Renedffd
/

Renedffd ZhengmingYu commited on
Commit
9d3c7b1
·
0 Parent(s):

Duplicate from ZhengmingYu/DMAD

Browse files

Co-authored-by: Zhengming <ZhengmingYu@users.noreply.huggingface.co>

Files changed (31) hide show
  1. .gitattributes +42 -0
  2. LICENSE +84 -0
  3. LICENSE_EDM +439 -0
  4. LICENSE_SDXL +60 -0
  5. LICENSE_Wan2.1 +201 -0
  6. NOTICE +3 -0
  7. README.md +275 -0
  8. assets/teaser.gif +3 -0
  9. imagenet64/dmad_imagenet_frozencritic/checkpoint_model_221000/pytorch_model_ema.bin +3 -0
  10. imagenet64/dmad_imagenet_gaproute/checkpoint_model_568000/pytorch_model_ema.bin +3 -0
  11. imagenet64/dmad_imagenet_pgcritic/checkpoint_model_108000/pytorch_model_ema.bin +3 -0
  12. minimax_h3/dmad_minimax_h3_4step_full_critic.safetensors +3 -0
  13. minimax_h3/dmad_minimax_h3_4step_full_critic_comfyui.safetensors +3 -0
  14. minimax_h3/dmad_minimax_h3_4step_lora_critic.safetensors +3 -0
  15. minimax_h3/dmad_minimax_h3_4step_lora_critic_comfyui.safetensors +3 -0
  16. minimax_h3/dmad_minimax_h3_8step_full_critic.safetensors +3 -0
  17. minimax_h3/dmad_minimax_h3_8step_full_critic_comfyui.safetensors +3 -0
  18. minimax_h3/workflows/dmad_h3_4step_15s_podcast.json +1167 -0
  19. minimax_h3/workflows/dmad_h3_4step_15s_podcast.mp4 +3 -0
  20. minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.json +1214 -0
  21. minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.mp4 +3 -0
  22. minimax_h3/workflows/dmad_h3_8step_15s_wok.json +1167 -0
  23. minimax_h3/workflows/dmad_h3_8step_15s_wok.mp4 +3 -0
  24. minimax_h3/workflows/screenshot_4step_15s.png +3 -0
  25. minimax_h3/workflows/screenshot_4step_5s_8gb.png +3 -0
  26. minimax_h3/workflows/screenshot_8step_15s.png +3 -0
  27. sdxl/dmad_sdxl_1step_frozencritic_unet_fp16.bin +3 -0
  28. sdxl/dmad_sdxl_1step_unet_fp16.bin +3 -0
  29. sdxl/dmad_sdxl_4step_unet_fp16.bin +3 -0
  30. wan2.1/dmad_wan2pt1_14B.pth +3 -0
  31. wan2.1/dmad_wan2pt1_1pt3B.pth +3 -0
.gitattributes ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ assets/teaser.gif filter=lfs diff=lfs merge=lfs -text
37
+ minimax_h3/workflows/dmad_h3_4step_15s_podcast.mp4 filter=lfs diff=lfs merge=lfs -text
38
+ minimax_h3/workflows/dmad_h3_8step_15s_wok.mp4 filter=lfs diff=lfs merge=lfs -text
39
+ minimax_h3/workflows/screenshot_4step_15s.png filter=lfs diff=lfs merge=lfs -text
40
+ minimax_h3/workflows/screenshot_8step_15s.png filter=lfs diff=lfs merge=lfs -text
41
+ minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.mp4 filter=lfs diff=lfs merge=lfs -text
42
+ minimax_h3/workflows/screenshot_4step_5s_8gb.png filter=lfs diff=lfs merge=lfs -text
LICENSE ADDED
@@ -0,0 +1,84 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ MiniMax H3 COMMUNITY LICENSE AGREEMENT
2
+ MiniMax H3 release date/License date: August 2, 2026.
3
+ The scope of this License Agreement (this “Agreement”) is expressly limited to the “Applicable Territory” as defined below.
4
+ By clicking to accept, or by using, reproducing, modifying, distributing, running, or displaying any portion or element of the MiniMax H3 Works (including through any Hosted Services) in any manner, you acknowledge and accept the terms of this Agreement, and this Agreement shall take immediate effect upon the occurrence of such act.
5
+ I. Definitions
6
+ 1. “Acceptable Use Policy” means the policy published by MiniMax in Exhibit A.
7
+ 2. “Agreement” means the terms and conditions set forth herein that govern the use, reproduction, distribution, modification, running, and display of the MiniMax H3 Works or any portion or element thereof.
8
+ 3. “Applicable Territory” means worldwide, excluding the Excluded Territories.
9
+ 4. “Documentation” means the specifications, manuals, and documentation concerning MiniMax H3 that are publicly released by MiniMax.
10
+ 5. “Excluded Territories” means the European Union, the United Kingdom, the Republic of Korea and the United States of America.
11
+ 6. “MiniMax H3” means the video generation model, together with its software and algorithms, including trained model weights, parameters (including optimizer states), machine-learning model code, inference-supporting code, and other elements thereof made publicly available by Us, as released at https://huggingface.co/MiniMaxAI/MiniMax-H3.
12
+ 7. “MiniMax H3 Works” means (i) the Materials, (ii) the Model Derivatives, and (iii) all derivatives thereof.
13
+ 8. “Hosted Services” means hosted services provided via application programming interfaces (APIs), web access, or any other electronic or remote means.
14
+ 9. “Licensee,” “you,” or “your” means the natural or legal person exercising rights and/or using the MiniMax H3 Works for any purpose in any field of use under this Agreement.
15
+ 10. “Materials” means, collectively, MiniMax H3 and the Documentation (and any portion thereof), in each case as made available by MiniMax under this Agreement and proprietary to MiniMax.
16
+ 11. “Model Derivatives” means all of the following: (i) any modification of MiniMax H3 or any Model Derivative thereof; (ii) any work based on MiniMax H3 or any Model Derivative thereof; or (iii) any other machine learning model created by transferring the patterns of the weights, parameters, operational patterns, or Outputs of MiniMax H3 or any Model Derivative thereof to another model, such that the latter model exhibits behavior similar to MiniMax H3 or its Model Derivatives, including by distillation methods, methods using intermediate data representations, or methods based on training using synthetic-data Outputs generated by MiniMax H3 or its Model Derivatives. For the avoidance of doubt, Outputs are not deemed Model Derivatives.
17
+ 12. “Output” means any result of operating or otherwise using MiniMax H3 or any Model Derivatives (including through Hosted Services).
18
+ 13. “Third Party” means any natural or legal person that is not under common control with us or with you.
19
+ 14. “Including” means “including but not limited to.”
20
+ 15. “We,” “Us” or “MiniMax” means Nanonoble Pte. Ltd..
21
+ II. Grant of Rights
22
+ Solely within the Applicable Territory, we grant you a non-exclusive, non-transferable, royalty-free, limited license to use, reproduce, distribute, create derivative works (including Model Derivatives), and modify the Materials in accordance with the terms of this Agreement and the Acceptable Use Policy, based on the intellectual property and other rights owned by MiniMax that are embodied in or used by the Materials. You shall not violate (or encourage or permit any person to violate) any term of this Agreement or the Acceptable Use Policy.
23
+ We will continuously evaluate the applicable laws, regulations and compliance requirements for the Excluded Territories. In the meantime, should any person in such Excluded Territories be interested in deploying our models, you are welcome to contact us about obtaining a license, which will be granted based on robust controls and guardrails for purposes of complying with the laws, regulations and compliance requirements of the Excluded Territories.
24
+ III. Distribution and Redistribution
25
+ Subject to and conditioned on your continuing compliance with this Agreement, including its territorial restrictions and the Acceptable Use Policy, and solely within the Applicable Territory, you may distribute or make available the MiniMax H3 Works to Third Parties within the Applicable Territory; provided, that all of the following conditions are met:
26
+ 1. You must provide a copy of this Agreement to all such Third Parties who receive the MiniMax H3 Works or use your products or services related thereto;
27
+ 2. You must cause any modified files to carry prominent notices stating that you have modified such files;
28
+ 3. You are encouraged to:
29
+ a. display a notice on any product or service developed using MiniMax H3 indicating that the product or service is “Powered by MiniMax H3”;
30
+ b. add an AI-generation identifier to files produced using generative AI models including MiniMax H3; and
31
+ c. publish at least one technical blog post or a public statement describing your experience using MiniMax H3 Works;
32
+ 4. All distributions to Third Parties (other than through Hosted Services) must be accompanied by a “NOTICE” text file containing the following notice:
33
+ “MiniMax H3 is licensed under the MiniMax H3 Community License Agreement, Copyright © 2026 MiniMax. All Rights Reserved.”
34
+ You may add your own copyright notices on your modifications; except as provided in this Section and in Section V, however, you may not impose additional or different terms and conditions on the use, reproduction, or distribution of your modifications or of any aggregate Model Derivatives, and your use, reproduction, modification, distribution, running, and display of the work must otherwise comply with the terms and conditions of this Agreement (including the provisions concerning the Applicable Territory). If you receive the MiniMax H3 Works from a Licensee as part of an integrated end-user product, the provisions of Section III of this Agreement do not apply to you, but Section V and Exhibit A remain applicable.
35
+ IV. Additional Commercial Terms
36
+ 1. You shall obtain a separate, prior written authorization from MiniMax by contacting api@minimax.io with the subject line “MiniMax H3 licensing - authorization request”, if your commercial products and services generate more than 20 million US dollars (or equivalent in other currencies) in yearly revenue.
37
+ 2. You shall prominently display “MiniMax H3”on the user interface of commercial product or service that uses MiniMax H3 or MiniMax H3 Works.
38
+ V. Use Restrictions
39
+ 1. Your use of the MiniMax H3 Works must comply with applicable laws and regulations (including trade-compliance laws and regulations) and must comply with the Acceptable Use Policy for the MiniMax H3 Works, which is incorporated into this Agreement by reference.
40
+ 2. Before providing access to the MiniMax H3 Works or any product, service, or Hosted Service incorporating them, you must bind each recipient or user to enforceable terms at least as protective as the use restrictions in this Section V and Exhibit A, and you must notify each recipient or user that those restrictions apply.
41
+ 3. You may not use the MiniMax H3 Works or any of their Outputs or results to improve any other artificial intelligence model (other than MiniMax H3 or its Model Derivatives).
42
+ 4. You may not use, reproduce, modify, distribute, or display the MiniMax H3 Works or any of their Outputs or results outside the Applicable Territory. Any such use outside the Applicable Territory is not authorized by this Agreement.
43
+ 5. If you provide or make available to any Third Party a product, service, or Hosted Service that permits the generation of Outputs using MiniMax H3 or any Model Derivative, you must, before making that product or service available and throughout its operation, implement, maintain, test, and periodically review reasonable and proportionate technical and organizational safeguards designed to prevent and mitigate access, uses, and Outputs that violate this Section V or Exhibit A, including uses or Outputs that infringe, misappropriate, or otherwise violate any Third Party’s intellectual-property or other rights. You must not knowingly disable, materially weaken, or permit the circumvention of those safeguards. You must maintain a reasonably accessible mechanism for reporting suspected violations. Upon receiving a good-faith report or otherwise obtaining actual knowledge of a violation, you must promptly investigate and take reasonable steps within your control to stop or mitigate the violation, including removing or disabling access to offending content or services and suspending or terminating repeat violators where appropriate. You are responsible for implementing and enforcing these requirements with respect to your products, services, systems, users, and downstream recipients.
44
+ VI. Intellectual Property
45
+ 1. Subject to MiniMax’s rights in the MiniMax H3 Works (and the intellectual property therein), and to your compliance with the terms and conditions of this Agreement, as between you and MiniMax, you will own the derivative works and modifications of the Materials that you have created or had created, as well as any Model Derivatives.
46
+ 2. Except for the limited license expressly granted in this paragraph, no trademark license is granted under this Agreement; with respect to MiniMax H3 Works, the Licensee may not use any name or mark owned by or associated with MiniMax or any of its affiliates, except as reasonably and customarily necessary to describe and distribute the MiniMax H3 Works. MiniMax hereby grants you a license to use the “MiniMax H3” mark (the “Mark”) within the Applicable Territory solely for the purpose of complying with Section III.3; provided, that you comply with all applicable trademark-protection laws. All goodwill arising from your use of the Mark shall inure to the benefit of MiniMax.
47
+ 3. If you bring or assert any suit or other legal proceeding (including a cross-claim or counterclaim in any action) against us or any other natural or legal person alleging that the Materials, any Output, or any portion of the foregoing infringes any intellectual property right or other right owned by you or for which you can obtain a license, all licenses granted to you under this Agreement will terminate as of the date such suit or proceeding is filed. You shall defend, indemnify, and hold us harmless against any Third-Party claim arising out of or related to the use or distribution of the MiniMax H3 Works by you or by any Third Party.
48
+ 4. MiniMax claims no rights over the Outputs you generate. You and your users are entirely responsible for the Outputs and any subsequent use thereof.
49
+ VII. Disclaimers and Limitations of Liability
50
+ 1. We have no obligation to support, update, provide training for, or develop any further version of the MiniMax H3 Works, or to grant any license with respect thereto.
51
+ 2. UNLESS AND ONLY TO THE EXTENT REQUIRED BY APPLICABLE LAW, THE MINIMAX H3 WORKS AND ANY OUTPUT AND RESULTS THEREFROM ARE PROVIDED “AS IS” WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES OF ANY KIND INCLUDING ANY WARRANTIES OF TITLE, MERCHANTABILITY, NONINFRINGEMENT, COURSE OF DEALING, USAGE OF TRADE, OR FITNESS FOR A PARTICULAR PURPOSE. YOU ARE SOLELY RESPONSIBLE FOR DETERMINING THE APPROPRIATENESS OF USING, REPRODUCING, MODIFYING, PERFORMING, DISPLAYING OR DISTRIBUTING ANY OF THE MINIMAX H3 WORKS OR OUTPUTS AND ASSUME ANY AND ALL RISKS ASSOCIATED WITH YOUR OR A THIRD PARTY’S USE OR DISTRIBUTION OF ANY OF THE MINIMAX H3 WORKS OR OUTPUTS AND YOUR EXERCISE OF RIGHTS AND PERMISSIONS UNDER THIS AGREEMENT.
52
+ 3. TO THE FULLEST EXTENT PERMITTED BY APPLICABLE LAW, IN NO EVENT SHALL MINIMAX OR ITS AFFILIATES BE LIABLE UNDER ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, TORT, NEGLIGENCE, PRODUCTS LIABILITY, OR OTHERWISE, FOR ANY DAMAGES, INCLUDING ANY DIRECT, INDIRECT, SPECIAL, INCIDENTAL, EXEMPLARY, CONSEQUENTIAL OR PUNITIVE DAMAGES, OR LOST PROFITS OF ANY KIND ARISING FROM THIS AGREEMENT OR RELATED TO ANY OF THE MINIMAX H3 WORKS OR OUTPUTS, EVEN IF MINIMAX OR ITS AFFILIATES HAVE BEEN ADVISED OF THE POSSIBILITY OF ANY OF THE FOREGOING.
53
+ VIII. Term and Termination
54
+ 1. This Agreement is effective from the moment you accept this Agreement or begin accessing the Materials, and, subject to your compliance with its terms and conditions, will remain in effect until terminated as provided herein.
55
+ 2. If you breach any term or condition of this Agreement, we have the right to terminate this Agreement. Upon termination, you must immediately cease accessing, using, and distributing the MiniMax H3 Works; delete or destroy all copies within your possession or control; and notify each downstream recipient that your authorization has ended. The obligations in the preceding sentence and Sections VI.1, VI.3, VII, and IX survive termination.
56
+ IX. Governing Law and Jurisdiction
57
+ 1. This Agreement, and any dispute arising out of or related to this Agreement, shall be governed by the laws of the Hong Kong Special Administrative Region of the People’s Republic of China, without regard to its conflict-of-laws rules. The United Nations Convention on Contracts for the International Sale of Goods does not apply to this Agreement.
58
+ 2. Any dispute arising out of or related to this Agreement shall be subject to the exclusive jurisdiction of the courts of the Hong Kong Special Administrative Region of the People’s Republic of China with competent jurisdiction. Both MiniMax and the Licensee hereby consent to the exclusive jurisdiction of such courts for any such dispute.
59
+ Additional Note: Please note that the encoder of MiniMax H3 uses Qwen3-VL-32B, which is licensed under Apache 2.0 License: https://github.com/QwenLM/Qwen3-VL/blob/main/LICENSE.
60
+
61
+ Exhibit A — Acceptable Use Policy
62
+ MiniMax reserves the right to update this Acceptable Use Policy from time to time.
63
+ Last revised: August 2, 2026.
64
+ MiniMax is committed to promoting the safe and fair use of its tools and features, including MiniMax H3. You agree not to use MiniMax H3, any Model Derivatives, or any Output in any of the following ways:
65
+ 1. Use outside the Applicable Territory;
66
+ 2. Use in any manner that violates any applicable national, federal, state, local, or international law, regulation, or other legal requirement, or that infringes, misappropriates, or otherwise violates any Third Party’s intellectual-property or other proprietary rights, including through unauthorized reproduction, distribution, public display, public performance, or creation of derivative works;
67
+ 3. Use in any manner that may harm yourself or others;
68
+ 4. Use to repurpose or distribute the Outputs of MiniMax H3 or any Model Derivatives in order to harm yourself or others;
69
+ 5. Use to circumvent or bypass any safety guardrails or safeguards we have implemented;
70
+ 6. Use in any manner that exploits or harms, or intends to exploit or harm, minors;
71
+ 7. Use to generate or disseminate verifiably false information and/or content for the purpose of harming others or influencing elections;
72
+ 8. Use to manufacture or facilitate false online engagement, including fake reviews and other means of false online engagement;
73
+ 9. Use to intentionally defame, disparage, or otherwise harass others;
74
+ 10. Use to generate and/or disseminate malware (including ransomware) or any other content intended to damage electronic systems;
75
+ 11. Use to generate or disseminate personally identifiable information for the purpose of harming others;
76
+ 12. Use to generate or disseminate information (including images, code, posts, or articles) in or to any public environment (including via bot tweets or similar means) without clearly and prominently disclosing that such information and/or content is machine-generated;
77
+ 13. Use to impersonate another person without that person’s consent, authorization, or lawful right to do so;
78
+ 14. Use to make high-risk automated decisions in critical domains that affect individual safety, rights, or well-being (such as law enforcement, immigration, healthcare or medical services, critical-infrastructure management, product-safety components, essential services, credit, employment, housing, education, social scoring, or insurance);
79
+ 15. Use in any manner that violates or disregards the social, ethical, or moral standards of other countries or regions;
80
+ 16. Use to carry out, assist, threaten, incite, plan, advocate for, or encourage violent extremism or terrorism;
81
+ 17. Use for any purpose intended to discriminate against, or harm, individuals or groups based on protected characteristics or categories, online or offline social behavior, or known or predicted personality traits;
82
+ 18. Use to intentionally exploit the vulnerabilities of specific populations based on age, social, physical, or psychological characteristics, so as to materially distort the behavior of a member of that group in a manner that causes, or is likely to cause, physical or psychological harm to that person or to others;
83
+ 19. Use for military purposes;
84
+ 20. Use to engage in any unauthorized or unlicensed professional activity, including but not limited to financial, legal, medical or healthcare, or other professional practice.
LICENSE_EDM ADDED
@@ -0,0 +1,439 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
2
+
3
+ Attribution-NonCommercial-ShareAlike 4.0 International
4
+
5
+ =======================================================================
6
+
7
+ Creative Commons Corporation ("Creative Commons") is not a law firm and
8
+ does not provide legal services or legal advice. Distribution of
9
+ Creative Commons public licenses does not create a lawyer-client or
10
+ other relationship. Creative Commons makes its licenses and related
11
+ information available on an "as-is" basis. Creative Commons gives no
12
+ warranties regarding its licenses, any material licensed under their
13
+ terms and conditions, or any related information. Creative Commons
14
+ disclaims all liability for damages resulting from their use to the
15
+ fullest extent possible.
16
+
17
+ Using Creative Commons Public Licenses
18
+
19
+ Creative Commons public licenses provide a standard set of terms and
20
+ conditions that creators and other rights holders may use to share
21
+ original works of authorship and other material subject to copyright
22
+ and certain other rights specified in the public license below. The
23
+ following considerations are for informational purposes only, are not
24
+ exhaustive, and do not form part of our licenses.
25
+
26
+ Considerations for licensors: Our public licenses are
27
+ intended for use by those authorized to give the public
28
+ permission to use material in ways otherwise restricted by
29
+ copyright and certain other rights. Our licenses are
30
+ irrevocable. Licensors should read and understand the terms
31
+ and conditions of the license they choose before applying it.
32
+ Licensors should also secure all rights necessary before
33
+ applying our licenses so that the public can reuse the
34
+ material as expected. Licensors should clearly mark any
35
+ material not subject to the license. This includes other CC-
36
+ licensed material, or material used under an exception or
37
+ limitation to copyright. More considerations for licensors:
38
+ wiki.creativecommons.org/Considerations_for_licensors
39
+
40
+ Considerations for the public: By using one of our public
41
+ licenses, a licensor grants the public permission to use the
42
+ licensed material under specified terms and conditions. If
43
+ the licensor's permission is not necessary for any reason--for
44
+ example, because of any applicable exception or limitation to
45
+ copyright--then that use is not regulated by the license. Our
46
+ licenses grant only permissions under copyright and certain
47
+ other rights that a licensor has authority to grant. Use of
48
+ the licensed material may still be restricted for other
49
+ reasons, including because others have copyright or other
50
+ rights in the material. A licensor may make special requests,
51
+ such as asking that all changes be marked or described.
52
+ Although not required by our licenses, you are encouraged to
53
+ respect those requests where reasonable. More considerations
54
+ for the public:
55
+ wiki.creativecommons.org/Considerations_for_licensees
56
+
57
+ =======================================================================
58
+
59
+ Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
60
+ Public License
61
+
62
+ By exercising the Licensed Rights (defined below), You accept and agree
63
+ to be bound by the terms and conditions of this Creative Commons
64
+ Attribution-NonCommercial-ShareAlike 4.0 International Public License
65
+ ("Public License"). To the extent this Public License may be
66
+ interpreted as a contract, You are granted the Licensed Rights in
67
+ consideration of Your acceptance of these terms and conditions, and the
68
+ Licensor grants You such rights in consideration of benefits the
69
+ Licensor receives from making the Licensed Material available under
70
+ these terms and conditions.
71
+
72
+
73
+ Section 1 -- Definitions.
