    if not torch.is_tensor(clip):
+        raise TypeError("clip should be Tensor. Got %s" % type(clip))
+
+    if not clip.ndimension() == 4:
+        raise ValueError("clip should be 4D. Got %dD" % clip.dim())
+
+    return True
+
+
+def center_crop_arr(pil_image, image_size):
+    """
+    Center cropping implementation from ADM.
+    https://github.com/openai/guided-diffusion/blob/8fb3ad9197f16bbc40620447b2742e13458d2831/guided_diffusion/image_datasets.py#L126
+    """
+    while min(*pil_image.size) >= 2 * image_size:
+        pil_image = pil_image.resize(tuple(x // 2 for x in pil_image.size), resample=Image.BOX)
+
+    scale = image_size / min(*pil_image.size)
+    pil_image = pil_image.resize(tuple(round(x * scale) for x in pil_image.size), resample=Image.BICUBIC)
+
+    arr = np.array(pil_image)
+    crop_y = (arr.shape[0] - image_size) // 2
+    crop_x = (arr.shape[1] - image_size) // 2
+    return Image.fromarray(arr[crop_y : crop_y + image_size, crop_x : crop_x + image_size])
+
+
+def crop(clip, i, j, h, w):
+    """
+    Args:
+        clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
+    """
+    if len(clip.size()) != 4:
+        raise ValueError("clip should be a 4D tensor")
+    return clip[..., i : i + h, j : j + w]
+
+
+def resize(clip, target_size, interpolation_mode):
+    if len(target_size) != 2:
+        raise ValueError(f"target size should be tuple (height, width), instead got {target_size}")
+    return torch.nn.functional.interpolate(clip, size=target_size, mode=interpolation_mode, align_corners=False)
+
+
+def resize_scale(clip, target_size, interpolation_mode):
+    if len(target_size) != 2:
+        raise ValueError(f"target size should be tuple (height, width), instead got {target_size}")
+    H, W = clip.size(-2), clip.size(-1)
+    scale_ = target_size[0] / min(H, W)
+    return torch.nn.functional.interpolate(clip, scale_factor=scale_, mode=interpolation_mode, align_corners=False)
+
+
+def resized_crop(clip, i, j, h, w, size, interpolation_mode="bilinear"):
+    """
+    Do spatial cropping and resizing to the video clip
+    Args:
+        clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
+        i (int): i in (i,j) i.e coordinates of the upper left corner.
+        j (int): j in (i,j) i.e coordinates of the upper left corner.
+        h (int): Height of the cropped region.
+        w (int): Width of the cropped region.
+        size (tuple(int, int)): height and width of resized clip
+    Returns:
+        clip (torch.tensor): Resized and cropped clip. Size is (T, C, H, W)
+    """
+    if not _is_tensor_video_clip(clip):
+        raise ValueError("clip should be a 4D torch.tensor")
+    clip = crop(clip, i, j, h, w)
+    clip = resize(clip, size, interpolation_mode)
+    return clip
+
+
+def center_crop(clip, crop_size):
+    if not _is_tensor_video_clip(clip):
+        raise ValueError("clip should be a 4D torch.tensor")
+    h, w = clip.size(-2), clip.size(-1)
+    th, tw = crop_size
+    if h < th or w < tw:
+        raise ValueError("height and width must be no smaller than crop_size")
+
+    i = int(round((h - th) / 2.0))
+    j = int(round((w - tw) / 2.0))
+    return crop(clip, i, j, th, tw)
+
+
+def center_crop_using_short_edge(clip):
+    if not _is_tensor_video_clip(clip):
+        raise ValueError("clip should be a 4D torch.tensor")
+    h, w = clip.size(-2), clip.size(-1)
+    if h < w:
+        th, tw = h, h
+        i = 0
+        j = int(round((w - tw) / 2.0))
+    else:
+        th, tw = w, w
+        i = int(round((h - th) / 2.0))
+        j = 0
+    return crop(clip, i, j, th, tw)
+
+
+def random_shift_crop(clip):
+    """
+    Slide along the long edge, with the short edge as crop size
+    """
+    if not _is_tensor_video_clip(clip):
+        raise ValueError("clip should be a 4D torch.tensor")
