Add files using upload-large-folder tool

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+examples/media/before-after.png filter=lfs diff=lfs merge=lfs -text
+examples/media/output.mp4 filter=lfs diff=lfs merge=lfs -text
+examples/media/reference.png filter=lfs diff=lfs merge=lfs -text

LICENSE

@@ -0,0 +1,84 @@
+MiniMax H3 COMMUNITY LICENSE AGREEMENT
+MiniMax H3 release date/License date: August 2, 2026.
+The scope of this License Agreement (this “Agreement”) is expressly limited to the “Applicable Territory” as defined below.
+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.
+I.  Definitions
+1. “Acceptable Use Policy” means the policy published by MiniMax in Exhibit A.
+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.
+3. “Applicable Territory” means worldwide, excluding the Excluded Territories.
+4. “Documentation” means the specifications, manuals, and documentation concerning MiniMax H3 that are publicly released by MiniMax.
+5. “Excluded Territories” means the European Union, the United Kingdom, the Republic of Korea and the United States of America.  
+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.
+7. “MiniMax H3 Works” means (i) the Materials, (ii) the Model Derivatives, and (iii) all derivatives thereof.
+8. “Hosted Services” means hosted services provided via application programming interfaces (APIs), web access, or any other electronic or remote means.
+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.
+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.
+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.
+12. “Output” means any result of operating or otherwise using MiniMax H3 or any Model Derivatives (including through Hosted Services).
+13. “Third Party” means any natural or legal person that is not under common control with us or with you.
+14. “Including” means “including but not limited to.”
+15. “We,” “Us”  or  “MiniMax” means Nanonoble Pte. Ltd..
+II.  Grant of Rights
+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.
+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.
+III.  Distribution and Redistribution
+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:
+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;
+2. You must cause any modified files to carry prominent notices stating that you have modified such files;
+3. You are encouraged to:
+  a. display a notice on any product or service developed using MiniMax H3 indicating that the product or service is “Powered by MiniMax H3”; 
+  b. add an AI-generation identifier to files produced using generative AI models including MiniMax H3; and
+  c. publish at least one technical blog post or a public statement describing your experience using MiniMax H3 Works;
+4. All distributions to Third Parties (other than through Hosted Services) must be accompanied by a “NOTICE” text file containing the following notice:
+  “MiniMax H3 is licensed under the MiniMax H3 Community License Agreement, Copyright © 2026 MiniMax. All Rights Reserved.”
+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.
+IV.  Additional Commercial Terms
+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.
+2. You shall prominently display “MiniMax H3”on the user interface of commercial product or service that uses MiniMax H3 or MiniMax H3 Works.
+V. Use Restrictions
+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. 
+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.
+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).
+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.
+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.
+VI.  Intellectual Property
+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.
+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.
+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.
+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.
+VII.  Disclaimers and Limitations of Liability
+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.
+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.
+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.
+VIII.  Term and Termination
+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.
+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.
+IX.  Governing Law and Jurisdiction
+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.
+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.
+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.
+
+Exhibit A — Acceptable Use Policy
+MiniMax reserves the right to update this Acceptable Use Policy from time to time. 
+Last revised: August 2, 2026.
+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:
+1. Use outside the Applicable Territory;
+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;
+3. Use in any manner that may harm yourself or others;
+4. Use to repurpose or distribute the Outputs of MiniMax H3 or any Model Derivatives in order to harm yourself or others;
+5. Use to circumvent or bypass any safety guardrails or safeguards we have implemented;
+6. Use in any manner that exploits or harms, or intends to exploit or harm, minors;
+7. Use to generate or disseminate verifiably false information and/or content for the purpose of harming others or influencing elections;
+8. Use to manufacture or facilitate false online engagement, including fake reviews and other means of false online engagement;
+9. Use to intentionally defame, disparage, or otherwise harass others;
+10. Use to generate and/or disseminate malware (including ransomware) or any other content intended to damage electronic systems;
+11. Use to generate or disseminate personally identifiable information for the purpose of harming others;
+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;
+13. Use to impersonate another person without that person’s consent, authorization, or lawful right to do so;
+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);
+15. Use in any manner that violates or disregards the social, ethical, or moral standards of other countries or regions;
+16. Use to carry out, assist, threaten, incite, plan, advocate for, or encourage violent extremism or terrorism;
+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;
+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;
+19. Use for military purposes;
+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.