74
+
75
+ a. Adapted Material means material subject to Copyright and Similar
76
+ Rights that is derived from or based upon the Licensed Material
77
+ and in which the Licensed Material is translated, altered,
78
+ arranged, transformed, or otherwise modified in a manner requiring
79
+ permission under the Copyright and Similar Rights held by the
80
+ Licensor. For purposes of this Public License, where the Licensed
81
+ Material is a musical work, performance, or sound recording,
82
+ Adapted Material is always produced where the Licensed Material is
83
+ synched in timed relation with a moving image.
84
+
85
+ b. Adapter's License means the license You apply to Your Copyright
86
+ and Similar Rights in Your contributions to Adapted Material in
87
+ accordance with the terms and conditions of this Public License.
88
+
89
+ c. BY-NC-SA Compatible License means a license listed at
90
+ creativecommons.org/compatiblelicenses, approved by Creative
91
+ Commons as essentially the equivalent of this Public License.
92
+
93
+ d. Copyright and Similar Rights means copyright and/or similar rights
94
+ closely related to copyright including, without limitation,
95
+ performance, broadcast, sound recording, and Sui Generis Database
96
+ Rights, without regard to how the rights are labeled or
97
+ categorized. For purposes of this Public License, the rights
98
+ specified in Section 2(b)(1)-(2) are not Copyright and Similar
99
+ Rights.
100
+
101
+ e. Effective Technological Measures means those measures that, in the
102
+ absence of proper authority, may not be circumvented under laws
103
+ fulfilling obligations under Article 11 of the WIPO Copyright
104
+ Treaty adopted on December 20, 1996, and/or similar international
105
+ agreements.
106
+
107
+ f. Exceptions and Limitations means fair use, fair dealing, and/or
108
+ any other exception or limitation to Copyright and Similar Rights
109
+ that applies to Your use of the Licensed Material.
110
+
111
+ g. License Elements means the license attributes listed in the name
112
+ of a Creative Commons Public License. The License Elements of this
113
+ Public License are Attribution, NonCommercial, and ShareAlike.
114
+
115
+ h. Licensed Material means the artistic or literary work, database,
116
+ or other material to which the Licensor applied this Public
117
+ License.
118
+
119
+ i. Licensed Rights means the rights granted to You subject to the
120
+ terms and conditions of this Public License, which are limited to
121
+ all Copyright and Similar Rights that apply to Your use of the
122
+ Licensed Material and that the Licensor has authority to license.
123
+
124
+ j. Licensor means the individual(s) or entity(ies) granting rights
125
+ under this Public License.
126
+
127
+ k. NonCommercial means not primarily intended for or directed towards
128
+ commercial advantage or monetary compensation. For purposes of
129
+ this Public License, the exchange of the Licensed Material for
130
+ other material subject to Copyright and Similar Rights by digital
131
+ file-sharing or similar means is NonCommercial provided there is
132
+ no payment of monetary compensation in connection with the
133
+ exchange.
134
+
135
+ l. Share means to provide material to the public by any means or
136
+ process that requires permission under the Licensed Rights, such
137
+ as reproduction, public display, public performance, distribution,
138
+ dissemination, communication, or importation, and to make material
139
+ available to the public including in ways that members of the
140
+ public may access the material from a place and at a time
141
+ individually chosen by them.
142
+
143
+ m. Sui Generis Database Rights means rights other than copyright
144
+ resulting from Directive 96/9/EC of the European Parliament and of
145
+ the Council of 11 March 1996 on the legal protection of databases,
146
+ as amended and/or succeeded, as well as other essentially
147
+ equivalent rights anywhere in the world.
148
+
149
+ n. You means the individual or entity exercising the Licensed Rights
150
+ under this Public License. Your has a corresponding meaning.
151
+
152
+
153
+ Section 2 -- Scope.
154
+
155
+ a. License grant.
156
+
157
+ 1. Subject to the terms and conditions of this Public License,
158
+ the Licensor hereby grants You a worldwide, royalty-free,
159
+ non-sublicensable, non-exclusive, irrevocable license to
160
+ exercise the Licensed Rights in the Licensed Material to:
161
+
162
+ a. reproduce and Share the Licensed Material, in whole or
163
+ in part, for NonCommercial purposes only; and
164
+
165
+ b. produce, reproduce, and Share Adapted Material for
166
+ NonCommercial purposes only.
167
+
168
+ 2. Exceptions and Limitations. For the avoidance of doubt, where
169
+ Exceptions and Limitations apply to Your use, this Public
170
+ License does not apply, and You do not need to comply with
171
+ its terms and conditions.
172
+
173
+ 3. Term. The term of this Public License is specified in Section
174
+ 6(a).
175
+
176
+ 4. Media and formats; technical modifications allowed. The
177
+ Licensor authorizes You to exercise the Licensed Rights in
178
+ all media and formats whether now known or hereafter created,
179
+ and to make technical modifications necessary to do so. The
180
+ Licensor waives and/or agrees not to assert any right or
181
+ authority to forbid You from making technical modifications
182
+ necessary to exercise the Licensed Rights, including
183
+ technical modifications necessary to circumvent Effective
184
+ Technological Measures. For purposes of this Public License,
185
+ simply making modifications authorized by this Section 2(a)
186
+ (4) never produces Adapted Material.
187
+
188
+ 5. Downstream recipients.
189
+
190
+ a. Offer from the Licensor -- Licensed Material. Every
191
+ recipient of the Licensed Material automatically
192
+ receives an offer from the Licensor to exercise the
193
+ Licensed Rights under the terms and conditions of this
194
+ Public License.
195
+
196
+ b. Additional offer from the Licensor -- Adapted Material.
197
+ Every recipient of Adapted Material from You
198
+ automatically receives an offer from the Licensor to
199
+ exercise the Licensed Rights in the Adapted Material
200
+ under the conditions of the Adapter's License You apply.
201
+
202
+ c. No downstream restrictions. You may not offer or impose
203
+ any additional or different terms or conditions on, or
204
+ apply any Effective Technological Measures to, the
205
+ Licensed Material if doing so restricts exercise of the
206
+ Licensed Rights by any recipient of the Licensed
207
+ Material.
208
+
209
+ 6. No endorsement. Nothing in this Public License constitutes or
210
+ may be construed as permission to assert or imply that You
211
+ are, or that Your use of the Licensed Material is, connected
212
+ with, or sponsored, endorsed, or granted official status by,
213
+ the Licensor or others designated to receive attribution as
214
+ provided in Section 3(a)(1)(A)(i).
215
+
216
+ b. Other rights.
217
+
218
+ 1. Moral rights, such as the right of integrity, are not
219
+ licensed under this Public License, nor are publicity,
220
+ privacy, and/or other similar personality rights; however, to
221
+ the extent possible, the Licensor waives and/or agrees not to
222
+ assert any such rights held by the Licensor to the limited
223
+ extent necessary to allow You to exercise the Licensed
224
+ Rights, but not otherwise.
225
+
226
+ 2. Patent and trademark rights are not licensed under this
227
+ Public License.
228
+
229
+ 3. To the extent possible, the Licensor waives any right to
230
+ collect royalties from You for the exercise of the Licensed
231
+ Rights, whether directly or through a collecting society
232
+ under any voluntary or waivable statutory or compulsory
233
+ licensing scheme. In all other cases the Licensor expressly
234
+ reserves any right to collect such royalties, including when
235
+ the Licensed Material is used other than for NonCommercial
236
+ purposes.
237
+
238
+
239
+ Section 3 -- License Conditions.
240
+
241
+ Your exercise of the Licensed Rights is expressly made subject to the
242
+ following conditions.
243
+
244
+ a. Attribution.
245
+
246
+ 1. If You Share the Licensed Material (including in modified
247
+ form), You must:
248
+
249
+ a. retain the following if it is supplied by the Licensor
250
+ with the Licensed Material:
251
+
252
+ i. identification of the creator(s) of the Licensed
253
+ Material and any others designated to receive
254
+ attribution, in any reasonable manner requested by
255
+ the Licensor (including by pseudonym if
256
+ designated);
257
+
258
+ ii. a copyright notice;
259
+
260
+ iii. a notice that refers to this Public License;
261
+
262
+ iv. a notice that refers to the disclaimer of
263
+ warranties;
264
+
265
+ v. a URI or hyperlink to the Licensed Material to the
266
+ extent reasonably practicable;
267
+
268
+ b. indicate if You modified the Licensed Material and
269
+ retain an indication of any previous modifications; and
270
+
271
+ c. indicate the Licensed Material is licensed under this
272
+ Public License, and include the text of, or the URI or
273
+ hyperlink to, this Public License.
274
+
275
+ 2. You may satisfy the conditions in Section 3(a)(1) in any
276
+ reasonable manner based on the medium, means, and context in
277
+ which You Share the Licensed Material. For example, it may be
278
+ reasonable to satisfy the conditions by providing a URI or
279
+ hyperlink to a resource that includes the required
280
+ information.
281
+ 3. If requested by the Licensor, You must remove any of the
282
+ information required by Section 3(a)(1)(A) to the extent
283
+ reasonably practicable.
284
+
285
+ b. ShareAlike.
286
+
287
+ In addition to the conditions in Section 3(a), if You Share
288
+ Adapted Material You produce, the following conditions also apply.
289
+
290
+ 1. The Adapter's License You apply must be a Creative Commons
291
+ license with the same License Elements, this version or
292
+ later, or a BY-NC-SA Compatible License.
293
+
294
+ 2. You must include the text of, or the URI or hyperlink to, the
295
+ Adapter's License You apply. You may satisfy this condition
296
+ in any reasonable manner based on the medium, means, and
297
+ context in which You Share Adapted Material.
298
+
299
+ 3. You may not offer or impose any additional or different terms
300
+ or conditions on, or apply any Effective Technological
301
+ Measures to, Adapted Material that restrict exercise of the
302
+ rights granted under the Adapter's License You apply.
303
+
304
+
305
+ Section 4 -- Sui Generis Database Rights.
306
+
307
+ Where the Licensed Rights include Sui Generis Database Rights that
308
+ apply to Your use of the Licensed Material:
309
+
310
+ a. for the avoidance of doubt, Section 2(a)(1) grants You the right
311
+ to extract, reuse, reproduce, and Share all or a substantial
312
+ portion of the contents of the database for NonCommercial purposes
313
+ only;
314
+
315
+ b. if You include all or a substantial portion of the database
316
+ contents in a database in which You have Sui Generis Database
317
+ Rights, then the database in which You have Sui Generis Database
318
+ Rights (but not its individual contents) is Adapted Material,
319
+ including for purposes of Section 3(b); and
320
+
321
+ c. You must comply with the conditions in Section 3(a) if You Share
322
+ all or a substantial portion of the contents of the database.
323
+
324
+ For the avoidance of doubt, this Section 4 supplements and does not
325
+ replace Your obligations under this Public License where the Licensed
326
+ Rights include other Copyright and Similar Rights.
327
+
328
+
329
+ Section 5 -- Disclaimer of Warranties and Limitation of Liability.
330
+
331
+ a. UNLESS OTHERWISE SEPARATELY UNDERTAKEN BY THE LICENSOR, TO THE
332
+ EXTENT POSSIBLE, THE LICENSOR OFFERS THE LICENSED MATERIAL AS-IS
333
+ AND AS-AVAILABLE, AND MAKES NO REPRESENTATIONS OR WARRANTIES OF
334
+ ANY KIND CONCERNING THE LICENSED MATERIAL, WHETHER EXPRESS,
335
+ IMPLIED, STATUTORY, OR OTHER. THIS INCLUDES, WITHOUT LIMITATION,
336
+ WARRANTIES OF TITLE, MERCHANTABILITY, FITNESS FOR A PARTICULAR
337
+ PURPOSE, NON-INFRINGEMENT, ABSENCE OF LATENT OR OTHER DEFECTS,
338
+ ACCURACY, OR THE PRESENCE OR ABSENCE OF ERRORS, WHETHER OR NOT
339
+ KNOWN OR DISCOVERABLE. WHERE DISCLAIMERS OF WARRANTIES ARE NOT
340
+ ALLOWED IN FULL OR IN PART, THIS DISCLAIMER MAY NOT APPLY TO YOU.
341
+
342
+ b. TO THE EXTENT POSSIBLE, IN NO EVENT WILL THE LICENSOR BE LIABLE
343
+ TO YOU ON ANY LEGAL THEORY (INCLUDING, WITHOUT LIMITATION,
344
+ NEGLIGENCE) OR OTHERWISE FOR ANY DIRECT, SPECIAL, INDIRECT,
345
+ INCIDENTAL, CONSEQUENTIAL, PUNITIVE, EXEMPLARY, OR OTHER LOSSES,
346
+ COSTS, EXPENSES, OR DAMAGES ARISING OUT OF THIS PUBLIC LICENSE OR
347
+ USE OF THE LICENSED MATERIAL, EVEN IF THE LICENSOR HAS BEEN
348
+ ADVISED OF THE POSSIBILITY OF SUCH LOSSES, COSTS, EXPENSES, OR
349
+ DAMAGES. WHERE A LIMITATION OF LIABILITY IS NOT ALLOWED IN FULL OR
350
+ IN PART, THIS LIMITATION MAY NOT APPLY TO YOU.
351
+
352
+ c. The disclaimer of warranties and limitation of liability provided
353
+ above shall be interpreted in a manner that, to the extent
354
+ possible, most closely approximates an absolute disclaimer and
355
+ waiver of all liability.
356
+
357
+
358
+ Section 6 -- Term and Termination.
359
+
360
+ a. This Public License applies for the term of the Copyright and
361
+ Similar Rights licensed here. However, if You fail to comply with
362
+ this Public License, then Your rights under this Public License
363
+ terminate automatically.
364
+
365
+ b. Where Your right to use the Licensed Material has terminated under
366
+ Section 6(a), it reinstates:
367
+
368
+ 1. automatically as of the date the violation is cured, provided
369
+ it is cured within 30 days of Your discovery of the
370
+ violation; or
371
+
372
+ 2. upon express reinstatement by the Licensor.
373
+
374
+ For the avoidance of doubt, this Section 6(b) does not affect any
375
+ right the Licensor may have to seek remedies for Your violations
376
+ of this Public License.
377
+
378
+ c. For the avoidance of doubt, the Licensor may also offer the
379
+ Licensed Material under separate terms or conditions or stop
380
+ distributing the Licensed Material at any time; however, doing so
381
+ will not terminate this Public License.
382
+
383
+ d. Sections 1, 5, 6, 7, and 8 survive termination of this Public
384
+ License.
385
+
386
+
387
+ Section 7 -- Other Terms and Conditions.
388
+
389
+ a. The Licensor shall not be bound by any additional or different
390
+ terms or conditions communicated by You unless expressly agreed.
391
+
392
+ b. Any arrangements, understandings, or agreements regarding the
393
+ Licensed Material not stated herein are separate from and
394
+ independent of the terms and conditions of this Public License.
395
+
396
+
397
+ Section 8 -- Interpretation.
398
+
399
+ a. For the avoidance of doubt, this Public License does not, and
400
+ shall not be interpreted to, reduce, limit, restrict, or impose
401
+ conditions on any use of the Licensed Material that could lawfully
402
+ be made without permission under this Public License.
403
+
404
+ b. To the extent possible, if any provision of this Public License is
405
+ deemed unenforceable, it shall be automatically reformed to the
406
+ minimum extent necessary to make it enforceable. If the provision
407
+ cannot be reformed, it shall be severed from this Public License
408
+ without affecting the enforceability of the remaining terms and
409
+ conditions.
410
+
411
+ c. No term or condition of this Public License will be waived and no
412
+ failure to comply consented to unless expressly agreed to by the
413
+ Licensor.
414
+
415
+ d. Nothing in this Public License constitutes or may be interpreted
416
+ as a limitation upon, or waiver of, any privileges and immunities
417
+ that apply to the Licensor or You, including from the legal
418
+ processes of any jurisdiction or authority.
419
+
420
+ =======================================================================
421
+
422
+ Creative Commons is not a party to its public
423
+ licenses. Notwithstanding, Creative Commons may elect to apply one of
424
+ its public licenses to material it publishes and in those instances
425
+ will be considered the "Licensor." The text of the Creative Commons
426
+ public licenses is dedicated to the public domain under the CC0 Public
427
+ Domain Dedication. Except for the limited purpose of indicating that
428
+ material is shared under a Creative Commons public license or as
429
+ otherwise permitted by the Creative Commons policies published at
430
+ creativecommons.org/policies, Creative Commons does not authorize the
431
+ use of the trademark "Creative Commons" or any other trademark or logo
432
+ of Creative Commons without its prior written consent including,
433
+ without limitation, in connection with any unauthorized modifications
434
+ to any of its public licenses or any other arrangements,
435
+ understandings, or agreements concerning use of licensed material. For
436
+ the avoidance of doubt, this paragraph does not form part of the
437
+ public licenses.
438
+
439
+ Creative Commons may be contacted at creativecommons.org.
LICENSE_SDXL ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Copyright (c) 2023 Stability AI
2
+ CreativeML Open RAIL++-M License dated July 26, 2023
3
+
4
+ Section I: PREAMBLE
5
+ Multimodal generative models are being widely adopted and used, and have the potential to transform the way artists, among other individuals, conceive and benefit from AI or ML technologies as a tool for content creation.
6
+ Notwithstanding the current and potential benefits that these artifacts can bring to society at large, there are also concerns about potential misuses of them, either due to their technical limitations or ethical considerations.
7
+ In short, this license strives for both the open and responsible downstream use of the accompanying model. When it comes to the open character, we took inspiration from open source permissive licenses regarding the grant of IP rights. Referring to the downstream responsible use, we added use-based restrictions not permitting the use of the model in very specific scenarios, in order for the licensor to be able to enforce the license in case potential misuses of the Model may occur. At the same time, we strive to promote open and responsible research on generative models for art and content generation.
8
+ Even though downstream derivative versions of the model could be released under different licensing terms, the latter will always have to include - at minimum - the same use-based restrictions as the ones in the original license (this license). We believe in the intersection between open and responsible AI development; thus, this agreement aims to strike a balance between both in order to enable responsible open-science in the field of AI.
9
+ This CreativeML Open RAIL++-M License governs the use of the model (and its derivatives) and is informed by the model card associated with the model.
10
+ NOW THEREFORE, You and Licensor agree as follows:
11
+ Definitions
12
+ "License" means the terms and conditions for use, reproduction, and Distribution as defined in this document.
13
+ "Data" means a collection of information and/or content extracted from the dataset used with the Model, including to train, pretrain, or otherwise evaluate the Model. The Data is not licensed under this License.
14
+ "Output" means the results of operating a Model as embodied in informational content resulting therefrom.
15
+ "Model" means any accompanying machine-learning based assemblies (including checkpoints), consisting of learnt weights, parameters (including optimizer states), corresponding to the model architecture as embodied in the Complementary Material, that have been trained or tuned, in whole or in part on the Data, using the Complementary Material.
16
+ "Derivatives of the Model" means all modifications to the Model, works based on the Model, or any other model which is created or initialized by transfer of patterns of the weights, parameters, activations or output of the Model, to the other model, in order to cause the other model to perform similarly to the Model, including - but not limited to - distillation methods entailing the use of intermediate data representations or methods based on the generation of synthetic data by the Model for training the other model.
17
+ "Complementary Material" means the accompanying source code and scripts used to define, run, load, benchmark or evaluate the Model, and used to prepare data for training or evaluation, if any. This includes any accompanying documentation, tutorials, examples, etc, if any.
18
+ "Distribution" means any transmission, reproduction, publication or other sharing of the Model or Derivatives of the Model to a third party, including providing the Model as a hosted service made available by electronic or other remote means - e.g. API-based or web access.
19
+ "Licensor" means the copyright owner or entity authorized by the copyright owner that is granting the License, including the persons or entities that may have rights in the Model and/or distributing the Model.
20
+ "You" (or "Your") means an individual or Legal Entity exercising permissions granted by this License and/or making use of the Model for whichever purpose and in any field of use, including usage of the Model in an end-use application - e.g. chatbot, translator, image generator.
21
+ "Third Parties" means individuals or legal entities that are not under common control with Licensor or You.
22
+ "Contribution" means any work of authorship, including the original version of the Model and any modifications or additions to that Model or Derivatives of the Model thereof, that is intentionally submitted to Licensor for inclusion in the Model 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 Model, but excluding communication that is conspicuously marked or otherwise designated in writing by the copyright owner as "Not a Contribution."
23
+ "Contributor" means Licensor and any individual or Legal Entity on behalf of whom a Contribution has been received by Licensor and subsequently incorporated within the Model.
24
+
25
+ Section II: INTELLECTUAL PROPERTY RIGHTS
26
+ Both copyright and patent grants apply to the Model, Derivatives of the Model and Complementary Material. The Model and Derivatives of the Model are subject to additional terms as described in
27
+
28
+ Section III.
29
+ 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, publicly display, publicly perform, sublicense, and distribute the Complementary Material, the Model, and Derivatives of the Model.
30
+ Grant of Patent License. Subject to the terms and conditions of this License and where and as applicable, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this paragraph) patent license to make, have made, use, offer to sell, sell, import, and otherwise transfer the Model and the Complementary Material, 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 Model 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 Model and/or Complementary Material or a Contribution incorporated within the Model and/or Complementary Material constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for the Model and/or Work shall terminate as of the date such litigation is asserted or filed.
31
+ Section III: CONDITIONS OF USAGE, DISTRIBUTION AND REDISTRIBUTION
32
+ Distribution and Redistribution. You may host for Third Party remote access purposes (e.g. software-as-a-service), reproduce and distribute copies of the Model or Derivatives of the Model thereof in any medium, with or without modifications, provided that You meet the following conditions: Use-based restrictions as referenced in paragraph 5 MUST be included as an enforceable provision by You in any type of legal agreement (e.g. a license) governing the use and/or distribution of the Model or Derivatives of the Model, and You shall give notice to subsequent users You Distribute to, that the Model or Derivatives of the Model are subject to paragraph 5. This provision does not apply to the use of Complementary Material. You must give any Third Party recipients of the Model or Derivatives of the Model a copy of this License; You must cause any modified files to carry prominent notices stating that You changed the files; You must retain all copyright, patent, trademark, and attribution notices excluding those notices that do not pertain to any part of the Model, Derivatives of the Model. You may add Your own copyright statement to Your modifications and may provide additional or different license terms and conditions - respecting paragraph 4.a. - for use, reproduction, or Distribution of Your modifications, or for any such Derivatives of the Model as a whole, provided Your use, reproduction, and Distribution of the Model otherwise complies with the conditions stated in this License.
33
+ Use-based restrictions. The restrictions set forth in Attachment A are considered Use-based restrictions. Therefore You cannot use the Model and the Derivatives of the Model for the specified restricted uses. You may use the Model subject to this License, including only for lawful purposes and in accordance with the License. Use may include creating any content with, finetuning, updating, running, training, evaluating and/or reparametrizing the Model. You shall require all of Your users who use the Model or a Derivative of the Model to comply with the terms of this paragraph (paragraph 5).