+    h, w = clip.size(-2), clip.size(-1)
+
+    if h <= w:
+        short_edge = h
+    else:
+        short_edge = w
+
+    th, tw = short_edge, short_edge
+
+    i = torch.randint(0, h - th + 1, size=(1,)).item()
+    j = torch.randint(0, w - tw + 1, size=(1,)).item()
+    return crop(clip, i, j, th, tw)
+
+
+def to_tensor(clip):
+    """
+    Convert tensor data type from uint8 to float, divide value by 255.0 and
+    permute the dimensions of clip tensor
+    Args:
+        clip (torch.tensor, dtype=torch.uint8): Size is (T, C, H, W)
+    Return:
+        clip (torch.tensor, dtype=torch.float): Size is (T, C, H, W)
+    """
+    _is_tensor_video_clip(clip)
+    if not clip.dtype == torch.uint8:
+        raise TypeError("clip tensor should have data type uint8. Got %s" % str(clip.dtype))
+    # return clip.float().permute(3, 0, 1, 2) / 255.0
+    return clip.float() / 255.0
+
+
+def normalize(clip, mean, std, inplace=False):
+    """
+    Args:
+        clip (torch.tensor): Video clip to be normalized. Size is (T, C, H, W)
+        mean (tuple): pixel RGB mean. Size is (3)
+        std (tuple): pixel standard deviation. Size is (3)
+    Returns:
+        normalized clip (torch.tensor): Size is (T, C, H, W)
+    """
+    if not _is_tensor_video_clip(clip):
+        raise ValueError("clip should be a 4D torch.tensor")
+    if not inplace:
+        clip = clip.clone()
+    mean = torch.as_tensor(mean, dtype=clip.dtype, device=clip.device)
+    # print(mean)
+    std = torch.as_tensor(std, dtype=clip.dtype, device=clip.device)
+    clip.sub_(mean[:, None, None, None]).div_(std[:, None, None, None])
+    return clip
+
+
+def hflip(clip):
+    """
+    Args:
+        clip (torch.tensor): Video clip to be normalized. Size is (T, C, H, W)
+    Returns:
+        flipped clip (torch.tensor): Size is (T, C, H, W)
+    """
+    if not _is_tensor_video_clip(clip):
+        raise ValueError("clip should be a 4D torch.tensor")
+    return clip.flip(-1)
+
+
+class RandomCropVideo:
+    def __init__(self, size):
+        if isinstance(size, numbers.Number):
+            self.size = (int(size), int(size))
+        else:
+            self.size = size
+
+    def __call__(self, clip):
+        """
+        Args:
+            clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
+        Returns:
+            torch.tensor: randomly cropped video clip.
+                size is (T, C, OH, OW)
+        """
+        i, j, h, w = self.get_params(clip)
+        return crop(clip, i, j, h, w)
+
+    def get_params(self, clip):
+        h, w = clip.shape[-2:]
+        th, tw = self.size
+
+        if h < th or w < tw:
+            raise ValueError(f"Required crop size {(th, tw)} is larger than input image size {(h, w)}")
+
+        if w == tw and h == th:
+            return 0, 0, h, w
+
+        i = torch.randint(0, h - th + 1, size=(1,)).item()
+        j = torch.randint(0, w - tw + 1, size=(1,)).item()
+
+        return i, j, th, tw
+
+    def __repr__(self) -> str:
+        return f"{self.__class__.__name__}(size={self.size})"
+
+
+class CenterCropResizeVideo:
+    """
+    First use the short side for cropping length,
+    center crop video, then resize to the specified size
+    """
+
+    def __init__(
+        self,
+        size,
+        interpolation_mode="bilinear",
+    ):
+        if isinstance(size, tuple):
+            if len(size) != 2:
+                raise ValueError(f"size should be tuple (height, width), instead got {size}")
+            self.size = size
+        else:
+            self.size = (size, size)
+
+        self.interpolation_mode = interpolation_mode
+
+    def __call__(self, clip):
+        """
+        Args:
+            clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
+        Returns:
+            torch.tensor: scale resized / center cropped video clip.