34
+ The Output You Generate. Except as set forth herein, Licensor claims no rights in the Output You generate using the Model. You are accountable for the Output you generate and its subsequent uses. No use of the output can contravene any provision as stated in the License.
35
+
36
+ Section IV: OTHER PROVISIONS
37
+ Updates and Runtime Restrictions. To the maximum extent permitted by law, Licensor reserves the right to restrict (remotely or otherwise) usage of the Model in violation of this License.
38
+ Trademarks and related. Nothing in this License permits You to make use of Licensors’ trademarks, trade names, logos or to otherwise suggest endorsement or misrepresent the relationship between the parties; and any rights not expressly granted herein are reserved by the Licensors.
39
+ Disclaimer of Warranty. Unless required by applicable law or agreed to in writing, Licensor provides the Model and the Complementary Material (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 Model, Derivatives of the Model, and the Complementary Material and assume any risks associated with Your exercise of permissions under this License.
40
+ 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 Model and the Complementary Material (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.
41
+ Accepting Warranty or Additional Liability. While redistributing the Model, Derivatives of the Model and the Complementary Material 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.
42
+ If any provision of this License is held to be invalid, illegal or unenforceable, the remaining provisions shall be unaffected thereby and remain valid as if such provision had not been set forth herein.
43
+
44
+ END OF TERMS AND CONDITIONS
45
+
46
+ Attachment A
47
+ Use Restrictions
48
+ You agree not to use the Model or Derivatives of the Model:
49
+ In any way that violates any applicable national, federal, state, local or international law or regulation;
50
+ For the purpose of exploiting, harming or attempting to exploit or harm minors in any way;
51
+ To generate or disseminate verifiably false information and/or content with the purpose of harming others;
52
+ To generate or disseminate personal identifiable information that can be used to harm an individual;
53
+ To defame, disparage or otherwise harass others;
54
+ For fully automated decision making that adversely impacts an individual’s legal rights or otherwise creates or modifies a binding, enforceable obligation;
55
+ For any use intended to or which has the effect of discriminating against or harming individuals or groups based on online or offline social behavior or known or predicted personal or personality characteristics;
56
+ To exploit any of the vulnerabilities of a specific group of persons based on their age, social, physical or mental characteristics, in order to materially distort the behavior of a person pertaining to that group in a manner that causes or is likely to cause that person or another person physical or psychological harm;
57
+ For any use intended to or which has the effect of discriminating against individuals or groups based on legally protected characteristics or categories;
58
+ To provide medical advice and medical results interpretation;
59
+ To generate or disseminate information for the purpose to be used for administration of justice, law enforcement, immigration or asylum processes, such as predicting an individual will commit fraud/crime commitment (e.g. by text profiling, drawing causal relationships between assertions made in documents, indiscriminate and arbitrarily-targeted use).
60
+
LICENSE_Wan2.1 ADDED
@@ -0,0 +1,201 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Apache License
2
+ Version 2.0, January 2004
3
+ http://www.apache.org/licenses/
4
+
5
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
6
+
7
+ 1. Definitions.
8
+
9
+ "License" shall mean the terms and conditions for use, reproduction,
10
+ and distribution as defined by Sections 1 through 9 of this document.
11
+
12
+ "Licensor" shall mean the copyright owner or entity authorized by
13
+ the copyright owner that is granting the License.
14
+
15
+ "Legal Entity" shall mean the union of the acting entity and all
16
+ other entities that control, are controlled by, or are under common
17
+ control with that entity. For the purposes of this definition,
18
+ "control" means (i) the power, direct or indirect, to cause the
19
+ direction or management of such entity, whether by contract or
20
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
21
+ outstanding shares, or (iii) beneficial ownership of such entity.
22
+
23
+ "You" (or "Your") shall mean an individual or Legal Entity
24
+ exercising permissions granted by this License.
25
+
26
+ "Source" form shall mean the preferred form for making modifications,
27
+ including but not limited to software source code, documentation
28
+ source, and configuration files.
29
+
30
+ "Object" form shall mean any form resulting from mechanical
31
+ transformation or translation of a Source form, including but
32
+ not limited to compiled object code, generated documentation,
33
+ and conversions to other media types.
34
+
35
+ "Work" shall mean the work of authorship, whether in Source or
36
+ Object form, made available under the License, as indicated by a
37
+ copyright notice that is included in or attached to the work
38
+ (an example is provided in the Appendix below).
39
+
40
+ "Derivative Works" shall mean any work, whether in Source or Object
41
+ form, that is based on (or derived from) the Work and for which the
42
+ editorial revisions, annotations, elaborations, or other modifications
43
+ represent, as a whole, an original work of authorship. For the purposes
44
+ of this License, Derivative Works shall not include works that remain
45
+ separable from, or merely link (or bind by name) to the interfaces of,
46
+ the Work and Derivative Works thereof.
47
+
48
+ "Contribution" shall mean any work of authorship, including
49
+ the original version of the Work and any modifications or additions
50
+ to that Work or Derivative Works thereof, that is intentionally
51
+ submitted to Licensor for inclusion in the Work by the copyright owner
52
+ or by an individual or Legal Entity authorized to submit on behalf of
53
+ the copyright owner. For the purposes of this definition, "submitted"
54
+ means any form of electronic, verbal, or written communication sent
55
+ to the Licensor or its representatives, including but not limited to
56
+ communication on electronic mailing lists, source code control systems,
57
+ and issue tracking systems that are managed by, or on behalf of, the
58
+ Licensor for the purpose of discussing and improving the Work, but
59
+ excluding communication that is conspicuously marked or otherwise
60
+ designated in writing by the copyright owner as "Not a Contribution."
61
+
62
+ "Contributor" shall mean Licensor and any individual or Legal Entity
63
+ on behalf of whom a Contribution has been received by Licensor and
64
+ subsequently incorporated within the Work.
65
+
66
+ 2. Grant of Copyright License. Subject to the terms and conditions of
67
+ this License, each Contributor hereby grants to You a perpetual,
68
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
69
+ copyright license to reproduce, prepare Derivative Works of,
70
+ publicly display, publicly perform, sublicense, and distribute the
71
+ Work and such Derivative Works in Source or Object form.
72
+
73
+ 3. Grant of Patent License. Subject to the terms and conditions of
74
+ this License, each Contributor hereby grants to You a perpetual,
75
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
76
+ (except as stated in this section) patent license to make, have made,
77
+ use, offer to sell, sell, import, and otherwise transfer the Work,
78
+ where such license applies only to those patent claims licensable
79
+ by such Contributor that are necessarily infringed by their
80
+ Contribution(s) alone or by combination of their Contribution(s)
81
+ with the Work to which such Contribution(s) was submitted. If You
82
+ institute patent litigation against any entity (including a
83
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
84
+ or a Contribution incorporated within the Work constitutes direct
85
+ or contributory patent infringement, then any patent licenses
86
+ granted to You under this License for that Work shall terminate
87
+ as of the date such litigation is filed.
88
+
89
+ 4. Redistribution. You may reproduce and distribute copies of the
90
+ Work or Derivative Works thereof in any medium, with or without
91
+ modifications, and in Source or Object form, provided that You
92
+ meet the following conditions:
93
+
94
+ (a) You must give any other recipients of the Work or
95
+ Derivative Works a copy of this License; and
96
+
97
+ (b) You must cause any modified files to carry prominent notices
98
+ stating that You changed the files; and
99
+
100
+ (c) You must retain, in the Source form of any Derivative Works
101
+ that You distribute, all copyright, patent, trademark, and
102
+ attribution notices from the Source form of the Work,
103
+ excluding those notices that do not pertain to any part of
104
+ the Derivative Works; and
105
+
106
+ (d) If the Work includes a "NOTICE" text file as part of its
107
+ distribution, then any Derivative Works that You distribute must
108
+ include a readable copy of the attribution notices contained
109
+ within such NOTICE file, excluding those notices that do not
110
+ pertain to any part of the Derivative Works, in at least one
111
+ of the following places: within a NOTICE text file distributed
112
+ as part of the Derivative Works; within the Source form or
113
+ documentation, if provided along with the Derivative Works; or,
114
+ within a display generated by the Derivative Works, if and
115
+ wherever such third-party notices normally appear. The contents
116
+ of the NOTICE file are for informational purposes only and
117
+ do not modify the License. You may add Your own attribution
118
+ notices within Derivative Works that You distribute, alongside
119
+ or as an addendum to the NOTICE text from the Work, provided
120
+ that such additional attribution notices cannot be construed
121
+ as modifying the License.
122
+
123
+ You may add Your own copyright statement to Your modifications and
124
+ may provide additional or different license terms and conditions
125
+ for use, reproduction, or distribution of Your modifications, or
126
+ for any such Derivative Works as a whole, provided Your use,
127
+ reproduction, and distribution of the Work otherwise complies with
128
+ the conditions stated in this License.
129
+
130
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
131
+ any Contribution intentionally submitted for inclusion in the Work
132
+ by You to the Licensor shall be under the terms and conditions of
133
+ this License, without any additional terms or conditions.
134
+ Notwithstanding the above, nothing herein shall supersede or modify
135
+ the terms of any separate license agreement you may have executed
136
+ with Licensor regarding such Contributions.
137
+
138
+ 6. Trademarks. This License does not grant permission to use the trade
139
+ names, trademarks, service marks, or product names of the Licensor,
140
+ except as required for reasonable and customary use in describing the
141
+ origin of the Work and reproducing the content of the NOTICE file.
142
+
143
+ 7. Disclaimer of Warranty. Unless required by applicable law or
144
+ agreed to in writing, Licensor provides the Work (and each
145
+ Contributor provides its Contributions) on an "AS IS" BASIS,
146
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
147
+ implied, including, without limitation, any warranties or conditions
148
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
149
+ PARTICULAR PURPOSE. You are solely responsible for determining the
150
+ appropriateness of using or redistributing the Work and assume any
151
+ risks associated with Your exercise of permissions under this License.
152
+
153
+ 8. Limitation of Liability. In no event and under no legal theory,
154
+ whether in tort (including negligence), contract, or otherwise,
155
+ unless required by applicable law (such as deliberate and grossly
156
+ negligent acts) or agreed to in writing, shall any Contributor be
157
+ liable to You for damages, including any direct, indirect, special,
158
+ incidental, or consequential damages of any character arising as a
159
+ result of this License or out of the use or inability to use the
160
+ Work (including but not limited to damages for loss of goodwill,
161
+ work stoppage, computer failure or malfunction, or any and all
162
+ other commercial damages or losses), even if such Contributor
163
+ has been advised of the possibility of such damages.
164
+
165
+ 9. Accepting Warranty or Additional Liability. While redistributing
166
+ the Work or Derivative Works thereof, You may choose to offer,
167
+ and charge a fee for, acceptance of support, warranty, indemnity,
168
+ or other liability obligations and/or rights consistent with this
169
+ License. However, in accepting such obligations, You may act only
170
+ on Your own behalf and on Your sole responsibility, not on behalf
171
+ of any other Contributor, and only if You agree to indemnify,
172
+ defend, and hold each Contributor harmless for any liability
173
+ incurred by, or claims asserted against, such Contributor by reason
174
+ of your accepting any such warranty or additional liability.
175
+
176
+ END OF TERMS AND CONDITIONS
177
+
178
+ APPENDIX: How to apply the Apache License to your work.
179
+
180
+ To apply the Apache License to your work, attach the following
181
+ boilerplate notice, with the fields enclosed by brackets "[]"
182
+ replaced with your own identifying information. (Don't include
183
+ the brackets!) The text should be enclosed in the appropriate
184
+ comment syntax for the file format. We also recommend that a
185
+ file or class name and description of purpose be included on the
186
+ same "printed page" as the copyright notice for easier
187
+ identification within third-party archives.
188
+
189
+ Copyright [yyyy] [name of copyright owner]
190
+
191
+ Licensed under the Apache License, Version 2.0 (the "License");
192
+ you may not use this file except in compliance with the License.
193
+ You may obtain a copy of the License at
194
+
195
+ http://www.apache.org/licenses/LICENSE-2.0
196
+
197
+ Unless required by applicable law or agreed to in writing, software
198
+ distributed under the License is distributed on an "AS IS" BASIS,
199
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
200
+ See the License for the specific language governing permissions and
201
+ limitations under the License.
NOTICE ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ MiniMax H3 is licensed under the MiniMax H3 Community License Agreement, Copyright © 2026 MiniMax. All Rights Reserved.
2
+
3
+ The DMAD checkpoints in this repository are Model Derivatives of MiniMax H3: the MiniMax-H3 text-to-audio-video transformer was modified by LoRA fine-tuning (DMAD distillation to 4 steps). They are distributed under the same agreement (see LICENSE).
README.md ADDED
@@ -0,0 +1,275 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: minimax-h3-community-license-agreement
4
+ license_link: LICENSE
5
+ base_model: MiniMaxAI/MiniMax-H3
6
+ pipeline_tag: text-to-video
7
+ library_name: minimax-h3
8
+ tags:
9
+ - diffusers
10
+ - text-to-audio-video
11
+ - audio-video-generation
12
+ - distillation
13
+ - lora
14
+ - minimax-h3
15
+ ---
16
+
17
+ # DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
18
+
19
+ ### 4-step & 8-step MiniMax-H3 students for joint audio-video generation
20
+
21
+ [![Project Page](https://img.shields.io/badge/Project-Page-yellow?logo=googlechrome&logoColor=yellow)](https://yzmblog.github.io/projects/DMAD/)
22
+ [![Paper](https://img.shields.io/badge/Paper-arXiv-b31b1b?logo=arxiv&logoColor=red)](https://arxiv.org/abs/2610.02188)
23
+ [![Code](https://img.shields.io/badge/Code-GitHub-black?logo=github)](https://github.com/Yzmblog/DMAD)
24
+ [![ModelScope](https://img.shields.io/badge/ModelScope-DMAD-624aff?logo=modelscope)](https://modelscope.ai/models/ZhengmingYu-Community/DMAD)
25
+ [![ModelScope CN](https://img.shields.io/badge/ModelScope_CN-DMAD-624aff?logo=modelscope)](https://modelscope.cn/models/ZhengmingYu-Community/DMAD)
26
+ [![Demo Video](https://img.shields.io/badge/Demo-Video-red?logo=youtube&logoColor=white)](https://www.youtube.com/watch?v=cOCUCYZzAtE)
27
+
28
+ [Zhengming Yu](https://yzmblog.github.io/)<sup>1,2</sup>,
29
+ [Junkun Yuan](https://junkunyuan.github.io/)<sup>2</sup>,
30
+ [Haotian Yang](https://yanght321.github.io/)<sup>2</sup>,
31
+ [Gordon Guocheng Qian](https://guochengqian.github.io/)<sup>2</sup>,
32
+ [Yizhi Wang](https://yizhiwang96.github.io/)<sup>2</sup>,
33
+ [Angtian Wang](https://angtianwang.github.io/)<sup>2</sup>,
34
+ [Yiding Yang](https://ihollywhy.github.io/)<sup>2</sup>,
35
+ [Bo Liu](https://scholar.google.com/citations?user=NOgz-HsAAAAJ&hl=en)<sup>2</sup>,
36
+ [Xin Li](https://people.tamu.edu/~xinli/)<sup>1</sup>,
37
+ [Wenping Wang](https://engineering.tamu.edu/cse/profiles/Wang-Wenping.html)<sup>1</sup>,
38
+ [Chongyang Ma](http://www.chongyangma.com/)<sup>2</sup><br/>
39
+ <sup>1</sup>Texas A&amp;M University, <sup>2</sup>ByteDance<br/>
40
+
41
+ <p align="center">
42
+ <img src="https://huggingface.co/ZhengmingYu/DMAD/resolve/main/assets/teaser.gif" alt="Videos generated by the 4-step DMAD student of MiniMax-H3 (video only; every clip also has generated audio)">
43
+ </p>
44
+
45
+ This repository holds the DMAD students of the paper: 4-step students of **[MiniMax-H3](https://huggingface.co/MiniMaxAI/MiniMax-H3)**
46
+ (33B, text-to-audio-video) and of **Wan2.1-T2V** (1.3B and 14B), 4- and 1-step students of **SDXL**, and 1-step
47
+ students of the **EDM ImageNet-64** teacher. The inference and training code is in the
48
+ [code repository](https://github.com/Yzmblog/DMAD) (`train/h3`, `train/wan`, `train/image`).
49
+
50
+ ## Updates
51
+
52
+ * **2026/10/29:** New 8-step student [`8step_full_critic`](#minimax-h3-text-to-audio-video) 🚀 Freshly trained for 8-step sampling,
53
+ it brings a clear step up in audio-video quality over our 4-step students. Ships with a ComfyUI-layout LoRA: drop it into the
54
+ [8-step workflow](#ready-to-run-workflows-8-gb-gpu-and-up). Feel free to try it out! 🎛️
55
+ * **2026/10/15:** The [Wan2.1 training data](https://huggingface.co/datasets/ZhengmingYu/DMAD-Wan-data) (42,158
56
+ Wan2.1-14B teacher samples with their real-video latents), so `train/wan` can start training right away 🤗
57
+ * **2026/10/10:** 8 GB GPUs: a 5 s [ComfyUI workflow](#ready-to-run-workflows-8-gb-gpu-and-up) for 8 GB GPUs, and
58
+ `--low-vram` in the [code repository](https://github.com/Yzmblog/DMAD#on-consumer-gpus) now makes 5 s videos in under
59
+ 7 GiB of GPU memory and 15 s videos in 12 GiB 🪶
60
+ * **2026/10/08:** 16 GB GPUs: the ComfyUI workflows and `--low-vram` fit in 14 GiB of GPU memory, bit-identical to the
61
+ larger-GPU runs 🪶
62
+ * **2026/10/07:** Ready-to-run [ComfyUI workflows](#ready-to-run-workflows-8-gb-gpu-and-up) (4 and 8 steps, 15 s of
63
+ video with audio) in `minimax_h3/workflows` 🎛️
64
+ * **2026/10/06:** The DMAD weights of Wan2.1 (1.3B, 14B), SDXL and ImageNet-64, and the
65
+ [MiniMax-H3 training data](https://huggingface.co/datasets/ZhengmingYu/DMAD-H3-data) 🤗
66
+ * **2026/10/05:** [ComfyUI-layout LoRAs](#comfyui) of the MiniMax-H3 students, and the training code in the
67
+ [code repository](https://github.com/Yzmblog/DMAD) 🏋️
68
+ * **2026/10/02:** The 4-step MiniMax-H3 students (`lora_critic`, `full_critic`) released 🚀
69
+
70
+ ## MiniMax-H3 (text-to-audio-video)
71
+
72
+ Rank-128 LoRAs on the H3 transformer that turn the 50-step teacher into a 4-step (or, with `8step_full_critic`, an
73
+ 8-step) generator of 1344x768 video with native stereo audio.
74
+
75
+ | File | Checkpoint | Size |
76
+ |------|------------|------|
77
+ | `minimax_h3/dmad_minimax_h3_4step_lora_critic.safetensors` | the checkpoint of the paper: EMA of the student at iteration 800 of the main run | 1.4 GB |
78
+ | `minimax_h3/dmad_minimax_h3_4step_full_critic.safetensors` | the student of a run whose critic backbone is fully trained (the paper's run keeps it frozen under a LoRA): iteration 1600, live weights; it scores higher on AVGen-Bench | 1.4 GB |
79
+ | `minimax_h3/dmad_minimax_h3_8step_full_critic.safetensors` | the 8-step student: the `full_critic` recipe trained with 8-step sampling from scratch. Sample it with 8 steps (`--steps 8`) | 1.4 GB |
80
+ | `minimax_h3/dmad_minimax_h3_4step_lora_critic_comfyui.safetensors` | `lora_critic` in ComfyUI's MiniMax-H3 key layout (exact conversion) | 2.0 GB |
81
+ | `minimax_h3/dmad_minimax_h3_4step_full_critic_comfyui.safetensors` | `full_critic` in ComfyUI's MiniMax-H3 key layout (exact conversion) | 2.0 GB |
82
+ | `minimax_h3/dmad_minimax_h3_8step_full_critic_comfyui.safetensors` | `8step_full_critic` in ComfyUI's MiniMax-H3 key layout (exact conversion); 8 steps in the scheduler | 2.0 GB |
83
+
84
+ LoRA layout of the first three: Diffusers keys (`<module>.lora.down.weight` = A `[128, in]`, `<module>.lora.up.weight` = B
85
+ `[out, 128]`) over `attn.to_q/to_k/to_v/to_out.0`, `ff.net.0.proj`, `ff.net.2` of all 50 transformer blocks and the 2
86
+ token-refiner blocks (312 modules). alpha = rank = 128. The safetensors metadata repeats this.
87
+
88
+ The inference code lives in the [code repository](https://github.com/Yzmblog/DMAD): `inference.py` with the sampler
89
+ the paper used (re-noise step rule) and a Diffusers-pipeline example; its README covers the environment. Sampling
90
+ settings: 4 steps (8 for `8step_full_critic`: `--steps 8`), time shift 12 (video) and 2 (audio), no classifier-free guidance, 124 frames at 24 fps. `--low-vram` runs it on an 8 GB GPU (5 s; 15 s needs 12 GB; `--low-vram 8 / 12 / 16 / 24` picks the settings for your GPU, same output).
91
+
92
+ ```bash
93
+ git clone https://github.com/Yzmblog/DMAD.git && cd DMAD # code + environment setup (see its README)
94
+ hf download MiniMaxAI/MiniMax-H3 --local-dir models/MiniMax-H3 --exclude "FL2VA/*" --exclude "Ref2VA/*" --exclude "transformer_ref/*"
95
+ hf download ZhengmingYu/DMAD --include "minimax_h3/*_critic.safetensors" --local-dir ckpt
96
+ python inference.py --model-dir models/MiniMax-H3 --lora ckpt/minimax_h3/dmad_minimax_h3_4step_lora_critic.safetensors \
97
+ --prompt-file prompts/dmad_sweater.txt --seed 42 --output-dir outputs/dmad_sweater
98
+ ```
99
+
100
+ ## ComfyUI
101
+
102
+ `minimax_h3/dmad_minimax_h3_{4step_lora_critic,4step_full_critic,8step_full_critic}_comfyui.safetensors` are the same
103
+ LoRAs converted exactly to ComfyUI's MiniMax-H3 key layout (rank 128; q/k/v fused into `attn.qkv_proj` adapters of rank
104
+ 384, alpha = rank, scale 1.0 — no rank reduction; 2.0 GB each). Load with `LoraLoaderModelOnly` at strength 1.0, cfg 1.0,
105
+ `ModelSamplingMiniMaxH3` with shift 12 / audio shift 2, and sample with ComfyUI's `lcm` sampler and `simple` scheduler
106
+ (the re-noise multistep rule and sigma grid the students were trained with; 4 steps for the 4-step students, 8 for
107
+ `8step_full_critic`; ODE samplers such as euler are not their operating point), or with the equivalent
108
+ **DMAD Sampler** + **DMAD Sigmas** nodes from
109
+ [`comfyui/ComfyUI-DMAD`](https://github.com/Yzmblog/DMAD/tree/main/comfyui/ComfyUI-DMAD).
110
+
111
+ ```bash
112
+ wget -P /path/to/ComfyUI/models/loras https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/dmad_minimax_h3_4step_full_critic_comfyui.safetensors
113
+ wget -P /path/to/ComfyUI/models/loras https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/dmad_minimax_h3_4step_lora_critic_comfyui.safetensors
114
+ wget -P /path/to/ComfyUI/models/loras https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/dmad_minimax_h3_8step_full_critic_comfyui.safetensors
115
+ ```
116
+
117
+ ### Ready-to-run workflows (8 GB GPU and up)
118
+
119
+ Complete text-to-audio-video workflows in [`minimax_h3/workflows`](https://huggingface.co/ZhengmingYu/DMAD/tree/main/minimax_h3/workflows):
120
+ two that make 15 s of 1344x768 video with stereo audio on a 16 GB GPU or larger, and a 5 s one for 8 GB GPUs. Load the
121
+ `.json`, or drag the example `.mp4` (it embeds the workflow) into ComfyUI; missing models are offered for download from
122
+ links stored in the workflow. All sample with stock nodes (`lcm` + `simple`, `full_critic` LoRA); the ComfyUI-DMAD nodes
123
+ are not needed. To run the 8-step student `8step_full_critic` instead, pick
124
+ `dmad_minimax_h3_8step_full_critic_comfyui.safetensors` in `LoraLoaderModelOnly` of the 8-step workflow (its scheduler
125
+ is already at 8 steps); everything else stays the same.
126
+
127
+ | GPU | workflow | video |
128
+ |---|---|---|
129
+ | 24 GB | [`dmad_h3_4step_15s_podcast.json`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_4step_15s_podcast.json) (4 steps) or [`dmad_h3_8step_15s_wok.json`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_8step_15s_wok.json) (8 steps) | 15 s |
130
+ | 16 GB, 22 GB | the same two workflows, unchanged (see below) | 15 s |
131
+ | 8 GB | [`dmad_h3_4step_5s_8gb_guitar.json`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.json) (see below) | 5 s |
132
+
133
+ The 24 GB and 16 GB rows share the workflow files: ComfyUI keeps as much of the model on the GPU as fits and streams
134
+ the rest from system memory, so the same workflow adapts to the card (and gives the same video).
135
+
136
+ | | 4 steps | 8 steps |
137
+ |---|---|---|
138
+ | Workflow | [`dmad_h3_4step_15s_podcast.json`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_4step_15s_podcast.json) | [`dmad_h3_8step_15s_wok.json`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_8step_15s_wok.json) |
139
+ | Example (workflow embedded) | [`dmad_h3_4step_15s_podcast.mp4`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_4step_15s_podcast.mp4), seed 2 | [`dmad_h3_8step_15s_wok.mp4`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_8step_15s_wok.mp4), seed 4 |
140
+ | Output | 1344x768, 362 frames = 15 s at 24 fps, stereo audio | 1344x768, 362 frames = 15 s at 24 fps, stereo audio |
141
+ | Peak GPU memory, 24 GiB cap | 24.6 GiB | 24.6 GiB |
142
+ | Time, 24 GiB cap | 448 s, of which sampling 393 s | 845 s, of which sampling 785 s |
143
+ | Peak GPU memory, 14 GiB cap (16 GB GPUs) | 14.7 GiB | 14.7 GiB |
144
+ | Time, 14 GiB cap | 470 s, of which sampling 413 s | 886 s, of which sampling 832 s |
145
+
146
+ <p align="center">
147
+ <img src="https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/screenshot_4step_15s.png" alt="The 4-step workflow in ComfyUI" width="100%">
148
+ </p>
149
+
150
+ | folder | file (from [Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3) unless noted) |
151
+ |---|---|
152
+ | `models/diffusion_models/` | `minimax_h3_fl2va_pruned_int8_convrot.safetensors` |
153
+ | `models/text_encoders/` | `qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors` |
154
+ | `models/vae/` | `minimax_h3_video_vae_fp16.safetensors`, `minimax_h3_audio_vae_fp32.safetensors` |
155
+ | `models/loras/` | `dmad_minimax_h3_4step_full_critic_comfyui.safetensors` (this repository) |
156
+
157
+ Video lengths are 5 + 17k frames (124 = 5 s, 243 = 10 s, 362 = 15 s). On a 16 or 24 GB GPU, 15 s needs the
158
+ `H3 Memory Optimization` node of [H3-Optimizations](https://github.com/Zironic/H3-Optimizations) (install with
159
+ ComfyUI-Manager), which both workflows include. Without it, 5 s (length 124) fits in 16 GB with stock nodes only, and
160
+ 15 s needs more than 32 GB (it fits in 40 GB); the video keeps its composition and action but is not bit-identical to
161
+ the one made with the memory node. Times are end to end (model loading, text encoding, sampling, decoding) on an H200
162
+ with the PyTorch allocator capped at 24 or 14 GiB; a consumer GPU is slower, and sampling time scales linearly with steps.
163
+
164
+ **16 GB and 22 GB GPUs.** Both workflows also run unchanged with the GPU memory capped at 14 GiB (room for a 16 GB
165
+ card's desktop and CUDA context): ComfyUI's dynamic VRAM keeps less of the model resident and streams more of it from
166
+ system memory, and the videos are bit-identical to the 24 GiB runs. About 46 GiB of system memory is in use, so 64 GB
167
+ of RAM is recommended. Outside ComfyUI, `inference.py --low-vram` in the
168
+ [code repository](https://github.com/Yzmblog/DMAD) makes 5 s videos in under 7 GiB and 15 s videos in 12 GiB of GPU memory,
169
+ bit-identical to its larger-GPU runs.
170
+
171
+ **8 GB GPUs (5 s).** [`dmad_h3_4step_5s_8gb_guitar.json`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.json) is the 4-step
172
+ workflow set up for 8 GB GPUs: 5 s instead of 15 s, the `H3 Memory Optimization` node in its lowest-memory setting
173
+ (`Attention memory mode` Lower VRAM (slower), `Activation chunk rows` 1024), and the `H3 AIMDO Residency Limiter` node of
174
+ the same pack (`VBAR residency budget` 0 blocks), which streams every transformer block from system memory. Models and
175
+ sampling are the same as above.
176
+
177
+ | | 4 steps, 5 s |
178
+ |---|---|
179
+ | Workflow | [`dmad_h3_4step_5s_8gb_guitar.json`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.json) |
180
+ | Example (workflow embedded) | [`dmad_h3_4step_5s_8gb_guitar.mp4`](https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.mp4), seed 1 |
181
+ | Output | 1344x768, 124 frames = 5 s at 24 fps, stereo audio |
182
+ | Peak GPU memory, 7 GiB cap (8 GB GPUs) | 7.7 GiB |
183
+ | Peak GPU memory, 5 GiB cap | 5.9 GiB |
184
+ | Time (either cap) | about 210 s, of which sampling 152 s |
185
+
186
+ The video is the same at both caps, bit for bit; the peaks include memory ComfyUI's dynamic VRAM allocates outside the
187
+ PyTorch allocator. About 45 GiB of system memory is in use (64 GB of RAM recommended). The memory settings change the
188
+ attention arithmetic, so the same prompt and seed do not reproduce the video the 16 GB+ settings make. 15 s does not
189
+ fit in 8 GB.
190
+
191
+ ## Wan2.1 (text-to-video)
192
+
193
+ Full generators (EMA, spectral norm folded in) in the `.pth` format of
194
+ [`train/wan`](https://github.com/Yzmblog/DMAD/tree/main/train/wan): 4 steps, 480p, 81 frames, no classifier-free
195
+ guidance.
196
+
197
+ | File | Model | Size |
198
+ |------|-------|------|
199
+ | `wan2.1/dmad_wan2pt1_1pt3B.pth` | Wan2.1-T2V-1.3B student, iteration 17k | 2.8 GB |
200
+ | `wan2.1/dmad_wan2pt1_14B.pth` | Wan2.1-T2V-14B student, iteration 19.5k | 29 GB |
201
+
202
+ ```bash
203
+ hf download ZhengmingYu/DMAD --include "wan2.1/*" --local-dir ckpt
204
+ bash experiments/dmad/sample.sh 1.3B ckpt/wan2.1/dmad_wan2pt1_1pt3B.pth my_prompts.json outputs/samples # in train/wan
205
+ ```
206
+
207
+ The training data of both students (42,158 rows: Wan2.1-14B teacher sample, umT5 embedding and the caption-paired
208
+ UltraVideo clip latent, 531 GB of WebDataset shards) is at
209
+ [ZhengmingYu/DMAD-Wan-data](https://huggingface.co/datasets/ZhengmingYu/DMAD-Wan-data).
210
+
211
+ ## SDXL (text-to-image)
212
+
213
+ UNet state dicts in fp16 (spectral norm folded in) that load into the standard SDXL pipeline; see
214
+ [`train/image`](https://github.com/Yzmblog/DMAD/tree/main/train/image) for the sampling code.
215
+
216
+ | File | Model | Size |
217
+ |------|-------|------|
218
+ | `sdxl/dmad_sdxl_4step_unet_fp16.bin` | 4-step student (backward simulation), iteration 17k | 5.1 GB |
219
+ | `sdxl/dmad_sdxl_1step_unet_fp16.bin` | 1-step student (ODE init), iteration 49k | 5.1 GB |
220
+ | `sdxl/dmad_sdxl_1step_frozencritic_unet_fp16.bin` | 1-step student (ODE init, frozen critic backbone), iteration 17.5k | 5.1 GB |
221
+
222
+ ```python
223
+ import torch
224
+ from diffusers import DiffusionPipeline, LCMScheduler, UNet2DConditionModel
225
+ from huggingface_hub import hf_hub_download
226
+
227
+ base_model_id = "stabilityai/stable-diffusion-xl-base-1.0"
228
+ unet = UNet2DConditionModel.from_config(base_model_id, subfolder="unet").to("cuda", torch.float16)
229
+ unet.load_state_dict(torch.load(hf_hub_download("ZhengmingYu/DMAD", "sdxl/dmad_sdxl_4step_unet_fp16.bin"), map_location="cuda"))
230
+ pipe = DiffusionPipeline.from_pretrained(base_model_id, unet=unet, torch_dtype=torch.float16, variant="fp16").to("cuda")
231
+ pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
232
+ image = pipe(prompt="a photo of a cat", num_inference_steps=4, guidance_scale=0, timesteps=[999, 749, 499, 249]).images[0]
233
+ # 1-step models: num_inference_steps=1, timesteps=[399]
234
+ ```
235
+
236
+ ## ImageNet-64 (class-conditional)
237
+
238
+ 1-step EMA generators of the EDM ImageNet-64 teacher, one folder per critic setting of the paper, in the
239
+ `checkpoint_model_<iteration>/pytorch_model_ema.bin` layout that `train/image`'s evaluation reads directly.
240
+
241
+ | Folder | Setting | Size |
242
+ |--------|---------|------|
243
+ | `imagenet64/dmad_imagenet_gaproute/` | teacher-UNet critic + gap routing, iteration 568k | 1.2 GB |
244
+ | `imagenet64/dmad_imagenet_pgcritic/` | pretrained-feature critic, iteration 108k | 1.2 GB |
245
+ | `imagenet64/dmad_imagenet_frozencritic/` | frozen teacher-UNet critic + gap routing, iteration 221k | 1.2 GB |
246
+
247
+ ```bash
248
+ hf download ZhengmingYu/DMAD --include "imagenet64/dmad_imagenet_gaproute/*" --local-dir ckpt
249
+ python main/edm/test_folder_edm.py --folder ckpt/imagenet64/dmad_imagenet_gaproute --run_once ... # in train/image
250
+ ```
251
+
252
+ ## Citation
253
+
254
+ ```bibtex
255
+ @article{yu2026dmad,
256
+ title={DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation},
257
+ author={Yu, Zhengming and Yuan, Junkun and Yang, Haotian and Qian, Gordon Guocheng and Wang, Yizhi and Wang, Angtian and Yang, Yiding and Liu, Bo and Li, Xin and Wang, Wenping and Ma, Chongyang},
258
+ journal={arXiv preprint arXiv:2610.02188},
259
+ year={2026}
260
+ }
261
+ ```
262
+
263
+ ## License
264
+
265
+ Each family of weights is a derivative of its base model and is distributed under that model's license:
266
+
267
+ * `minimax_h3/`: Model Derivatives of MiniMax H3, under the [MiniMax H3 Community License Agreement](LICENSE)
268
+ (see also [NOTICE](NOTICE)). The weights were modified from MiniMax-H3 by LoRA fine-tuning.
269
+ * `wan2.1/`: derived from Wan2.1-T2V-1.3B / 14B, under the [Apache License 2.0](LICENSE_Wan2.1).
270
+ * `sdxl/`: derived from Stable Diffusion XL base 1.0, under the
271
+ [CreativeML Open RAIL++-M License](LICENSE_SDXL), including its use-based restrictions.
272
+ * `imagenet64/`: derived from the NVIDIA EDM ImageNet-64 model, under
273
+ [CC BY-NC-SA 4.0](LICENSE_EDM) (non-commercial).
274
+
275
+ Images and videos produced with these weights are AI-generated.
assets/teaser.gif ADDED

Git LFS Details

  • SHA256: ff8c375e432e677369a78539c4a85846aff9afcce6529e669b9b5cc78a077822
  • Pointer size: 133 Bytes
  • Size of remote file: 15.3 MB
imagenet64/dmad_imagenet_frozencritic/checkpoint_model_221000/pytorch_model_ema.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04b895acbca554fd2f7a12cba5f6c364762e39b7babb62c84156952268363ff8
3
+ size 1184538562
imagenet64/dmad_imagenet_gaproute/checkpoint_model_568000/pytorch_model_ema.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99c76505fb66f89f4cd1d202c5aa1b62cbb57014420f82653fa3e03976fc8e36
3
+ size 1184538562
imagenet64/dmad_imagenet_pgcritic/checkpoint_model_108000/pytorch_model_ema.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd4569f7152e7d4be48c30fdc366a19464c896a10711f00f224be8f97c0422fd
3
+ size 1184538562
minimax_h3/dmad_minimax_h3_4step_full_critic.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac23eaef3a359666610769e4ef5f8af9530262c265f25ac211b0b54e51efc91e
3
+ size 1383674320
minimax_h3/dmad_minimax_h3_4step_full_critic_comfyui.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d3ee9f599ef1ac2d76850e90372ce9064fdc2c16e7f3de6a7ba9bc90cb91eb1b
3
+ size 1956195976
minimax_h3/dmad_minimax_h3_4step_lora_critic.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61d0865d6bf426e328caeab6610a28a677c68e9f2f04c607657c6451e38b5980
3
+ size 1383674304
minimax_h3/dmad_minimax_h3_4step_lora_critic_comfyui.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a4f54d9eb4eed9102be1adea8376e8e933cc2805f4017b95c14586290395b3a7
3
+ size 1956195960
minimax_h3/dmad_minimax_h3_8step_full_critic.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cc176f1defbb70d14fec2a7e27d397f60279d0239656c0b11d1697705462395
3
+ size 1383674424
minimax_h3/dmad_minimax_h3_8step_full_critic_comfyui.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:acb2c015ca171eb175165fca962ba29e2ace8a78f6ce015017561c4b78474933
3
+ size 1956196016
minimax_h3/workflows/dmad_h3_4step_15s_podcast.json ADDED
@@ -0,0 +1,1167 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "id": "dmad-minimax-h3-t2av-24gb-4step",
3
+ "revision": 0,
4
+ "last_node_id": 18,
5
+ "last_link_id": 20,
6
+ "nodes": [
7
+ {
8
+ "id": 1,
9
+ "type": "UNETLoader",
10
+ "pos": [
11
+ 0,
12
+ 0
13
+ ],
14
+ "size": [
15
+ 430,
16
+ 90
17
+ ],
18
+ "flags": {},
19
+ "order": 0,
20
+ "mode": 0,
21
+ "inputs": [
22
+ {
23
+ "name": "unet_name",
24
+ "type": "COMBO",
25
+ "widget": {
26
+ "name": "unet_name"
27
+ },
28
+ "link": null
29
+ }
30
+ ],
31
+ "outputs": [
32
+ {
33
+ "name": "MODEL",
34
+ "type": "MODEL",
35
+ "links": [
36
+ 1
37
+ ]
38
+ }
39
+ ],
40
+ "properties": {
41
+ "cnr_id": "comfy-core",
42
+ "ver": "0.33.0",
43
+ "Node name for S&R": "UNETLoader",
44
+ "models": [
45
+ {
46
+ "name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
47
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
48
+ "directory": "diffusion_models"
49
+ }
50
+ ]
51
+ },
52
+ "widgets_values": [
53
+ "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
54
+ "default"
55
+ ]
56
+ },
57
+ {
58
+ "id": 2,
59
+ "type": "LoraLoaderModelOnly",
60
+ "pos": [
61
+ 0,
62
+ 140
63
+ ],
64
+ "size": [
65
+ 430,
66
+ 90
67
+ ],
68
+ "flags": {},
69
+ "order": 1,
70
+ "mode": 0,
71
+ "inputs": [
72
+ {
73
+ "name": "model",
74
+ "type": "MODEL",
75
+ "link": 1
76
+ },
77
+ {
78
+ "name": "lora_name",
79
+ "type": "COMBO",
80
+ "widget": {
81
+ "name": "lora_name"
82
+ },
83
+ "link": null
84
+ },
85
+ {
86
+ "name": "strength_model",
87
+ "type": "FLOAT",
88
+ "widget": {
89
+ "name": "strength_model"
90
+ },
91
+ "link": null
92
+ }
93
+ ],
94
+ "outputs": [
95
+ {
96
+ "name": "MODEL",
97
+ "type": "MODEL",
98
+ "links": [
99
+ 2
100
+ ]
101
+ }
102
+ ],
103
+ "properties": {
104
+ "cnr_id": "comfy-core",
105
+ "ver": "0.33.0",
106
+ "Node name for S&R": "LoraLoaderModelOnly",
107
+ "models": [
108
+ {
109
+ "name": "dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
110
+ "url": "https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
111
+ "directory": "loras"
112
+ }
113
+ ]
114
+ },
115
+ "widgets_values": [
116
+ "dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
117
+ 1
118
+ ]
119
+ },
120
+ {
121
+ "id": 3,
122
+ "type": "MiniMaxH3SigmaShift",
123
+ "pos": [
124
+ 0,
125
+ 280
126
+ ],
127
+ "size": [
128
+ 430,
129
+ 90
130
+ ],
131
+ "flags": {},
132
+ "order": 2,
133
+ "mode": 0,
134
+ "inputs": [
135
+ {
136
+ "name": "model",
137
+ "type": "MODEL",
138
+ "link": 2
139
+ },
140
+ {
141
+ "name": "shift_video",
142
+ "type": "FLOAT",
143
+ "widget": {
144
+ "name": "shift_video"
145
+ },
146
+ "link": null
147
+ },
148
+ {
149
+ "name": "shift_audio",
150
+ "type": "FLOAT",
151
+ "widget": {
152
+ "name": "shift_audio"
153
+ },
154
+ "link": null
155
+ }
156
+ ],