+                size is (T, C, crop_size, crop_size)
+        """
+        clip_center_crop = center_crop_using_short_edge(clip)
+        clip_center_crop_resize = resize(
+            clip_center_crop, target_size=self.size, interpolation_mode=self.interpolation_mode
+        )
+        return clip_center_crop_resize
+
+    def __repr__(self) -> str:
+        return f"{self.__class__.__name__}(size={self.size}, interpolation_mode={self.interpolation_mode}"
+
+
+class UCFCenterCropVideo:
+    """
+    First scale to the specified size in equal proportion to the short edge,
+    then center cropping
+    """
+
+    def __init__(
+        self,
+        size,
+        interpolation_mode="bilinear",
+    ):
+        if isinstance(size, tuple):
+            if len(size) != 2:
+                raise ValueError(f"size should be tuple (height, width), instead got {size}")
+            self.size = size
+        else:
+            self.size = (size, size)
+
+        self.interpolation_mode = interpolation_mode
+
+    def __call__(self, clip):
+        """
+        Args:
+            clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
+        Returns:
+            torch.tensor: scale resized / center cropped video clip.
+                size is (T, C, crop_size, crop_size)
+        """
+        clip_resize = resize_scale(clip=clip, target_size=self.size, interpolation_mode=self.interpolation_mode)
+        clip_center_crop = center_crop(clip_resize, self.size)
+        return clip_center_crop
+
+    def __repr__(self) -> str:
+        return f"{self.__class__.__name__}(size={self.size}, interpolation_mode={self.interpolation_mode}"
+
+
+class KineticsRandomCropResizeVideo:
+    """
+    Slide along the long edge, with the short edge as crop size. And resie to the desired size.
+    """
+
+    def __init__(
+        self,
+        size,
+        interpolation_mode="bilinear",
+    ):
+        if isinstance(size, tuple):
+            if len(size) != 2:
+                raise ValueError(f"size should be tuple (height, width), instead got {size}")
+            self.size = size
+        else:
+            self.size = (size, size)
+
+        self.interpolation_mode = interpolation_mode
+
+    def __call__(self, clip):
+        clip_random_crop = random_shift_crop(clip)
+        clip_resize = resize(clip_random_crop, self.size, self.interpolation_mode)
+        return clip_resize
+
+
+class CenterCropVideo:
+    def __init__(
+        self,
+        size,
+        interpolation_mode="bilinear",
+    ):
+        if isinstance(size, tuple):
+            if len(size) != 2:
+                raise ValueError(f"size should be tuple (height, width), instead got {size}")
+            self.size = size
+        else:
+            self.size = (size, size)
+
+        self.interpolation_mode = interpolation_mode
+
+    def __call__(self, clip):
+        """
+        Args:
+            clip (torch.tensor): Video clip to be cropped. Size is (T, C, H, W)
+        Returns:
+            torch.tensor: center cropped video clip.
+                size is (T, C, crop_size, crop_size)
+        """
+        clip_center_crop = center_crop(clip, self.size)
+        return clip_center_crop
+
+    def __repr__(self) -> str:
+        return f"{self.__class__.__name__}(size={self.size}, interpolation_mode={self.interpolation_mode}"
+
+
+class NormalizeVideo:
+    """
+    Normalize the video clip by mean subtraction and division by standard deviation
+    Args:
+        mean (3-tuple): pixel RGB mean
+        std (3-tuple): pixel RGB standard deviation
+        inplace (boolean): whether do in-place normalization
+    """
+
+    def __init__(self, mean, std, inplace=False):
+        self.mean = mean
+        self.std = std
+        self.inplace = inplace
+
+    def __call__(self, clip):
+        """
+        Args:
+            clip (torch.tensor): video clip must be normalized. Size is (C, T, H, W)
+        """
+        return normalize(clip, self.mean, self.std, self.inplace)
+
+    def __repr__(self) -> str:
+        return f"{self.__class__.__name__}(mean={self.mean}, std={self.std}, inplace={self.inplace})"