157
+ "outputs": [
158
+ {
159
+ "name": "MODEL",
160
+ "type": "MODEL",
161
+ "links": [
162
+ 3
163
+ ]
164
+ }
165
+ ],
166
+ "properties": {
167
+ "cnr_id": "comfy-core",
168
+ "ver": "0.35.1",
169
+ "Node name for S&R": "MiniMaxH3SigmaShift"
170
+ },
171
+ "widgets_values": [
172
+ 12,
173
+ 2
174
+ ]
175
+ },
176
+ {
177
+ "id": 4,
178
+ "type": "CLIPLoader",
179
+ "pos": [
180
+ 0,
181
+ 420
182
+ ],
183
+ "size": [
184
+ 430,
185
+ 110
186
+ ],
187
+ "flags": {},
188
+ "order": 3,
189
+ "mode": 0,
190
+ "inputs": [
191
+ {
192
+ "name": "clip_name",
193
+ "type": "COMBO",
194
+ "widget": {
195
+ "name": "clip_name"
196
+ },
197
+ "link": null
198
+ }
199
+ ],
200
+ "outputs": [
201
+ {
202
+ "name": "CLIP",
203
+ "type": "CLIP",
204
+ "links": [
205
+ 6
206
+ ]
207
+ }
208
+ ],
209
+ "properties": {
210
+ "cnr_id": "comfy-core",
211
+ "ver": "0.33.0",
212
+ "Node name for S&R": "CLIPLoader",
213
+ "models": [
214
+ {
215
+ "name": "qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
216
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
217
+ "directory": "text_encoders"
218
+ }
219
+ ]
220
+ },
221
+ "widgets_values": [
222
+ "qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
223
+ "minimax",
224
+ "default"
225
+ ]
226
+ },
227
+ {
228
+ "id": 5,
229
+ "type": "VAELoader",
230
+ "pos": [
231
+ 0,
232
+ 580
233
+ ],
234
+ "size": [
235
+ 430,
236
+ 60
237
+ ],
238
+ "flags": {},
239
+ "order": 4,
240
+ "mode": 0,
241
+ "inputs": [
242
+ {
243
+ "name": "vae_name",
244
+ "type": "COMBO",
245
+ "widget": {
246
+ "name": "vae_name"
247
+ },
248
+ "link": null
249
+ }
250
+ ],
251
+ "outputs": [
252
+ {
253
+ "name": "VAE",
254
+ "type": "VAE",
255
+ "links": [
256
+ 7,
257
+ 15
258
+ ]
259
+ }
260
+ ],
261
+ "properties": {
262
+ "cnr_id": "comfy-core",
263
+ "ver": "0.33.0",
264
+ "Node name for S&R": "VAELoader",
265
+ "models": [
266
+ {
267
+ "name": "minimax_h3_video_vae_fp16.safetensors",
268
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors",
269
+ "directory": "vae"
270
+ }
271
+ ]
272
+ },
273
+ "widgets_values": [
274
+ "minimax_h3_video_vae_fp16.safetensors"
275
+ ]
276
+ },
277
+ {
278
+ "id": 6,
279
+ "type": "VAELoader",
280
+ "pos": [
281
+ 0,
282
+ 690
283
+ ],
284
+ "size": [
285
+ 430,
286
+ 60
287
+ ],
288
+ "flags": {},
289
+ "order": 5,
290
+ "mode": 0,
291
+ "inputs": [
292
+ {
293
+ "name": "vae_name",
294
+ "type": "COMBO",
295
+ "widget": {
296
+ "name": "vae_name"
297
+ },
298
+ "link": null
299
+ }
300
+ ],
301
+ "outputs": [
302
+ {
303
+ "name": "VAE",
304
+ "type": "VAE",
305
+ "links": [
306
+ 17
307
+ ]
308
+ }
309
+ ],
310
+ "properties": {
311
+ "cnr_id": "comfy-core",
312
+ "ver": "0.33.0",
313
+ "Node name for S&R": "VAELoader",
314
+ "models": [
315
+ {
316
+ "name": "minimax_h3_audio_vae_fp32.safetensors",
317
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors",
318
+ "directory": "vae"
319
+ }
320
+ ]
321
+ },
322
+ "widgets_values": [
323
+ "minimax_h3_audio_vae_fp32.safetensors"
324
+ ]
325
+ },
326
+ {
327
+ "id": 7,
328
+ "type": "MiniMaxH3ImageToVideo",
329
+ "pos": [
330
+ 480,
331
+ 0
332
+ ],
333
+ "size": [
334
+ 520,
335
+ 560
336
+ ],
337
+ "flags": {},
338
+ "order": 6,
339
+ "mode": 0,
340
+ "inputs": [
341
+ {
342
+ "name": "clip",
343
+ "type": "CLIP",
344
+ "link": 6
345
+ },
346
+ {
347
+ "name": "vae",
348
+ "type": "VAE",
349
+ "link": 7
350
+ },
351
+ {
352
+ "name": "first_frame",
353
+ "shape": 7,
354
+ "type": "IMAGE",
355
+ "link": null
356
+ },
357
+ {
358
+ "name": "last_frame",
359
+ "shape": 7,
360
+ "type": "IMAGE",
361
+ "link": null
362
+ },
363
+ {
364
+ "name": "prompt",
365
+ "type": "STRING",
366
+ "widget": {
367
+ "name": "prompt"
368
+ },
369
+ "link": null
370
+ },
371
+ {
372
+ "name": "width",
373
+ "type": "INT",
374
+ "widget": {
375
+ "name": "width"
376
+ },
377
+ "link": null
378
+ },
379
+ {
380
+ "name": "height",
381
+ "type": "INT",
382
+ "widget": {
383
+ "name": "height"
384
+ },
385
+ "link": null
386
+ },
387
+ {
388
+ "name": "length",
389
+ "type": "INT",
390
+ "widget": {
391
+ "name": "length"
392
+ },
393
+ "link": null
394
+ }
395
+ ],
396
+ "outputs": [
397
+ {
398
+ "name": "positive",
399
+ "type": "CONDITIONING",
400
+ "links": [
401
+ 8
402
+ ]
403
+ },
404
+ {
405
+ "name": "LATENT",
406
+ "type": "LATENT",
407
+ "links": [
408
+ 13
409
+ ]
410
+ }
411
+ ],
412
+ "properties": {
413
+ "cnr_id": "comfy-core",
414
+ "ver": "0.33.0",
415
+ "Node name for S&R": "MiniMaxH3ImageToVideo"
416
+ },
417
+ "widgets_values": [
418
+ "integrated_multimodal_description: [Shot 1] Live-action, cozy late-night podcast studio with warm practical lights, a wide shot frames two hosts at a wooden table with microphones, headphones and two white coffee mugs printed with \"DMAD\" in bold black letters, big glowing magenta backlit LED channel letters spelling \"DMAD\" on the brick wall behind them, each solid letter lit evenly from within with a soft halo on the bricks. [Shot 2] At 00:02.500, the camera cuts to a medium close-up of a young man with curly hair leaning toward his microphone, and the man (S1) says: <d>[English] So you're telling me it makes a whole video in four steps?</d> [Shot 3] At 00:06.000, the camera cuts to a medium close-up of a woman with short black hair and round glasses who smiles, raises one hand beside her face and holds up exactly four fingers toward the camera, thumb folded in, keeping the hand steady, and the woman (S2) says: <d>[English] Four steps. Video and sound, together.</d> [Shot 4] At 00:09.500, the camera cuts back to the man, who throws his hands up in disbelief and laughs. [Shot 5] At 00:12.000, the camera cuts to a wide shot as both hosts laugh and clink their white \"DMAD\" coffee mugs under the glowing DMAD letters, the letters on the mugs staying crisp as they move.\n\noverall_soundscape: A quiet studio room tone, close-miked voices, chairs creaking, laughter, and the clink of ceramic mugs.\n\nnon_diegetic_music: A mellow lo-fi beat playing softly underneath, swelling slightly at the end.",
419
+ 1344,
420
+ 768,
421
+ 362
422
+ ]
423
+ },
424
+ {
425
+ "id": 8,
426
+ "type": "RandomNoise",
427
+ "pos": [
428
+ 1050,
429
+ 0
430
+ ],
431
+ "size": [
432
+ 320,
433
+ 90
434
+ ],
435
+ "flags": {},
436
+ "order": 7,
437
+ "mode": 0,
438
+ "inputs": [
439
+ {
440
+ "name": "noise_seed",
441
+ "type": "INT",
442
+ "widget": {
443
+ "name": "noise_seed"
444
+ },
445
+ "link": null
446
+ }
447
+ ],
448
+ "outputs": [
449
+ {
450
+ "name": "NOISE",
451
+ "type": "NOISE",
452
+ "links": [
453
+ 9
454
+ ]
455
+ }
456
+ ],
457
+ "properties": {
458
+ "cnr_id": "comfy-core",
459
+ "ver": "0.33.0",
460
+ "Node name for S&R": "RandomNoise"
461
+ },
462
+ "widgets_values": [
463
+ 2,
464
+ "fixed"
465
+ ]
466
+ },
467
+ {
468
+ "id": 9,
469
+ "type": "BasicGuider",
470
+ "pos": [
471
+ 1050,
472
+ 140
473
+ ],
474
+ "size": [
475
+ 320,
476
+ 60
477
+ ],
478
+ "flags": {},
479
+ "order": 8,
480
+ "mode": 0,
481
+ "inputs": [
482
+ {
483
+ "name": "model",
484
+ "type": "MODEL",
485
+ "link": 4
486
+ },
487
+ {
488
+ "name": "conditioning",
489
+ "type": "CONDITIONING",
490
+ "link": 8
491
+ }
492
+ ],
493
+ "outputs": [
494
+ {
495
+ "name": "GUIDER",
496
+ "type": "GUIDER",
497
+ "links": [
498
+ 10
499
+ ]
500
+ }
501
+ ],
502
+ "properties": {
503
+ "cnr_id": "comfy-core",
504
+ "ver": "0.33.0",
505
+ "Node name for S&R": "BasicGuider"
506
+ }
507
+ },
508
+ {
509
+ "id": 10,
510
+ "type": "KSamplerSelect",
511
+ "pos": [
512
+ 1050,
513
+ 250
514
+ ],
515
+ "size": [
516
+ 320,
517
+ 60
518
+ ],
519
+ "flags": {},
520
+ "order": 9,
521
+ "mode": 0,
522
+ "inputs": [],
523
+ "outputs": [
524
+ {
525
+ "name": "SAMPLER",
526
+ "type": "SAMPLER",
527
+ "links": [
528
+ 11
529
+ ]
530
+ }
531
+ ],
532
+ "properties": {
533
+ "cnr_id": "comfy-core",
534
+ "ver": "0.33.0",
535
+ "Node name for S&R": "KSamplerSelect"
536
+ },
537
+ "widgets_values": [
538
+ "lcm"
539
+ ]
540
+ },
541
+ {
542
+ "id": 11,
543
+ "type": "BasicScheduler",
544
+ "pos": [
545
+ 1050,
546
+ 360
547
+ ],
548
+ "size": [
549
+ 320,
550
+ 110
551
+ ],
552
+ "flags": {},
553
+ "order": 10,
554
+ "mode": 0,
555
+ "inputs": [
556
+ {
557
+ "name": "model",
558
+ "type": "MODEL",
559
+ "link": 5
560
+ },
561
+ {
562
+ "name": "steps",
563
+ "type": "INT",
564
+ "widget": {
565
+ "name": "steps"
566
+ },
567
+ "link": null
568
+ }
569
+ ],
570
+ "outputs": [
571
+ {
572
+ "name": "SIGMAS",
573
+ "type": "SIGMAS",
574
+ "links": [
575
+ 12
576
+ ]
577
+ }
578
+ ],
579
+ "properties": {
580
+ "cnr_id": "comfy-core",
581
+ "ver": "0.33.0",
582
+ "Node name for S&R": "BasicScheduler"
583
+ },
584
+ "widgets_values": [
585
+ "simple",
586
+ 4,
587
+ 1
588
+ ]
589
+ },
590
+ {
591
+ "id": 12,
592
+ "type": "SamplerCustomAdvanced",
593
+ "pos": [
594
+ 1050,
595
+ 520
596
+ ],
597
+ "size": [
598
+ 320,
599
+ 110
600
+ ],
601
+ "flags": {},
602
+ "order": 11,
603
+ "mode": 0,
604
+ "inputs": [
605
+ {
606
+ "name": "noise",
607
+ "type": "NOISE",
608
+ "link": 9
609
+ },
610
+ {
611
+ "name": "guider",
612
+ "type": "GUIDER",
613
+ "link": 10
614
+ },
615
+ {
616
+ "name": "sampler",
617
+ "type": "SAMPLER",
618
+ "link": 11
619
+ },
620
+ {
621
+ "name": "sigmas",
622
+ "type": "SIGMAS",
623
+ "link": 12
624
+ },
625
+ {
626
+ "name": "latent_image",
627
+ "type": "LATENT",
628
+ "link": 13
629
+ }
630
+ ],
631
+ "outputs": [
632
+ {
633
+ "name": "output",
634
+ "type": "LATENT",
635
+ "links": [
636
+ 14,
637
+ 16
638
+ ]
639
+ },
640
+ {
641
+ "name": "denoised_output",
642
+ "type": "LATENT",
643
+ "links": []
644
+ }
645
+ ],
646
+ "properties": {
647
+ "cnr_id": "comfy-core",
648
+ "ver": "0.33.0",
649
+ "Node name for S&R": "SamplerCustomAdvanced"
650
+ }
651
+ },
652
+ {
653
+ "id": 13,
654
+ "type": "VAEDecode",
655
+ "pos": [
656
+ 1420,
657
+ 0
658
+ ],
659
+ "size": [
660
+ 260,
661
+ 50
662
+ ],
663
+ "flags": {
664
+ "collapsed": false
665
+ },
666
+ "order": 12,
667
+ "mode": 0,
668
+ "inputs": [
669
+ {
670
+ "name": "samples",
671
+ "type": "LATENT",
672
+ "link": 14
673
+ },
674
+ {
675
+ "name": "vae",
676
+ "type": "VAE",
677
+ "link": 15
678
+ }
679
+ ],
680
+ "outputs": [
681
+ {
682
+ "name": "IMAGE",
683
+ "type": "IMAGE",
684
+ "links": [
685
+ 18
686
+ ]
687
+ }
688
+ ],
689
+ "properties": {
690
+ "cnr_id": "comfy-core",
691
+ "ver": "0.33.0",
692
+ "Node name for S&R": "VAEDecode"
693
+ }
694
+ },
695
+ {
696
+ "id": 14,
697
+ "type": "VAEDecodeAudio",
698
+ "pos": [
699
+ 1420,
700
+ 100
701
+ ],
702
+ "size": [
703
+ 260,
704
+ 50
705
+ ],
706
+ "flags": {
707
+ "collapsed": false
708
+ },
709
+ "order": 13,
710
+ "mode": 0,
711
+ "inputs": [
712
+ {
713
+ "name": "samples",
714
+ "type": "LATENT",
715
+ "link": 16
716
+ },
717
+ {
718
+ "name": "vae",
719
+ "type": "VAE",
720
+ "link": 17
721
+ }
722
+ ],
723
+ "outputs": [
724
+ {
725
+ "name": "AUDIO",
726
+ "type": "AUDIO",
727
+ "links": [
728
+ 19
729
+ ]
730
+ }
731
+ ],
732
+ "properties": {
733
+ "cnr_id": "comfy-core",
734
+ "ver": "0.33.0",
735
+ "Node name for S&R": "VAEDecodeAudio"
736
+ }
737
+ },
738
+ {
739
+ "id": 15,
740
+ "type": "CreateVideo",
741
+ "pos": [
742
+ 1420,
743
+ 200
744
+ ],
745
+ "size": [
746
+ 260,
747
+ 130
748
+ ],
749
+ "flags": {},
750
+ "order": 14,
751
+ "mode": 0,
752
+ "inputs": [
753
+ {
754
+ "name": "images",
755
+ "type": "IMAGE",
756
+ "link": 18
757
+ },
758
+ {
759
+ "name": "audio",
760
+ "shape": 7,
761
+ "type": "AUDIO",
762
+ "link": 19
763
+ }
764
+ ],
765
+ "outputs": [
766
+ {
767
+ "name": "VIDEO",
768
+ "type": "VIDEO",
769
+ "links": [
770
+ 20
771
+ ]
772
+ }
773
+ ],
774
+ "properties": {
775
+ "cnr_id": "comfy-core",
776
+ "ver": "0.33.0",
777
+ "Node name for S&R": "CreateVideo"
778
+ },
779
+ "widgets_values": [
780
+ 24,
781
+ 8,
782
+ "sRGB",
783
+ "none"
784
+ ]
785
+ },
786
+ {
787
+ "id": 16,
788
+ "type": "SaveVideo",
789
+ "pos": [
790
+ 1720,
791
+ 0
792
+ ],
793
+ "size": [
794
+ 520,
795
+ 560
796
+ ],
797
+ "flags": {},
798
+ "order": 15,
799
+ "mode": 0,
800
+ "inputs": [
801
+ {
802
+ "name": "video",
803
+ "type": "VIDEO",
804
+ "link": 20
805
+ },
806
+ {
807
+ "name": "filename_prefix",
808
+ "type": "STRING",
809
+ "widget": {
810
+ "name": "filename_prefix"
811
+ },
812
+ "link": null
813
+ },
814
+ {
815
+ "name": "format",
816
+ "type": "COMFY_DYNAMICCOMBO_V3",
817
+ "widget": {
818
+ "name": "format"
819
+ },
820
+ "link": null
821
+ },
822
+ {
823
+ "name": "format.codec",
824
+ "type": "COMFY_DYNAMICCOMBO_V3",
825
+ "widget": {
826
+ "name": "format.codec"
827
+ },
828
+ "link": null
829
+ },
830
+ {
831
+ "name": "codec",
832
+ "shape": 7,
833
+ "type": "COMFY_DYNAMICCOMBO_V3",
834
+ "widget": {
835
+ "name": "codec"
836
+ },
837
+ "link": null
838
+ }
839
+ ],
840
+ "outputs": [
841
+ {
842
+ "name": "video",
843
+ "type": "VIDEO",
844
+ "links": []
845
+ }
846
+ ],
847
+ "properties": {
848
+ "cnr_id": "comfy-core",
849
+ "ver": "0.33.0",
850
+ "Node name for S&R": "SaveVideo"
851
+ },
852
+ "widgets_values": [
853
+ "video/dmad_h3_4step_podcast",
854
+ "auto",
855
+ "auto",
856
+ "auto"
857
+ ]
858
+ },
859
+ {
860
+ "id": 18,
861
+ "type": "H3MemoryOptimization",
862
+ "pos": [
863
+ 0,
864
+ 800
865
+ ],
866
+ "size": [
867
+ 430,
868
+ 200
869
+ ],
870
+ "flags": {},
871
+ "order": 16,
872
+ "mode": 0,
873
+ "inputs": [
874
+ {
875
+ "name": "model",
876
+ "type": "MODEL",
877
+ "link": 3
878
+ },
879
+ {
880
+ "name": "fused_qkv",
881
+ "type": "COMBO",
882
+ "widget": {
883
+ "name": "fused_qkv"
884
+ },
885
+ "link": null
886
+ },
887
+ {
888
+ "name": "mlp_memory",
889
+ "type": "COMBO",
890
+ "widget": {
891
+ "name": "mlp_memory"
892
+ },
893
+ "link": null
894
+ },
895
+ {
896
+ "name": "chunk_rows",
897
+ "type": "INT",
898
+ "widget": {
899
+ "name": "chunk_rows"
900
+ },
901
+ "link": null
902
+ },
903
+ {
904
+ "name": "preserve_precision",
905
+ "type": "BOOLEAN",
906
+ "widget": {
907
+ "name": "preserve_precision"
908
+ },
909
+ "link": null
910
+ },
911
+ {
912
+ "name": "precision_mode",
913
+ "type": "COMBO",
914
+ "widget": {
915
+ "name": "precision_mode"
916
+ },
917
+ "link": null
918
+ },
919
+ {
920
+ "name": "qkv_streaming_mode",
921
+ "type": "COMBO",
922
+ "widget": {
923
+ "name": "qkv_streaming_mode"
924
+ },
925
+ "link": null
926
+ },
927
+ {
928
+ "name": "embedding_memory_mode",
929
+ "type": "COMBO",
930
+ "widget": {
931
+ "name": "embedding_memory_mode"
932
+ },
933
+ "link": null
934
+ },
935
+ {
936
+ "name": "kitchen_v_memory_mode",
937
+ "type": "COMBO",
938
+ "widget": {
939
+ "name": "kitchen_v_memory_mode"
940
+ },
941
+ "link": null
942
+ }
943
+ ],
944
+ "outputs": [
945
+ {
946
+ "name": "MODEL",
947
+ "type": "MODEL",
948
+ "links": [
949
+ 4,
950
+ 5
951
+ ]
952
+ }
953
+ ],
954
+ "properties": {
955
+ "cnr_id": "h3-optimizations",
956
+ "ver": "0.2.45",
957
+ "Node name for S&R": "H3MemoryOptimization"
958
+ },
959
+ "widgets_values": [
960
+ "auto",
961
+ "auto",
962
+ 4096,
963
+ true,
964
+ "Auto",
965
+ "Auto",
966
+ "Auto",
967
+ "Standard"
968
+ ]
969
+ },
970
+ {
971
+ "id": 17,
972
+ "type": "MarkdownNote",
973
+ "pos": [
974
+ 480,
975
+ 720
976
+ ],
977
+ "size": [
978
+ 890,
979
+ 330
980
+ ],
981
+ "flags": {},
982
+ "order": 17,
983
+ "mode": 0,
984
+ "inputs": [],
985
+ "outputs": [],
986
+ "properties": {},
987
+ "widgets_values": [
988
+ "## DMAD × MiniMax-H3 — 4-step text-to-audio-video (24 GB GPU)\n\n**Models** (put in `ComfyUI/models/...`):\n- `diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors` — [Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3)\n- `text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors` — same repo\n- `vae/minimax_h3_video_vae_fp16.safetensors`, `vae/minimax_h3_audio_vae_fp32.safetensors` — same repo\n- `loras/dmad_minimax_h3_4step_full_critic_comfyui.safetensors` — [ZhengmingYu/DMAD](https://huggingface.co/ZhengmingYu/DMAD/tree/main/minimax_h3)\n\n**Settings that matter:** LoRA strength 1.0 · shift 12 / audio 2 · sampler `lcm` + scheduler `simple` · no CFG.\n**Steps:** 4 = fastest; 8 or 12 give finer detail on fast motion. Avoid `euler` (softer).\n**Prompt:** describe the shot, then `overall_soundscape:` and `non_diegetic_music:` for the audio.\n**Length:** 362 frames = 15 s at 24 fps (lengths are 5 + 17k: 124 = 5 s, 243 = 10 s, 362 = 15 s). On a 24 GB GPU this\nneeds the `H3 Memory Optimization` node of [H3-Optimizations](https://github.com/Zironic/H3-Optimizations) (install it\nwith ComfyUI-Manager); without it, 5 s clips still fit with stock nodes — bypass the node and set length 124.\n"