+
+
+class ToTensorVideo:
+    """
+    Convert tensor data type from uint8 to float, divide value by 255.0 and
+    permute the dimensions of clip tensor
+    """
+
+    def __init__(self):
+        pass
+
+    def __call__(self, clip):
+        """
+        Args:
+            clip (torch.tensor, dtype=torch.uint8): Size is (T, C, H, W)
+        Return:
+            clip (torch.tensor, dtype=torch.float): Size is (T, C, H, W)
+        """
+        return to_tensor(clip)
+
+    def __repr__(self) -> str:
+        return self.__class__.__name__
+
+
+class RandomHorizontalFlipVideo:
+    """
+    Flip the video clip along the horizontal direction with a given probability
+    Args:
+        p (float): probability of the clip being flipped. Default value is 0.5
+    """
+
+    def __init__(self, p=0.5):
+        self.p = p
+
+    def __call__(self, clip):
+        """
+        Args:
+            clip (torch.tensor): Size is (T, C, H, W)
+        Return:
+            clip (torch.tensor): Size is (T, C, H, W)
+        """
+        if random.random() < self.p:
+            clip = hflip(clip)
+        return clip
+
+    def __repr__(self) -> str:
+        return f"{self.__class__.__name__}(p={self.p})"
+
+
+#  ------------------------------------------------------------
+#  ---------------------  Sampling  ---------------------------
+#  ------------------------------------------------------------
+class TemporalRandomCrop(object):
+    """Temporally crop the given frame indices at a random location.
+
+    Args:
+            size (int): Desired length of frames will be seen in the model.
+    """
+
+    def __init__(self, size):
+        self.size = size
+
+    def __call__(self, total_frames):
+        rand_end = max(0, total_frames - self.size - 1)
+        begin_index = random.randint(0, rand_end)
+        end_index = min(begin_index + self.size, total_frames)
+        return begin_index, end_index
+
+
+if __name__ == "__main__":
+    import os
+
+    import numpy as np
+    import torchvision.io as io
+    from torchvision import transforms
+    from torchvision.utils import save_image
+
+    vframes, aframes, info = io.read_video(filename="./v_Archery_g01_c03.avi", pts_unit="sec", output_format="TCHW")
+
+    trans = transforms.Compose(
+        [
+            ToTensorVideo(),
+            RandomHorizontalFlipVideo(),
+            UCFCenterCropVideo(512),
+            # NormalizeVideo(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
+            transforms.Normalize(mean=[0.5, 0.5, 0.5], std=[0.5, 0.5, 0.5], inplace=True),
+        ]
+    )
+
+    target_video_len = 32
+    frame_interval = 1
+    total_frames = len(vframes)
+    print(total_frames)
+
+    temporal_sample = TemporalRandomCrop(target_video_len * frame_interval)
+
+    # Sampling video frames
+    start_frame_ind, end_frame_ind = temporal_sample(total_frames)
+    # print(start_frame_ind)
+    # print(end_frame_ind)
+    assert end_frame_ind - start_frame_ind >= target_video_len
+    frame_indice = np.linspace(start_frame_ind, end_frame_ind - 1, target_video_len, dtype=int)
+    print(frame_indice)
+
+    select_vframes = vframes[frame_indice]
+    print(select_vframes.shape)
+    print(select_vframes.dtype)
+
+    select_vframes_trans = trans(select_vframes)
+    print(select_vframes_trans.shape)
+    print(select_vframes_trans.dtype)
+
+    select_vframes_trans_int = ((select_vframes_trans * 0.5 + 0.5) * 255).to(dtype=torch.uint8)
+    print(select_vframes_trans_int.dtype)
+    print(select_vframes_trans_int.permute(0, 2, 3, 1).shape)
+
+    io.write_video("./test.avi", select_vframes_trans_int.permute(0, 2, 3, 1), fps=8)
+
+    for i in range(target_video_len):
+        save_image(
+            select_vframes_trans[i], os.path.join("./test000", "%04d.png" % i), normalize=True, value_range=(-1, 1)
+        )

opensora/models/__init__.py

@@ -0,0 +1,6 @@
+from .dit import *
+from .latte import *
+from .pixart import *
+from .stdit import *
+from .text_encoder import *
+from .vae import *

opensora/models/dit/__init__.py

@@ -0,0 +1 @@
+from .dit import DiT, DiT_XL_2, DiT_XL_2x2