989
+ ]
990
+ }
991
+ ],
992
+ "links": [
993
+ [
994
+ 1,
995
+ 1,
996
+ 0,
997
+ 2,
998
+ 0,
999
+ "MODEL"
1000
+ ],
1001
+ [
1002
+ 2,
1003
+ 2,
1004
+ 0,
1005
+ 3,
1006
+ 0,
1007
+ "MODEL"
1008
+ ],
1009
+ [
1010
+ 3,
1011
+ 3,
1012
+ 0,
1013
+ 18,
1014
+ 0,
1015
+ "MODEL"
1016
+ ],
1017
+ [
1018
+ 4,
1019
+ 18,
1020
+ 0,
1021
+ 9,
1022
+ 0,
1023
+ "MODEL"
1024
+ ],
1025
+ [
1026
+ 5,
1027
+ 18,
1028
+ 0,
1029
+ 11,
1030
+ 0,
1031
+ "MODEL"
1032
+ ],
1033
+ [
1034
+ 6,
1035
+ 4,
1036
+ 0,
1037
+ 7,
1038
+ 0,
1039
+ "CLIP"
1040
+ ],
1041
+ [
1042
+ 7,
1043
+ 5,
1044
+ 0,
1045
+ 7,
1046
+ 1,
1047
+ "VAE"
1048
+ ],
1049
+ [
1050
+ 8,
1051
+ 7,
1052
+ 0,
1053
+ 9,
1054
+ 1,
1055
+ "CONDITIONING"
1056
+ ],
1057
+ [
1058
+ 9,
1059
+ 8,
1060
+ 0,
1061
+ 12,
1062
+ 0,
1063
+ "NOISE"
1064
+ ],
1065
+ [
1066
+ 10,
1067
+ 9,
1068
+ 0,
1069
+ 12,
1070
+ 1,
1071
+ "GUIDER"
1072
+ ],
1073
+ [
1074
+ 11,
1075
+ 10,
1076
+ 0,
1077
+ 12,
1078
+ 2,
1079
+ "SAMPLER"
1080
+ ],
1081
+ [
1082
+ 12,
1083
+ 11,
1084
+ 0,
1085
+ 12,
1086
+ 3,
1087
+ "SIGMAS"
1088
+ ],
1089
+ [
1090
+ 13,
1091
+ 7,
1092
+ 1,
1093
+ 12,
1094
+ 4,
1095
+ "LATENT"
1096
+ ],
1097
+ [
1098
+ 14,
1099
+ 12,
1100
+ 0,
1101
+ 13,
1102
+ 0,
1103
+ "LATENT"
1104
+ ],
1105
+ [
1106
+ 15,
1107
+ 5,
1108
+ 0,
1109
+ 13,
1110
+ 1,
1111
+ "VAE"
1112
+ ],
1113
+ [
1114
+ 16,
1115
+ 12,
1116
+ 0,
1117
+ 14,
1118
+ 0,
1119
+ "LATENT"
1120
+ ],
1121
+ [
1122
+ 17,
1123
+ 6,
1124
+ 0,
1125
+ 14,
1126
+ 1,
1127
+ "VAE"
1128
+ ],
1129
+ [
1130
+ 18,
1131
+ 13,
1132
+ 0,
1133
+ 15,
1134
+ 0,
1135
+ "IMAGE"
1136
+ ],
1137
+ [
1138
+ 19,
1139
+ 14,
1140
+ 0,
1141
+ 15,
1142
+ 1,
1143
+ "AUDIO"
1144
+ ],
1145
+ [
1146
+ 20,
1147
+ 15,
1148
+ 0,
1149
+ 16,
1150
+ 0,
1151
+ "VIDEO"
1152
+ ]
1153
+ ],
1154
+ "groups": [],
1155
+ "config": {},
1156
+ "extra": {
1157
+ "ds": {
1158
+ "scale": 0.6,
1159
+ "offset": [
1160
+ 40,
1161
+ 60
1162
+ ]
1163
+ },
1164
+ "frontendVersion": "1.49.6"
1165
+ },
1166
+ "version": 0.4
1167
+ }
minimax_h3/workflows/dmad_h3_4step_15s_podcast.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5d0f847b69e07075d9835bb6c3a05f6f44c932e3b0c9f7359c6297a450c7c8e2
3
+ size 5067052
minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.json ADDED
@@ -0,0 +1,1214 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "id": "dmad-minimax-h3-t2av-8gb-4step",
3
+ "revision": 0,
4
+ "last_node_id": 19,
5
+ "last_link_id": 21,
6
+ "nodes": [
7
+ {
8
+ "id": 19,
9
+ "type": "H3AIMDOResidencyLimiter",
10
+ "pos": [
11
+ 0,
12
+ 1050
13
+ ],
14
+ "size": [
15
+ 430,
16
+ 60
17
+ ],
18
+ "flags": {},
19
+ "order": 17,
20
+ "mode": 0,
21
+ "inputs": [
22
+ {
23
+ "name": "model",
24
+ "type": "MODEL",
25
+ "link": 4
26
+ }
27
+ ],
28
+ "outputs": [
29
+ {
30
+ "name": "MODEL",
31
+ "type": "MODEL",
32
+ "links": [
33
+ 5,
34
+ 6
35
+ ]
36
+ }
37
+ ],
38
+ "properties": {
39
+ "cnr_id": "h3-optimizations",
40
+ "ver": "0.2.45",
41
+ "Node name for S&R": "H3AIMDOResidencyLimiter"
42
+ },
43
+ "widgets_values": [
44
+ "0 blocks"
45
+ ]
46
+ },
47
+ {
48
+ "id": 1,
49
+ "type": "UNETLoader",
50
+ "pos": [
51
+ 0,
52
+ 0
53
+ ],
54
+ "size": [
55
+ 430,
56
+ 90
57
+ ],
58
+ "flags": {},
59
+ "order": 0,
60
+ "mode": 0,
61
+ "inputs": [
62
+ {
63
+ "name": "unet_name",
64
+ "type": "COMBO",
65
+ "widget": {
66
+ "name": "unet_name"
67
+ },
68
+ "link": null
69
+ }
70
+ ],
71
+ "outputs": [
72
+ {
73
+ "name": "MODEL",
74
+ "type": "MODEL",
75
+ "links": [
76
+ 1
77
+ ]
78
+ }
79
+ ],
80
+ "properties": {
81
+ "cnr_id": "comfy-core",
82
+ "ver": "0.33.0",
83
+ "Node name for S&R": "UNETLoader",
84
+ "models": [
85
+ {
86
+ "name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
87
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
88
+ "directory": "diffusion_models"
89
+ }
90
+ ]
91
+ },
92
+ "widgets_values": [
93
+ "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
94
+ "default"
95
+ ]
96
+ },
97
+ {
98
+ "id": 2,
99
+ "type": "LoraLoaderModelOnly",
100
+ "pos": [
101
+ 0,
102
+ 140
103
+ ],
104
+ "size": [
105
+ 430,
106
+ 90
107
+ ],
108
+ "flags": {},
109
+ "order": 1,
110
+ "mode": 0,
111
+ "inputs": [
112
+ {
113
+ "name": "model",
114
+ "type": "MODEL",
115
+ "link": 1
116
+ },
117
+ {
118
+ "name": "lora_name",
119
+ "type": "COMBO",
120
+ "widget": {
121
+ "name": "lora_name"
122
+ },
123
+ "link": null
124
+ },
125
+ {
126
+ "name": "strength_model",
127
+ "type": "FLOAT",
128
+ "widget": {
129
+ "name": "strength_model"
130
+ },
131
+ "link": null
132
+ }
133
+ ],
134
+ "outputs": [
135
+ {
136
+ "name": "MODEL",
137
+ "type": "MODEL",
138
+ "links": [
139
+ 2
140
+ ]
141
+ }
142
+ ],
143
+ "properties": {
144
+ "cnr_id": "comfy-core",
145
+ "ver": "0.33.0",
146
+ "Node name for S&R": "LoraLoaderModelOnly",
147
+ "models": [
148
+ {
149
+ "name": "dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
150
+ "url": "https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
151
+ "directory": "loras"
152
+ }
153
+ ]
154
+ },
155
+ "widgets_values": [
156
+ "dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
157
+ 1
158
+ ]
159
+ },
160
+ {
161
+ "id": 3,
162
+ "type": "MiniMaxH3SigmaShift",
163
+ "pos": [
164
+ 0,
165
+ 280
166
+ ],
167
+ "size": [
168
+ 430,
169
+ 90
170
+ ],
171
+ "flags": {},
172
+ "order": 2,
173
+ "mode": 0,
174
+ "inputs": [
175
+ {
176
+ "name": "model",
177
+ "type": "MODEL",
178
+ "link": 2
179
+ },
180
+ {
181
+ "name": "shift_video",
182
+ "type": "FLOAT",
183
+ "widget": {
184
+ "name": "shift_video"
185
+ },
186
+ "link": null
187
+ },
188
+ {
189
+ "name": "shift_audio",
190
+ "type": "FLOAT",
191
+ "widget": {
192
+ "name": "shift_audio"
193
+ },
194
+ "link": null
195
+ }
196
+ ],
197
+ "outputs": [
198
+ {
199
+ "name": "MODEL",
200
+ "type": "MODEL",
201
+ "links": [
202
+ 3
203
+ ]
204
+ }
205
+ ],
206
+ "properties": {
207
+ "cnr_id": "comfy-core",
208
+ "ver": "0.35.1",
209
+ "Node name for S&R": "MiniMaxH3SigmaShift"
210
+ },
211
+ "widgets_values": [
212
+ 12,
213
+ 2
214
+ ]
215
+ },
216
+ {
217
+ "id": 4,
218
+ "type": "CLIPLoader",
219
+ "pos": [
220
+ 0,
221
+ 420
222
+ ],
223
+ "size": [
224
+ 430,
225
+ 110
226
+ ],
227
+ "flags": {},
228
+ "order": 3,
229
+ "mode": 0,
230
+ "inputs": [
231
+ {
232
+ "name": "clip_name",
233
+ "type": "COMBO",
234
+ "widget": {
235
+ "name": "clip_name"
236
+ },
237
+ "link": null
238
+ }
239
+ ],
240
+ "outputs": [
241
+ {
242
+ "name": "CLIP",
243
+ "type": "CLIP",
244
+ "links": [
245
+ 7
246
+ ]
247
+ }
248
+ ],
249
+ "properties": {
250
+ "cnr_id": "comfy-core",
251
+ "ver": "0.33.0",
252
+ "Node name for S&R": "CLIPLoader",
253
+ "models": [
254
+ {
255
+ "name": "qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
256
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
257
+ "directory": "text_encoders"
258
+ }
259
+ ]
260
+ },
261
+ "widgets_values": [
262
+ "qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
263
+ "minimax",
264
+ "default"
265
+ ]
266
+ },
267
+ {
268
+ "id": 5,
269
+ "type": "VAELoader",
270
+ "pos": [
271
+ 0,
272
+ 580
273
+ ],
274
+ "size": [
275
+ 430,
276
+ 60
277
+ ],
278
+ "flags": {},
279
+ "order": 4,
280
+ "mode": 0,
281
+ "inputs": [
282
+ {
283
+ "name": "vae_name",
284
+ "type": "COMBO",
285
+ "widget": {
286
+ "name": "vae_name"
287
+ },
288
+ "link": null
289
+ }
290
+ ],
291
+ "outputs": [
292
+ {
293
+ "name": "VAE",
294
+ "type": "VAE",
295
+ "links": [
296
+ 8,
297
+ 16
298
+ ]
299
+ }
300
+ ],
301
+ "properties": {
302
+ "cnr_id": "comfy-core",
303
+ "ver": "0.33.0",
304
+ "Node name for S&R": "VAELoader",
305
+ "models": [
306
+ {
307
+ "name": "minimax_h3_video_vae_fp16.safetensors",
308
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors",
309
+ "directory": "vae"
310
+ }
311
+ ]
312
+ },
313
+ "widgets_values": [
314
+ "minimax_h3_video_vae_fp16.safetensors"
315
+ ]
316
+ },
317
+ {
318
+ "id": 6,
319
+ "type": "VAELoader",
320
+ "pos": [
321
+ 0,
322
+ 690
323
+ ],
324
+ "size": [
325
+ 430,
326
+ 60
327
+ ],
328
+ "flags": {},
329
+ "order": 5,
330
+ "mode": 0,
331
+ "inputs": [
332
+ {
333
+ "name": "vae_name",
334
+ "type": "COMBO",
335
+ "widget": {
336
+ "name": "vae_name"
337
+ },
338
+ "link": null
339
+ }
340
+ ],
341
+ "outputs": [
342
+ {
343
+ "name": "VAE",
344
+ "type": "VAE",
345
+ "links": [
346
+ 18
347
+ ]
348
+ }
349
+ ],
350
+ "properties": {
351
+ "cnr_id": "comfy-core",
352
+ "ver": "0.33.0",
353
+ "Node name for S&R": "VAELoader",
354
+ "models": [
355
+ {
356
+ "name": "minimax_h3_audio_vae_fp32.safetensors",
357
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors",
358
+ "directory": "vae"
359
+ }
360
+ ]
361
+ },
362
+ "widgets_values": [
363
+ "minimax_h3_audio_vae_fp32.safetensors"
364
+ ]
365
+ },
366
+ {
367
+ "id": 7,
368
+ "type": "MiniMaxH3ImageToVideo",
369
+ "pos": [
370
+ 480,
371
+ 0
372
+ ],
373
+ "size": [
374
+ 520,
375
+ 560
376
+ ],
377
+ "flags": {},
378
+ "order": 6,
379
+ "mode": 0,
380
+ "inputs": [
381
+ {
382
+ "name": "clip",
383
+ "type": "CLIP",
384
+ "link": 7
385
+ },
386
+ {
387
+ "name": "vae",
388
+ "type": "VAE",
389
+ "link": 8
390
+ },
391
+ {
392
+ "name": "first_frame",
393
+ "shape": 7,
394
+ "type": "IMAGE",
395
+ "link": null
396
+ },
397
+ {
398
+ "name": "last_frame",
399
+ "shape": 7,
400
+ "type": "IMAGE",
401
+ "link": null
402
+ },
403
+ {
404
+ "name": "prompt",
405
+ "type": "STRING",
406
+ "widget": {
407
+ "name": "prompt"
408
+ },
409
+ "link": null
410
+ },
411
+ {
412
+ "name": "width",
413
+ "type": "INT",
414
+ "widget": {
415
+ "name": "width"
416
+ },
417
+ "link": null
418
+ },
419
+ {
420
+ "name": "height",
421
+ "type": "INT",
422
+ "widget": {
423
+ "name": "height"
424
+ },
425
+ "link": null
426
+ },
427
+ {
428
+ "name": "length",
429
+ "type": "INT",
430
+ "widget": {
431
+ "name": "length"
432
+ },
433
+ "link": null
434
+ }
435
+ ],
436
+ "outputs": [
437
+ {
438
+ "name": "positive",
439
+ "type": "CONDITIONING",
440
+ "links": [
441
+ 9
442
+ ]
443
+ },
444
+ {
445
+ "name": "LATENT",
446
+ "type": "LATENT",
447
+ "links": [
448
+ 14
449
+ ]
450
+ }
451
+ ],
452
+ "properties": {
453
+ "cnr_id": "comfy-core",
454
+ "ver": "0.33.0",
455
+ "Node name for S&R": "MiniMaxH3ImageToVideo"
456
+ },
457
+ "widgets_values": [
458
+ "integrated_multimodal_description: [Shot 1] Live-action, golden-hour city rooftop with the skyline glowing behind, a medium shot frames a young man with a denim jacket sitting on a ledge and strumming a natural-wood acoustic guitar with \"DMAD\" painted in bold white letters on its body below the sound hole, the warm sun flaring behind him. [Shot 2] At 00:02.500, the camera slowly arcs to a close-up as he sings a soft line with his eyes closed, and the young man (S1) sings: <d>[English] Golden light, stay with me tonight.</d>\n\noverall_soundscape: A gentle breeze, distant city traffic far below, the bright strum of the acoustic guitar, and his soft live vocal.\n\nnon_diegetic_music: N/A",
459
+ 1344,
460
+ 768,
461
+ 124
462
+ ]
463
+ },
464
+ {
465
+ "id": 8,
466
+ "type": "RandomNoise",
467
+ "pos": [
468
+ 1050,
469
+ 0
470
+ ],
471
+ "size": [
472
+ 320,
473
+ 90
474
+ ],
475
+ "flags": {},
476
+ "order": 7,
477
+ "mode": 0,
478
+ "inputs": [
479
+ {
480
+ "name": "noise_seed",
481
+ "type": "INT",
482
+ "widget": {
483
+ "name": "noise_seed"
484
+ },
485
+ "link": null
486
+ }
487
+ ],
488
+ "outputs": [
489
+ {
490
+ "name": "NOISE",
491
+ "type": "NOISE",
492
+ "links": [
493
+ 10
494
+ ]
495
+ }
496
+ ],
497
+ "properties": {
498
+ "cnr_id": "comfy-core",
499
+ "ver": "0.33.0",
500
+ "Node name for S&R": "RandomNoise"
501
+ },
502
+ "widgets_values": [
503
+ 1,
504
+ "fixed"
505
+ ]
506
+ },
507
+ {
508
+ "id": 9,
509
+ "type": "BasicGuider",
510
+ "pos": [
511
+ 1050,
512
+ 140
513
+ ],
514
+ "size": [
515
+ 320,
516
+ 60
517
+ ],
518
+ "flags": {},
519
+ "order": 8,
520
+ "mode": 0,
521
+ "inputs": [
522
+ {
523
+ "name": "model",
524
+ "type": "MODEL",
525
+ "link": 5
526
+ },
527
+ {
528
+ "name": "conditioning",
529
+ "type": "CONDITIONING",
530
+ "link": 9
531
+ }
532
+ ],
533
+ "outputs": [
534
+ {
535
+ "name": "GUIDER",
536
+ "type": "GUIDER",
537
+ "links": [
538
+ 11
539
+ ]
540
+ }
541
+ ],
542
+ "properties": {
543
+ "cnr_id": "comfy-core",
544
+ "ver": "0.33.0",
545
+ "Node name for S&R": "BasicGuider"
546
+ }
547
+ },
548
+ {
549
+ "id": 10,
550
+ "type": "KSamplerSelect",
551
+ "pos": [
552
+ 1050,
553
+ 250
554
+ ],
555
+ "size": [
556
+ 320,
557
+ 60
558
+ ],
559
+ "flags": {},
560
+ "order": 9,
561
+ "mode": 0,
562
+ "inputs": [],
563
+ "outputs": [
564
+ {
565
+ "name": "SAMPLER",
566
+ "type": "SAMPLER",
567
+ "links": [
568
+ 12
569
+ ]
570
+ }
571
+ ],
572
+ "properties": {
573
+ "cnr_id": "comfy-core",
574
+ "ver": "0.33.0",
575
+ "Node name for S&R": "KSamplerSelect"
576
+ },
577
+ "widgets_values": [
578
+ "lcm"
579
+ ]
580
+ },
581
+ {
582
+ "id": 11,
583
+ "type": "BasicScheduler",
584
+ "pos": [
585
+ 1050,
586
+ 360
587
+ ],
588
+ "size": [
589
+ 320,
590
+ 110
591
+ ],
592
+ "flags": {},
593
+ "order": 10,
594
+ "mode": 0,
595
+ "inputs": [
596
+ {
597
+ "name": "model",
598
+ "type": "MODEL",
599
+ "link": 6
600
+ },
601
+ {
602
+ "name": "steps",
603
+ "type": "INT",
604
+ "widget": {
605
+ "name": "steps"
606
+ },
607
+ "link": null
608
+ }
609
+ ],
610
+ "outputs": [
611
+ {
612
+ "name": "SIGMAS",
613
+ "type": "SIGMAS",
614
+ "links": [
615
+ 13
616
+ ]
617
+ }
618
+ ],
619
+ "properties": {
620
+ "cnr_id": "comfy-core",
621
+ "ver": "0.33.0",
622
+ "Node name for S&R": "BasicScheduler"
623
+ },
624
+ "widgets_values": [
625
+ "simple",
626
+ 4,
627
+ 1
628
+ ]
629
+ },
630
+ {
631
+ "id": 12,
632
+ "type": "SamplerCustomAdvanced",
633
+ "pos": [
634
+ 1050,
635
+ 520
636
+ ],
637
+ "size": [
638
+ 320,
639
+ 110
640
+ ],
641
+ "flags": {},
642
+ "order": 11,
643
+ "mode": 0,
644
+ "inputs": [
645
+ {
646
+ "name": "noise",
647
+ "type": "NOISE",
648
+ "link": 10
649
+ },
650
+ {
651
+ "name": "guider",
652
+ "type": "GUIDER",
653
+ "link": 11
654
+ },
655
+ {
656
+ "name": "sampler",
657
+ "type": "SAMPLER",
658
+ "link": 12
659
+ },
660
+ {
661
+ "name": "sigmas",
662
+ "type": "SIGMAS",
663
+ "link": 13
664
+ },
665
+ {
666
+ "name": "latent_image",
667
+ "type": "LATENT",
668
+ "link": 14
669
+ }
670
+ ],
671
+ "outputs": [
672
+ {
673
+ "name": "output",
674
+ "type": "LATENT",
675
+ "links": [
676
+ 15,
677
+ 17
678
+ ]
679
+ },
680
+ {
681
+ "name": "denoised_output",
682
+ "type": "LATENT",
683
+ "links": []
684
+ }
685
+ ],
686
+ "properties": {
687
+ "cnr_id": "comfy-core",
688
+ "ver": "0.33.0",
689
+ "Node name for S&R": "SamplerCustomAdvanced"
690
+ }
691
+ },
692
+ {
693
+ "id": 13,
694
+ "type": "VAEDecode",
695
+ "pos": [
696
+ 1420,
697
+ 0
698
+ ],
699
+ "size": [
700
+ 260,
701
+ 50
702
+ ],
703
+ "flags": {
704
+ "collapsed": false
705
+ },
706
+ "order": 12,
707
+ "mode": 0,
708
+ "inputs": [
709
+ {
710
+ "name": "samples",
711
+ "type": "LATENT",
712
+ "link": 15
713
+ },
714
+ {
715
+ "name": "vae",
716
+ "type": "VAE",
717
+ "link": 16
718
+ }
719
+ ],
720
+ "outputs": [
721
+ {
722
+ "name": "IMAGE",
723
+ "type": "IMAGE",
724
+ "links": [
725
+ 19
726
+ ]
727
+ }
728
+ ],
729
+ "properties": {
730
+ "cnr_id": "comfy-core",
731
+ "ver": "0.33.0",
732
+ "Node name for S&R": "VAEDecode"
733
+ }
734
+ },
735
+ {
736
+ "id": 14,
737
+ "type": "VAEDecodeAudio",
738
+ "pos": [
739
+ 1420,
740
+ 100
741
+ ],
742
+ "size": [
743
+ 260,
744
+ 50
745
+ ],
746
+ "flags": {
747
+ "collapsed": false
748
+ },
749
+ "order": 13,
750
+ "mode": 0,
751
+ "inputs": [
752
+ {
753
+ "name": "samples",
754
+ "type": "LATENT",
755
+ "link": 17
756
+ },
757
+ {
758
+ "name": "vae",
759
+ "type": "VAE",
760
+ "link": 18
761
+ }
762
+ ],
763
+ "outputs": [
764
+ {
765
+ "name": "AUDIO",
766
+ "type": "AUDIO",
767
+ "links": [
768
+ 20
769
+ ]
770
+ }
771
+ ],
772
+ "properties": {
773
+ "cnr_id": "comfy-core",
774
+ "ver": "0.33.0",
775
+ "Node name for S&R": "VAEDecodeAudio"
776
+ }
777
+ },
778
+ {
779
+ "id": 15,
780
+ "type": "CreateVideo",
781
+ "pos": [
782
+ 1420,
783
+ 200
784
+ ],
785
+ "size": [
786
+ 260,
787
+ 130
788
+ ],
789
+ "flags": {},
790
+ "order": 14,
791
+ "mode": 0,
792
+ "inputs": [
793
+ {
794
+ "name": "images",
795
+ "type": "IMAGE",
796
+ "link": 19
797
+ },
798
+ {
799
+ "name": "audio",
800
+ "shape": 7,
801
+ "type": "AUDIO",
802
+ "link": 20
803
+ }
804
+ ],
805
+ "outputs": [
806
+ {
807
+ "name": "VIDEO",
808
+ "type": "VIDEO",
809
+ "links": [
810
+ 21
811
+ ]
812
+ }
813
+ ],
814
+ "properties": {
815
+ "cnr_id": "comfy-core",
816
+ "ver": "0.33.0",
817
+ "Node name for S&R": "CreateVideo"
818
+ },
819
+ "widgets_values": [
820
+ 24,
821
+ 8,
822
+ "sRGB",
823
+ "none"
824
+ ]
825
+ },
826
+ {
827
+ "id": 16,
828
+ "type": "SaveVideo",
829
+ "pos": [
830
+ 1720,
831
+ 0
832
+ ],
833
+ "size": [
834
+ 520,
835
+ 560
836
+ ],
837
+ "flags": {},
838
+ "order": 15,
839
+ "mode": 0,
840
+ "inputs": [
841
+ {
842
+ "name": "video",
843
+ "type": "VIDEO",
844
+ "link": 21
845
+ },
846
+ {
847
+ "name": "filename_prefix",
848
+ "type": "STRING",
849
+ "widget": {
850
+ "name": "filename_prefix"
851
+ },
852
+ "link": null
853
+ },
854
+ {
855
+ "name": "format",
856
+ "type": "COMFY_DYNAMICCOMBO_V3",
857
+ "widget": {
858
+ "name": "format"
859
+ },
860
+ "link": null
861
+ },
862
+ {
863
+ "name": "format.codec",
864
+ "type": "COMFY_DYNAMICCOMBO_V3",
865
+ "widget": {
866
+ "name": "format.codec"
867
+ },
868
+ "link": null
869
+ },
870
+ {
871
+ "name": "codec",
872
+ "shape": 7,
873
+ "type": "COMFY_DYNAMICCOMBO_V3",
874
+ "widget": {
875
+ "name": "codec"
876
+ },
877
+ "link": null
878
+ }
879
+ ],
880
+ "outputs": [
881
+ {
882
+ "name": "video",
883
+ "type": "VIDEO",
884
+ "links": []
885
+ }
886
+ ],
887
+ "properties": {
888
+ "cnr_id": "comfy-core",
889
+ "ver": "0.33.0",
890
+ "Node name for S&R": "SaveVideo"
891
+ },
892
+ "widgets_values": [
893
+ "video/dmad_h3_4step_5s_8gb_guitar",
894
+ "auto",
895
+ "auto",
896
+ "auto"
897
+ ]
898
+ },
899
+ {
900
+ "id": 18,
901
+ "type": "H3MemoryOptimization",
902
+ "pos": [
903
+ 0,
904
+ 800
905
+ ],
906
+ "size": [
907
+ 430,
908
+ 200
909
+ ],
910
+ "flags": {},
911
+ "order": 16,
912
+ "mode": 0,
913
+ "inputs": [
914
+ {
915
+ "name": "model",
916
+ "type": "MODEL",
917
+ "link": 3
918
+ },
919
+ {
920
+ "name": "fused_qkv",
921
+ "type": "COMBO",
922
+ "widget": {
923
+ "name": "fused_qkv"
924
+ },
925
+ "link": null
926
+ },
927
+ {
928
+ "name": "mlp_memory",
929
+ "type": "COMBO",
930
+ "widget": {
931
+ "name": "mlp_memory"
932
+ },
933
+ "link": null
934
+ },
935
+ {
936
+ "name": "chunk_rows",
937
+ "type": "INT",
938
+ "widget": {
939
+ "name": "chunk_rows"
940
+ },
941
+ "link": null
942
+ },
943
+ {
944
+ "name": "preserve_precision",
945
+ "type": "BOOLEAN",
946
+ "widget": {
947
+ "name": "preserve_precision"
948
+ },
949
+ "link": null
950
+ },
951
+ {
952
+ "name": "precision_mode",
953
+ "type": "COMBO",
954
+ "widget": {
955
+ "name": "precision_mode"
956
+ },
957
+ "link": null
958
+ },
959
+ {
960
+ "name": "qkv_streaming_mode",
961
+ "type": "COMBO",
962
+ "widget": {
963
+ "name": "qkv_streaming_mode"
964
+ },
965
+ "link": null
966
+ },
967
+ {
968
+ "name": "embedding_memory_mode",
969
+ "type": "COMBO",
970
+ "widget": {
971
+ "name": "embedding_memory_mode"
972
+ },
973
+ "link": null
974
+ },
975
+ {
976
+ "name": "kitchen_v_memory_mode",
977
+ "type": "COMBO",
978
+ "widget": {
979
+ "name": "kitchen_v_memory_mode"
980
+ },
981
+ "link": null
982
+ }
983
+ ],
984
+ "outputs": [
985
+ {
986
+ "name": "MODEL",
987
+ "type": "MODEL",
988
+ "links": [
989
+ 4
990
+ ]
991
+ }
992
+ ],
993
+ "properties": {
994
+ "cnr_id": "h3-optimizations",
995
+ "ver": "0.2.45",
996
+ "Node name for S&R": "H3MemoryOptimization"
997
+ },
998
+ "widgets_values": [
999
+ "auto",
1000
+ "auto",
1001
+ 1024,
1002
+ true,
1003
+ "Auto",
1004
+ "Auto",
1005
+ "Auto",
1006
+ "Lower VRAM (slower)"
1007
+ ]
1008
+ },
1009
+ {
1010
+ "id": 17,
1011
+ "type": "MarkdownNote",
1012
+ "pos": [
1013
+ 480,
1014
+ 720
1015
+ ],
1016
+ "size": [
1017
+ 890,
1018
+ 330
1019
+ ],
1020
+ "flags": {},
1021
+ "order": 18,
1022
+ "mode": 0,
1023
+ "inputs": [],
1024
+ "outputs": [],
1025
+ "properties": {},
1026
+ "widgets_values": [
1027
+ "## DMAD × MiniMax-H3 — 4-step text-to-audio-video (8 GB GPU)\n\n**Models** (put in `ComfyUI/models/...`):\n- `diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors` — [Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3)\n- `text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors` — same repo\n- `vae/minimax_h3_video_vae_fp16.safetensors`, `vae/minimax_h3_audio_vae_fp32.safetensors` — same repo\n- `loras/dmad_minimax_h3_4step_full_critic_comfyui.safetensors` — [ZhengmingYu/DMAD](https://huggingface.co/ZhengmingYu/DMAD/tree/main/minimax_h3)\n\n**Settings that matter:** LoRA strength 1.0 · shift 12 / audio 2 · sampler `lcm` + scheduler `simple` · no CFG.\n**Steps:** 4 = fastest; 8 or 12 give finer detail on fast motion. Avoid `euler` (softer).\n**Prompt:** describe the shot, then `overall_soundscape:` and `non_diegetic_music:` for the audio.\n**8 GB GPUs:** 124 frames = 5 s at 24 fps. The `H3 Memory Optimization` node (Lower VRAM attention, 1024-row\nchunks) and the `H3 AIMDO Residency Limiter` (0 blocks) of [H3-Optimizations](https://github.com/Zironic/H3-Optimizations)\n(install with ComfyUI-Manager) make it fit in 8 GB (it runs with 6 GB free); the model streams from system memory (64 GB RAM\nrecommended).\nFor 15 s use the 4- or 8-step 15 s workflows on a 16 GB GPU or larger.\n"
1028
+ ]
1029
+ }
1030
+ ],
1031
+ "links": [
1032
+ [
1033
+ 1,
1034
+ 1,
1035
+ 0,
1036
+ 2,
1037
+ 0,
1038
+ "MODEL"
1039
+ ],
1040
+ [
1041
+ 2,
1042
+ 2,
1043
+ 0,
1044
+ 3,
1045
+ 0,
1046
+ "MODEL"
1047
+ ],
1048
+ [
1049
+ 3,
1050
+ 3,
1051
+ 0,
1052
+ 18,
1053
+ 0,
1054
+ "MODEL"
1055
+ ],
1056
+ [
1057
+ 4,
1058
+ 18,
1059
+ 0,
1060
+ 19,
1061
+ 0,
1062
+ "MODEL"
1063
+ ],
1064
+ [
1065
+ 5,
1066
+ 19,
1067
+ 0,
1068
+ 9,
1069
+ 0,
1070
+ "MODEL"
1071
+ ],
1072
+ [
1073
+ 6,
1074
+ 19,
1075
+ 0,
1076
+ 11,
1077
+ 0,
1078
+ "MODEL"
1079
+ ],
1080
+ [
1081
+ 7,
1082
+ 4,
1083
+ 0,
1084
+ 7,
1085
+ 0,
1086
+ "CLIP"
1087
+ ],
1088
+ [
1089
+ 8,
1090
+ 5,
1091
+ 0,
1092
+ 7,
1093
+ 1,
1094
+ "VAE"
1095
+ ],
1096
+ [
1097
+ 9,
1098
+ 7,
1099
+ 0,
1100
+ 9,
1101
+ 1,
1102
+ "CONDITIONING"
1103
+ ],
1104
+ [
1105
+ 10,
1106
+ 8,
1107
+ 0,
1108
+ 12,
1109
+ 0,
1110
+ "NOISE"
1111
+ ],
1112
+ [
1113
+ 11,
1114
+ 9,
1115
+ 0,
1116
+ 12,
1117
+ 1,
1118
+ "GUIDER"
1119
+ ],
1120
+ [
1121
+ 12,
1122
+ 10,
1123
+ 0,
1124
+ 12,
1125
+ 2,
1126
+ "SAMPLER"
1127
+ ],
1128
+ [
1129
+ 13,
1130
+ 11,
1131
+ 0,
1132
+ 12,
1133
+ 3,
1134
+ "SIGMAS"
1135
+ ],
1136
+ [
1137
+ 14,
1138
+ 7,
1139
+ 1,
1140
+ 12,
1141
+ 4,
1142
+ "LATENT"
1143
+ ],
1144
+ [
1145
+ 15,
1146
+ 12,
1147
+ 0,
1148
+ 13,
1149
+ 0,
1150
+ "LATENT"
1151
+ ],
1152
+ [
1153
+ 16,
1154
+ 5,
1155
+ 0,
1156
+ 13,
1157
+ 1,
1158
+ "VAE"
1159
+ ],
1160
+ [
1161
+ 17,
1162
+ 12,
1163
+ 0,
1164
+ 14,
1165
+ 0,
1166
+ "LATENT"
1167
+ ],
1168
+ [
1169
+ 18,
1170
+ 6,
1171
+ 0,
1172
+ 14,
1173
+ 1,
1174
+ "VAE"
1175
+ ],
1176
+ [
1177
+ 19,
1178
+ 13,
1179
+ 0,
1180
+ 15,
1181
+ 0,
1182
+ "IMAGE"
1183
+ ],
1184
+ [
1185
+ 20,
1186
+ 14,
1187
+ 0,
1188
+ 15,
1189
+ 1,
1190
+ "AUDIO"
1191
+ ],
1192
+ [
1193
+ 21,
1194
+ 15,
1195
+ 0,
1196
+ 16,
1197
+ 0,
1198
+ "VIDEO"
1199
+ ]
1200
+ ],
1201
+ "groups": [],
1202
+ "config": {},
1203
+ "extra": {
1204
+ "ds": {
1205
+ "scale": 0.6,
1206
+ "offset": [
1207
+ 40,
1208
+ 60
1209
+ ]
1210
+ },
1211
+ "frontendVersion": "1.49.6"
1212
+ },
1213
+ "version": 0.4
1214
+ }
minimax_h3/workflows/dmad_h3_4step_5s_8gb_guitar.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5892a42e099472eaa001df409139f74a312bd8194557a78aa8d50af73152ed51
3
+ size 2607452
minimax_h3/workflows/dmad_h3_8step_15s_wok.json ADDED
@@ -0,0 +1,1167 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "id": "dmad-minimax-h3-t2av-24gb-8step",
3
+ "revision": 0,
4
+ "last_node_id": 18,
5
+ "last_link_id": 20,
6
+ "nodes": [
7
+ {
8
+ "id": 1,
9
+ "type": "UNETLoader",
10
+ "pos": [
11
+ 0,
12
+ 0
13
+ ],
14
+ "size": [
15
+ 430,
16
+ 90
17
+ ],
18
+ "flags": {},
19
+ "order": 0,
20
+ "mode": 0,
21
+ "inputs": [
22
+ {
23
+ "name": "unet_name",
24
+ "type": "COMBO",
25
+ "widget": {
26
+ "name": "unet_name"
27
+ },
28
+ "link": null
29
+ }
30
+ ],
31
+ "outputs": [
32
+ {
33
+ "name": "MODEL",
34
+ "type": "MODEL",
35
+ "links": [
36
+ 1
37
+ ]
38
+ }
39
+ ],
40
+ "properties": {
41
+ "cnr_id": "comfy-core",
42
+ "ver": "0.33.0",
43
+ "Node name for S&R": "UNETLoader",
44
+ "models": [
45
+ {
46
+ "name": "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
47
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors",
48
+ "directory": "diffusion_models"
49
+ }
50
+ ]
51
+ },
52
+ "widgets_values": [
53
+ "minimax_h3_fl2va_pruned_int8_convrot.safetensors",
54
+ "default"
55
+ ]
56
+ },
57
+ {
58
+ "id": 2,
59
+ "type": "LoraLoaderModelOnly",
60
+ "pos": [
61
+ 0,
62
+ 140
63
+ ],
64
+ "size": [
65
+ 430,
66
+ 90
67
+ ],
68
+ "flags": {},
69
+ "order": 1,
70
+ "mode": 0,
71
+ "inputs": [
72
+ {
73
+ "name": "model",
74
+ "type": "MODEL",
75
+ "link": 1
76
+ },
77
+ {
78
+ "name": "lora_name",
79
+ "type": "COMBO",
80
+ "widget": {
81
+ "name": "lora_name"
82
+ },
83
+ "link": null
84
+ },
85
+ {
86
+ "name": "strength_model",
87
+ "type": "FLOAT",
88
+ "widget": {
89
+ "name": "strength_model"
90
+ },
91
+ "link": null
92
+ }
93
+ ],
94
+ "outputs": [
95
+ {
96
+ "name": "MODEL",
97
+ "type": "MODEL",
98
+ "links": [
99
+ 2
100
+ ]
101
+ }
102
+ ],
103
+ "properties": {
104
+ "cnr_id": "comfy-core",
105
+ "ver": "0.33.0",
106
+ "Node name for S&R": "LoraLoaderModelOnly",
107
+ "models": [
108
+ {
109
+ "name": "dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
110
+ "url": "https://huggingface.co/ZhengmingYu/DMAD/resolve/main/minimax_h3/dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
111
+ "directory": "loras"
112
+ }
113
+ ]
114
+ },
115
+ "widgets_values": [
116
+ "dmad_minimax_h3_4step_full_critic_comfyui.safetensors",
117
+ 1
118
+ ]
119
+ },
120
+ {
121
+ "id": 3,
122
+ "type": "MiniMaxH3SigmaShift",
123
+ "pos": [
124
+ 0,
125
+ 280
126
+ ],
127
+ "size": [
128
+ 430,
129
+ 90
130
+ ],
131
+ "flags": {},
132
+ "order": 2,
133
+ "mode": 0,
134
+ "inputs": [
135
+ {
136
+ "name": "model",
137
+ "type": "MODEL",
138
+ "link": 2
139
+ },
140
+ {
141
+ "name": "shift_video",
142
+ "type": "FLOAT",
143
+ "widget": {
144
+ "name": "shift_video"
145
+ },
146
+ "link": null
147
+ },
148
+ {
149
+ "name": "shift_audio",
150
+ "type": "FLOAT",
151
+ "widget": {
152
+ "name": "shift_audio"
153
+ },
154
+ "link": null
155
+ }
156
+ ],
157
+ "outputs": [
158
+ {
159
+ "name": "MODEL",
160
+ "type": "MODEL",
161
+ "links": [
162
+ 3
163
+ ]
164
+ }
165
+ ],
166
+ "properties": {
167
+ "cnr_id": "comfy-core",
168
+ "ver": "0.35.1",
169
+ "Node name for S&R": "MiniMaxH3SigmaShift"
170
+ },
171
+ "widgets_values": [
172
+ 12,
173
+ 2
174
+ ]
175
+ },
176
+ {
177
+ "id": 4,
178
+ "type": "CLIPLoader",
179
+ "pos": [
180
+ 0,
181
+ 420
182
+ ],
183
+ "size": [
184
+ 430,
185
+ 110
186
+ ],
187
+ "flags": {},
188
+ "order": 3,
189
+ "mode": 0,
190
+ "inputs": [
191
+ {
192
+ "name": "clip_name",
193
+ "type": "COMBO",
194
+ "widget": {
195
+ "name": "clip_name"
196
+ },
197
+ "link": null
198
+ }
199
+ ],
200
+ "outputs": [
201
+ {
202
+ "name": "CLIP",
203
+ "type": "CLIP",
204
+ "links": [
205
+ 6
206
+ ]
207
+ }
208
+ ],
209
+ "properties": {
210
+ "cnr_id": "comfy-core",
211
+ "ver": "0.33.0",
212
+ "Node name for S&R": "CLIPLoader",
213
+ "models": [
214
+ {
215
+ "name": "qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
216
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
217
+ "directory": "text_encoders"
218
+ }
219
+ ]
220
+ },
221
+ "widgets_values": [
222
+ "qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors",
223
+ "minimax",
224
+ "default"
225
+ ]
226
+ },
227
+ {
228
+ "id": 5,
229
+ "type": "VAELoader",
230
+ "pos": [
231
+ 0,
232
+ 580
233
+ ],
234
+ "size": [
235
+ 430,
236
+ 60
237
+ ],
238
+ "flags": {},
239
+ "order": 4,
240
+ "mode": 0,
241
+ "inputs": [
242
+ {
243
+ "name": "vae_name",
244
+ "type": "COMBO",
245
+ "widget": {
246
+ "name": "vae_name"
247
+ },
248
+ "link": null
249
+ }
250
+ ],
251
+ "outputs": [
252
+ {
253
+ "name": "VAE",
254
+ "type": "VAE",
255
+ "links": [
256
+ 7,
257
+ 15
258
+ ]
259
+ }
260
+ ],
261
+ "properties": {
262
+ "cnr_id": "comfy-core",
263
+ "ver": "0.33.0",
264
+ "Node name for S&R": "VAELoader",
265
+ "models": [
266
+ {
267
+ "name": "minimax_h3_video_vae_fp16.safetensors",
268
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors",
269
+ "directory": "vae"
270
+ }
271
+ ]
272
+ },
273
+ "widgets_values": [
274
+ "minimax_h3_video_vae_fp16.safetensors"
275
+ ]
276
+ },
277
+ {
278
+ "id": 6,
279
+ "type": "VAELoader",
280
+ "pos": [
281
+ 0,
282
+ 690
283
+ ],
284
+ "size": [
285
+ 430,
286
+ 60
287
+ ],
288
+ "flags": {},
289
+ "order": 5,
290
+ "mode": 0,
291
+ "inputs": [
292
+ {
293
+ "name": "vae_name",
294
+ "type": "COMBO",
295
+ "widget": {
296
+ "name": "vae_name"
297
+ },
298
+ "link": null
299
+ }
300
+ ],
301
+ "outputs": [
302
+ {
303
+ "name": "VAE",
304
+ "type": "VAE",
305
+ "links": [
306
+ 17
307
+ ]
308
+ }
309
+ ],
310
+ "properties": {
311
+ "cnr_id": "comfy-core",
312
+ "ver": "0.33.0",
313
+ "Node name for S&R": "VAELoader",
314
+ "models": [
315
+ {
316
+ "name": "minimax_h3_audio_vae_fp32.safetensors",
317
+ "url": "https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors",
318
+ "directory": "vae"
319
+ }
320
+ ]
321
+ },
322
+ "widgets_values": [
323
+ "minimax_h3_audio_vae_fp32.safetensors"
324
+ ]
325
+ },
326
+ {
327
+ "id": 7,
328
+ "type": "MiniMaxH3ImageToVideo",
329
+ "pos": [
330
+ 480,
331
+ 0
332
+ ],
333
+ "size": [
334
+ 520,
335
+ 560
336
+ ],
337
+ "flags": {},
338
+ "order": 6,
339
+ "mode": 0,
340
+ "inputs": [
341
+ {
342
+ "name": "clip",
343
+ "type": "CLIP",
344
+ "link": 6
345
+ },
346
+ {
347
+ "name": "vae",
348
+ "type": "VAE",
349
+ "link": 7
350
+ },
351
+ {
352
+ "name": "first_frame",
353
+ "shape": 7,
354
+ "type": "IMAGE",
355
+ "link": null
356
+ },
357
+ {
358
+ "name": "last_frame",
359
+ "shape": 7,
360
+ "type": "IMAGE",
361
+ "link": null
362
+ },
363
+ {
364
+ "name": "prompt",
365
+ "type": "STRING",
366
+ "widget": {
367
+ "name": "prompt"
368
+ },
369
+ "link": null
370
+ },
371
+ {
372
+ "name": "width",
373
+ "type": "INT",
374
+ "widget": {
375
+ "name": "width"
376
+ },
377
+ "link": null
378
+ },
379
+ {
380
+ "name": "height",
381
+ "type": "INT",
382
+ "widget": {
383
+ "name": "height"
384
+ },
385
+ "link": null
386
+ },
387
+ {
388
+ "name": "length",
389
+ "type": "INT",
390
+ "widget": {
391
+ "name": "length"
392
+ },
393
+ "link": null
394
+ }
395
+ ],
396
+ "outputs": [
397
+ {
398
+ "name": "positive",
399
+ "type": "CONDITIONING",
400
+ "links": [
401
+ 8
402
+ ]
403
+ },
404
+ {
405
+ "name": "LATENT",
406
+ "type": "LATENT",
407
+ "links": [
408
+ 13
409
+ ]
410
+ }
411
+ ],
412
+ "properties": {
413
+ "cnr_id": "comfy-core",
414
+ "ver": "0.33.0",
415
+ "Node name for S&R": "MiniMaxH3ImageToVideo"
416
+ },
417
+ "widgets_values": [
418
+ "integrated_multimodal_description: [Shot 1] Live-action, cinematic food film at a busy Chinese night market, a medium-wide shot of an old street chef working at his fried rice stall glowing in warm light, his face clearly visible and well lit, a big red sign reading \"DMAD 炒饭\" above the stall, steam rising from the wok behind him and crowds walking past under string lights. [Shot 2] At 00:02.000, the camera cuts to a medium shot of the old street chef in a white tank top at his stall under the red \"DMAD 炒饭\" sign, cracking eggs one-handed into a blackened wok over roaring flames and tossing each empty eggshell into a small bin beside the stove, then stirring the eggs with the spatula, breaking them into small fluffy bits of yellow egg, and tipping in the rice. [Shot 3] At 00:04.000, the camera cuts to a close-up as he tosses the classic Chinese egg fried rice: separate rice grains lightly tinted pale gold, evenly mixed with many small fluffy broken bits of yellow scrambled egg the size of peas, pink shrimp, chopped green scallions and diced carrots high into the air, the rice grains, small bits of yellow egg, shrimp and scallions glistening in slow motion. [Shot 4] At 00:06.500, the camera holds as he splashes wine into the wok and a huge column of orange fire erupts upward, lighting his face and the curious crowd. [Shot 5] At 00:09.000, the camera cuts to an overhead shot of his single wok held above an empty white plate, with no other pan in the frame, and in one continuous motion he tilts the wok and pours all of the classic Chinese egg fried rice: separate rice grains lightly tinted pale gold, evenly mixed with many small fluffy broken bits of yellow scrambled egg the size of peas, pink shrimp, chopped green scallions and diced carrots onto the plate, scraping every last grain out with the spatula, the yellow egg bits pouring out together with the rice, piling into a single mound with plenty of yellow egg, pink shrimp and green scallions clearly visible on top. [Shot 6] At 00:11.500, the camera cuts to a medium shot as he slides the steaming plate of classic Chinese egg fried rice: separate rice grains lightly tinted pale gold, evenly mixed with many small fluffy broken bits of yellow scrambled egg the size of peas, pink shrimp, chopped green scallions and diced carrots, the yellow egg bits clearly visible all over the mound, across the counter to a smiling customer, \"DMAD\" written on the paper napkin beside it, and the camera frames him from the chest up as he smiles and gives a quick thumbs-up to the camera, his face and hand both fully in frame.\n\noverall_soundscape: Bustling night-market chatter, loud sizzling and crackling oil, the clang of the spatula on the wok, a deep whoosh of fire, and the crowd's impressed gasps.\n\nnon_diegetic_music: A cool, rhythmic lo-fi hip hop beat with a jazzy bass line and crisp hand percussion.",
419
+ 1344,
420
+ 768,
421
+ 362
422
+ ]
423
+ },
424
+ {
425
+ "id": 8,
426
+ "type": "RandomNoise",
427
+ "pos": [
428
+ 1050,
429
+ 0
430
+ ],
431
+ "size": [
432
+ 320,
433
+ 90
434
+ ],
435
+ "flags": {},
436
+ "order": 7,
437
+ "mode": 0,
438
+ "inputs": [
439
+ {
440
+ "name": "noise_seed",
441
+ "type": "INT",
442
+ "widget": {
443
+ "name": "noise_seed"
444
+ },
445
+ "link": null
446
+ }
447
+ ],
448
+ "outputs": [
449
+ {
450
+ "name": "NOISE",
451
+ "type": "NOISE",
452
+ "links": [
453
+ 9
454
+ ]
455
+ }
456
+ ],
457
+ "properties": {
458
+ "cnr_id": "comfy-core",
459
+ "ver": "0.33.0",
460
+ "Node name for S&R": "RandomNoise"
461
+ },
462
+ "widgets_values": [
463
+ 4,
464
+ "fixed"
465
+ ]
466
+ },
467
+ {
468
+ "id": 9,
469
+ "type": "BasicGuider",
470
+ "pos": [
471
+ 1050,
472
+ 140
473
+ ],
474
+ "size": [
475
+ 320,
476
+ 60
477
+ ],
478
+ "flags": {},
479
+ "order": 8,
480
+ "mode": 0,
481
+ "inputs": [
482
+ {
483
+ "name": "model",
484
+ "type": "MODEL",
485
+ "link": 4
486
+ },
487
+ {
488
+ "name": "conditioning",
489
+ "type": "CONDITIONING",
490
+ "link": 8
491
+ }
492
+ ],
493
+ "outputs": [
494
+ {
495
+ "name": "GUIDER",
496
+ "type": "GUIDER",
497
+ "links": [
498
+ 10
499
+ ]
500
+ }
501
+ ],
502
+ "properties": {
503
+ "cnr_id": "comfy-core",
504
+ "ver": "0.33.0",
505
+ "Node name for S&R": "BasicGuider"
506
+ }
507
+ },
508
+ {
509
+ "id": 10,
510
+ "type": "KSamplerSelect",
511
+ "pos": [
512
+ 1050,
513
+ 250
514
+ ],
515
+ "size": [
516
+ 320,
517
+ 60
518
+ ],
519
+ "flags": {},
520
+ "order": 9,
521
+ "mode": 0,
522
+ "inputs": [],
523
+ "outputs": [
524
+ {
525
+ "name": "SAMPLER",
526
+ "type": "SAMPLER",
527
+ "links": [
528
+ 11
529
+ ]
530
+ }
531
+ ],
532
+ "properties": {
533
+ "cnr_id": "comfy-core",
534
+ "ver": "0.33.0",
535
+ "Node name for S&R": "KSamplerSelect"
536
+ },
537
+ "widgets_values": [
538
+ "lcm"
539
+ ]
540
+ },
541
+ {
542
+ "id": 11,
543
+ "type": "BasicScheduler",
544
+ "pos": [
545
+ 1050,
546
+ 360
547
+ ],
548
+ "size": [
549
+ 320,
550
+ 110
551
+ ],
552
+ "flags": {},
553
+ "order": 10,
554
+ "mode": 0,
555
+ "inputs": [
556
+ {
557
+ "name": "model",
558
+ "type": "MODEL",
559
+ "link": 5
560
+ },
561
+ {
562
+ "name": "steps",
563
+ "type": "INT",
564
+ "widget": {
565
+ "name": "steps"
566
+ },
567
+ "link": null
568
+ }
569
+ ],
570
+ "outputs": [
571
+ {
572
+ "name": "SIGMAS",
573
+ "type": "SIGMAS",
574
+ "links": [
575
+ 12
576
+ ]
577
+ }
578
+ ],
579
+ "properties": {
580
+ "cnr_id": "comfy-core",
581
+ "ver": "0.33.0",
582
+ "Node name for S&R": "BasicScheduler"
583
+ },
584
+ "widgets_values": [
585
+ "simple",
586
+ 8,
587
+ 1
588
+ ]
589
+ },
590
+ {
591
+ "id": 12,
592
+ "type": "SamplerCustomAdvanced",
593
+ "pos": [
594
+ 1050,
595
+ 520
596
+ ],
597
+ "size": [
598
+ 320,
599
+ 110
600
+ ],
601
+ "flags": {},
602
+ "order": 11,
603
+ "mode": 0,
604
+ "inputs": [
605
+ {
606
+ "name": "noise",
607
+ "type": "NOISE",
608
+ "link": 9
609
+ },
610
+ {
611
+ "name": "guider",
612
+ "type": "GUIDER",
613
+ "link": 10
614
+ },
615
+ {
616
+ "name": "sampler",
617
+ "type": "SAMPLER",
618
+ "link": 11
619
+ },
620
+ {
621
+ "name": "sigmas",
622
+ "type": "SIGMAS",
623
+ "link": 12
624
+ },
625
+ {
626
+ "name": "latent_image",
627
+ "type": "LATENT",
628
+ "link": 13
629
+ }
630
+ ],
631
+ "outputs": [
632
+ {
633
+ "name": "output",
634
+ "type": "LATENT",
635
+ "links": [
636
+ 14,
637
+ 16
638
+ ]
639
+ },
640
+ {
641
+ "name": "denoised_output",
642
+ "type": "LATENT",
643
+ "links": []
644
+ }
645
+ ],
646
+ "properties": {
647
+ "cnr_id": "comfy-core",
648
+ "ver": "0.33.0",
649
+ "Node name for S&R": "SamplerCustomAdvanced"
650
+ }
651
+ },
652
+ {
653
+ "id": 13,
654
+ "type": "VAEDecode",
655
+ "pos": [
656
+ 1420,
657
+ 0
658
+ ],
659
+ "size": [
660
+ 260,
661
+ 50
662
+ ],
663
+ "flags": {
664
+ "collapsed": false
665
+ },
666
+ "order": 12,
667
+ "mode": 0,
668
+ "inputs": [
669
+ {
670
+ "name": "samples",
671
+ "type": "LATENT",
672
+ "link": 14
673
+ },
674
+ {
675
+ "name": "vae",
676
+ "type": "VAE",
677
+ "link": 15
678
+ }
679
+ ],
680
+ "outputs": [
681
+ {
682
+ "name": "IMAGE",
683
+ "type": "IMAGE",
684
+ "links": [
685
+ 18
686
+ ]
687
+ }
688
+ ],
689
+ "properties": {
690
+ "cnr_id": "comfy-core",
691
+ "ver": "0.33.0",
692
+ "Node name for S&R": "VAEDecode"
693
+ }
694
+ },
695
+ {
696
+ "id": 14,
697
+ "type": "VAEDecodeAudio",
698
+ "pos": [
699
+ 1420,
700
+ 100
701
+ ],
702
+ "size": [
703
+ 260,
704
+ 50
705
+ ],
706
+ "flags": {
707
+ "collapsed": false
708
+ },
709
+ "order": 13,
710
+ "mode": 0,
711
+ "inputs": [
712
+ {
713
+ "name": "samples",
714
+ "type": "LATENT",
715
+ "link": 16
716
+ },
717
+ {
718
+ "name": "vae",
719
+ "type": "VAE",
720
+ "link": 17
721
+ }
722
+ ],
723
+ "outputs": [
724
+ {
725
+ "name": "AUDIO",
726
+ "type": "AUDIO",
727
+ "links": [
728
+ 19
729
+ ]
730
+ }
731
+ ],
732
+ "properties": {
733
+ "cnr_id": "comfy-core",
734
+ "ver": "0.33.0",
735
+ "Node name for S&R": "VAEDecodeAudio"
736
+ }
737
+ },
738
+ {
739
+ "id": 15,
740
+ "type": "CreateVideo",
741
+ "pos": [
742
+ 1420,
743
+ 200
744
+ ],
745
+ "size": [
746
+ 260,
747
+ 130
748
+ ],
749
+ "flags": {},
750
+ "order": 14,
751
+ "mode": 0,
752
+ "inputs": [
753
+ {
754
+ "name": "images",
755
+ "type": "IMAGE",
756
+ "link": 18
757
+ },
758
+ {
759
+ "name": "audio",
760
+ "shape": 7,
761
+ "type": "AUDIO",
762
+ "link": 19
763
+ }
764
+ ],
765
+ "outputs": [
766
+ {
767
+ "name": "VIDEO",
768
+ "type": "VIDEO",
769
+ "links": [
770
+ 20
771
+ ]
772
+ }
773
+ ],
774
+ "properties": {
775
+ "cnr_id": "comfy-core",
776
+ "ver": "0.33.0",
777
+ "Node name for S&R": "CreateVideo"
778
+ },
779
+ "widgets_values": [
780
+ 24,
781
+ 8,
782
+ "sRGB",
783
+ "none"
784
+ ]
785
+ },
786
+ {
787
+ "id": 16,
788
+ "type": "SaveVideo",
789
+ "pos": [
790
+ 1720,
791
+ 0
792
+ ],
793
+ "size": [
794
+ 520,
795
+ 560
796
+ ],
797
+ "flags": {},
798
+ "order": 15,
799
+ "mode": 0,
800
+ "inputs": [
801
+ {
802
+ "name": "video",
803
+ "type": "VIDEO",
804
+ "link": 20
805
+ },
806
+ {
807
+ "name": "filename_prefix",
808
+ "type": "STRING",
809
+ "widget": {
810
+ "name": "filename_prefix"
811
+ },
812
+ "link": null
813
+ },
814
+ {
815
+ "name": "format",
816
+ "type": "COMFY_DYNAMICCOMBO_V3",
817
+ "widget": {
818
+ "name": "format"
819
+ },
820
+ "link": null
821
+ },
822
+ {
823
+ "name": "format.codec",
824
+ "type": "COMFY_DYNAMICCOMBO_V3",
825
+ "widget": {
826
+ "name": "format.codec"
827
+ },
828
+ "link": null
829
+ },
830
+ {
831
+ "name": "codec",
832
+ "shape": 7,
833
+ "type": "COMFY_DYNAMICCOMBO_V3",
834
+ "widget": {
835
+ "name": "codec"
836
+ },
837
+ "link": null
838
+ }
839
+ ],
840
+ "outputs": [
841
+ {
842
+ "name": "video",
843
+ "type": "VIDEO",
844
+ "links": []
845
+ }
846
+ ],
847
+ "properties": {
848
+ "cnr_id": "comfy-core",
849
+ "ver": "0.33.0",
850
+ "Node name for S&R": "SaveVideo"
851
+ },
852
+ "widgets_values": [
853
+ "video/dmad_h3_8step_wok",
854
+ "auto",
855
+ "auto",
856
+ "auto"
857
+ ]
858
+ },
859
+ {
860
+ "id": 18,
861
+ "type": "H3MemoryOptimization",
862
+ "pos": [
863
+ 0,
864
+ 800
865
+ ],
866
+ "size": [
867
+ 430,
868
+ 200
869
+ ],
870
+ "flags": {},
871
+ "order": 16,
872
+ "mode": 0,
873
+ "inputs": [
874
+ {
875
+ "name": "model",
876
+ "type": "MODEL",
877
+ "link": 3
878
+ },
879
+ {
880
+ "name": "fused_qkv",
881
+ "type": "COMBO",
882
+ "widget": {
883
+ "name": "fused_qkv"
884
+ },
885
+ "link": null
886
+ },
887
+ {
888
+ "name": "mlp_memory",
889
+ "type": "COMBO",
890
+ "widget": {
891
+ "name": "mlp_memory"
892
+ },
893
+ "link": null
894
+ },
895
+ {
896
+ "name": "chunk_rows",
897
+ "type": "INT",
898
+ "widget": {
899
+ "name": "chunk_rows"
900
+ },
901
+ "link": null
902
+ },
903
+ {
904
+ "name": "preserve_precision",
905
+ "type": "BOOLEAN",
906
+ "widget": {
907
+ "name": "preserve_precision"
908
+ },
909
+ "link": null
910
+ },
911
+ {
912
+ "name": "precision_mode",
913
+ "type": "COMBO",
914
+ "widget": {
915
+ "name": "precision_mode"
916
+ },
917
+ "link": null
918
+ },
919
+ {
920
+ "name": "qkv_streaming_mode",
921
+ "type": "COMBO",
922
+ "widget": {
923
+ "name": "qkv_streaming_mode"
924
+ },
925
+ "link": null
926
+ },
927
+ {
928
+ "name": "embedding_memory_mode",
929
+ "type": "COMBO",
930
+ "widget": {
931
+ "name": "embedding_memory_mode"
932
+ },
933
+ "link": null
934
+ },
935
+ {
936
+ "name": "kitchen_v_memory_mode",
937
+ "type": "COMBO",
938
+ "widget": {
939
+ "name": "kitchen_v_memory_mode"
940
+ },
941
+ "link": null
942
+ }
943
+ ],
944
+ "outputs": [
945
+ {
946
+ "name": "MODEL",
947
+ "type": "MODEL",
948
+ "links": [
949
+ 4,
950
+ 5
951
+ ]
952
+ }
953
+ ],
954
+ "properties": {
955
+ "cnr_id": "h3-optimizations",
956
+ "ver": "0.2.45",
957
+ "Node name for S&R": "H3MemoryOptimization"
958
+ },
959
+ "widgets_values": [
960
+ "auto",
961
+ "auto",
962
+ 4096,
963
+ true,
964
+ "Auto",
965
+ "Auto",
966
+ "Auto",
967
+ "Standard"
968
+ ]
969
+ },
970
+ {
971
+ "id": 17,
972
+ "type": "MarkdownNote",
973
+ "pos": [
974
+ 480,
975
+ 720
976
+ ],
977
+ "size": [
978
+ 890,
979
+ 330
980
+ ],
981
+ "flags": {},
982
+ "order": 17,
983
+ "mode": 0,
984
+ "inputs": [],
985
+ "outputs": [],
986
+ "properties": {},
987
+ "widgets_values": [
988
+ "## DMAD × MiniMax-H3 — 8-step text-to-audio-video (24 GB GPU)\n\n**Models** (put in `ComfyUI/models/...`):\n- `diffusion_models/minimax_h3_fl2va_pruned_int8_convrot.safetensors` — [Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3)\n- `text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors` — same repo\n- `vae/minimax_h3_video_vae_fp16.safetensors`, `vae/minimax_h3_audio_vae_fp32.safetensors` — same repo\n- `loras/dmad_minimax_h3_4step_full_critic_comfyui.safetensors` — [ZhengmingYu/DMAD](https://huggingface.co/ZhengmingYu/DMAD/tree/main/minimax_h3)\n\n**Settings that matter:** LoRA strength 1.0 · shift 12 / audio 2 · sampler `lcm` + scheduler `simple` · no CFG.\n**Steps:** 4 = fastest; 8 or 12 give finer detail on fast motion. Avoid `euler` (softer).\n**Prompt:** describe the shot, then `overall_soundscape:` and `non_diegetic_music:` for the audio.\n**Length:** 362 frames = 15 s at 24 fps (lengths are 5 + 17k: 124 = 5 s, 243 = 10 s, 362 = 15 s). On a 24 GB GPU this\nneeds the `H3 Memory Optimization` node of [H3-Optimizations](https://github.com/Zironic/H3-Optimizations) (install it\nwith ComfyUI-Manager); without it, 5 s clips still fit with stock nodes — bypass the node and set length 124.\n"
989
+ ]
990
+ }
991
+ ],
992
+ "links": [
993
+ [
994
+ 1,
995
+ 1,
996
+ 0,
997
+ 2,
998
+ 0,
999
+ "MODEL"
1000
+ ],
1001
+ [
1002
+ 2,
1003
+ 2,
1004
+ 0,
1005
+ 3,
1006
+ 0,
1007
+ "MODEL"
1008
+ ],
1009
+ [
1010
+ 3,
1011
+ 3,
1012
+ 0,
1013
+ 18,
1014
+ 0,
1015
+ "MODEL"
1016
+ ],
1017
+ [
1018
+ 4,
1019
+ 18,
1020
+ 0,
1021
+ 9,
1022
+ 0,
1023
+ "MODEL"
1024
+ ],
1025
+ [
1026
+ 5,
1027
+ 18,
1028
+ 0,
1029
+ 11,
1030
+ 0,
1031
+ "MODEL"
1032
+ ],
1033
+ [
1034
+ 6,
1035
+ 4,
1036
+ 0,
1037
+ 7,
1038
+ 0,
1039
+ "CLIP"
1040
+ ],
1041
+ [
1042
+ 7,
1043
+ 5,
1044
+ 0,
1045
+ 7,
1046
+ 1,
1047
+ "VAE"
1048
+ ],
1049
+ [
1050
+ 8,
1051
+ 7,
1052
+ 0,
1053
+ 9,
1054
+ 1,
1055
+ "CONDITIONING"
1056
+ ],
1057
+ [
1058
+ 9,
1059
+ 8,
1060
+ 0,
1061
+ 12,
1062
+ 0,
1063
+ "NOISE"
1064
+ ],
1065
+ [
1066
+ 10,
1067
+ 9,
1068
+ 0,
1069
+ 12,
1070
+ 1,
1071
+ "GUIDER"
1072
+ ],
1073
+ [
1074
+ 11,
1075
+ 10,
1076
+ 0,
1077
+ 12,
1078
+ 2,
1079
+ "SAMPLER"
1080
+ ],
1081
+ [
1082
+ 12,
1083
+ 11,
1084
+ 0,
1085
+ 12,
1086
+ 3,
1087
+ "SIGMAS"
1088
+ ],
1089
+ [
1090
+ 13,
1091
+ 7,
1092
+ 1,
1093
+ 12,
1094
+ 4,
1095
+ "LATENT"
1096
+ ],
1097
+ [
1098
+ 14,
1099
+ 12,
1100
+ 0,
1101
+ 13,
1102
+ 0,
1103
+ "LATENT"
1104
+ ],
1105
+ [
1106
+ 15,
1107
+ 5,
1108
+ 0,
1109
+ 13,
1110
+ 1,
1111
+ "VAE"
1112
+ ],
1113
+ [
1114
+ 16,
1115
+ 12,
1116
+ 0,
1117
+ 14,
1118
+ 0,
1119
+ "LATENT"
1120
+ ],
1121
+ [
1122
+ 17,
1123
+ 6,
1124
+ 0,
1125
+ 14,
1126
+ 1,
1127
+ "VAE"
1128
+ ],
1129
+ [
1130
+ 18,
1131
+ 13,
1132
+ 0,
1133
+ 15,
1134
+ 0,
1135
+ "IMAGE"
1136
+ ],
1137
+ [
1138
+ 19,
1139
+ 14,
1140
+ 0,
1141
+ 15,
1142
+ 1,
1143
+ "AUDIO"
1144
+ ],
1145
+ [
1146
+ 20,
1147
+ 15,
1148
+ 0,
1149
+ 16,
1150
+ 0,
1151
+ "VIDEO"
1152
+ ]
1153
+ ],
1154
+ "groups": [],
1155
+ "config": {},
1156
+ "extra": {
1157
+ "ds": {
1158
+ "scale": 0.6,
1159
+ "offset": [
1160
+ 40,
1161
+ 60
1162
+ ]
1163
+ },
1164
+ "frontendVersion": "1.49.6"
1165
+ },
1166
+ "version": 0.4
1167
+ }
minimax_h3/workflows/dmad_h3_8step_15s_wok.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee9abfa4fa2507f994c8dd8d86c8a7e20d00c57facbad2a345e7bfd483045c44
3
+ size 9403948
minimax_h3/workflows/screenshot_4step_15s.png ADDED

Git LFS Details

  • SHA256: eb82817ef139bdc21a69f10dec34918b9ed8e8163a8f86c2c82cfd38a630ebf0
  • Pointer size: 131 Bytes
  • Size of remote file: 477 kB
minimax_h3/workflows/screenshot_4step_5s_8gb.png ADDED

Git LFS Details

  • SHA256: ccc450a1b5419d3cc8018e294d704c268d1f5ab4cb4311e514ef0e21411e9a50
  • Pointer size: 131 Bytes
  • Size of remote file: 425 kB
minimax_h3/workflows/screenshot_8step_15s.png ADDED

Git LFS Details

  • SHA256: 8e328e8cc8596714598e169c369312ee5dfc5b166a397ff6385d6ae79340a43f
  • Pointer size: 131 Bytes
  • Size of remote file: 523 kB
sdxl/dmad_sdxl_1step_frozencritic_unet_fp16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f743d8b2422a8c0f242a1b199f2b3289a872850d9e3c4b3c63d86eade6fb356
3
+ size 5135860277
sdxl/dmad_sdxl_1step_unet_fp16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17a765a5033162ed109e4b81c0b637fab744ce5f53dd75c291ccaf785d635d13
3
+ size 5135837515
sdxl/dmad_sdxl_4step_unet_fp16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02c18fd274bf50eec47b5769ee2a18e9d6a6b497e3824cb6cadeb4127b490a5c
3
+ size 5135837515
wan2.1/dmad_wan2pt1_14B.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e8c7cf5e3fd525258aecb6b34c5e5c767a72db732db21d90fd0cd25115df553d
3
+ size 28577348498
wan2.1/dmad_wan2pt1_1pt3B.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:763249cc7c27f43e64bb21240e840e242c513cdd1c9bcf92b9fe486de579921e
3
+ size 2838249419