MODIFICATIONS.md

@@ -0,0 +1,67 @@
+# Modified files
+
+Section III.2 of the MiniMax H3 Community License Agreement requires that modified
+files carry a prominent notice saying so. This file is that notice.
+
+Everything below is derived from
+[`MiniMaxAI/MiniMax-H3`](https://huggingface.co/MiniMaxAI/MiniMax-H3).
+
+## `transformer/` — modified
+
+**Every weight file in `transformer/` has been modified.** It started as the base
+model's `transformer_ref/` (the `ref2va` transformer, 33.1 B parameters) and every
+parameter was updated by a full finetune on a character-replacement objective. The
+architecture, `config.json` and tensor names are unchanged, so it is a drop-in
+replacement for the base `transformer_ref/`; the numbers in it are not the base
+model's numbers.
+
+The file layout also differs: the finetune was written as one 61.7 GiB safetensors
+file and re-sharded here into 14 parts, because HuggingFace rejects single files
+above 50 GB. The 638 tensors and their contents are unchanged by that re-sharding.
+
+## `lora/pytorch_lora_weights.safetensors` — new
+
+Not a MiniMax file. A rank-128 LoRA over 302 linear layers of `transformer/`,
+trained by us with DMD2 distillation. It is a delta on the finetuned transformer
+above, not on the base model — loading it onto stock `transformer_ref/` produces
+garbage.
+
+## `assets/fixed_embed_fwd_anyframe.pt` — new
+
+Not a MiniMax file. A frozen 362 × 5120 text-conditioning tensor we computed once
+with the base model's own text encoder, so that inference never has to load
+Qwen3-VL. It is an *output* of the base model's encoder in the sense of Section
+I.12, computed from the prompt in `assets/fixed_prompt.txt`.
+
+## `assets/fixed_prompt.txt` — new
+
+Not a MiniMax file. The prompt text the tensor above was computed from, included so
+that what conditions every render is readable rather than opaque.
+
+## `inference/sample.py`, `examples/demo.sh` — new
+
+Not MiniMax files. Written by us against the public `diffusers` API.
+
+## `examples/media/` — new, and derived from a MiniMax demo clip
+
+The clips this model was developed and evaluated against are user footage and are not
+ours to publish, so the demo here was built instead from `assets/ref2va.mp4`, a video
+MiniMax published with the base model. That clip is itself a MiniMax H3 generation, not
+camera footage, so no filmed performer appears anywhere in this repository. All three
+files derive from a 512 × 768 portrait crop of it (`crop=512:768:389:0`, no scaling):
+
+- `reference.png` — the crop's first frame with the young man repainted as an invented
+  elderly woman. Produced with OpenAI's `gpt-image-2`; the character is fictional and is
+  not a real person or an existing property.
+- `output.mp4` — that reference propagated across 124 frames by this model.
+- `before-after.png` — frames from the driving crop above frames from the output.
+
+The driving clip itself is **not** bundled. `examples/demo.sh` rebuilds it, with the
+documented crop, from your own copy of the base model.
+
+## Not redistributed here
+
+The VAE, audio VAE, schedulers, text encoder, tokenizer and processor are **not**
+included in this repository and are not modified. They are loaded at runtime from
+your own copy of `MiniMaxAI/MiniMax-H3`, which you must download separately and
+under its own license terms.

NOTICE

@@ -0,0 +1,13 @@
+MiniMax H3 is licensed under the MiniMax H3 Community License Agreement,
+Copyright © 2026 MiniMax. All Rights Reserved.
+
+---
+
+Viggle-Animate is a Model Derivative of MiniMax H3, as that term is defined in
+Section I.11 of the MiniMax H3 Community License Agreement. It is distributed
+under that same Agreement, a copy of which is included in this repository as
+LICENSE. See MODIFICATIONS.md for the list of files that were modified.
+
+Powered by MiniMax H3.
+
+Modifications and additions Copyright © 2026 Viggle AI.

README.md

@@ -0,0 +1,232 @@
+---
+license: other
+license_name: minimax-h3-community-license
+license_link: LICENSE
+base_model: MiniMaxAI/MiniMax-H3
+pipeline_tag: video-to-video
+library_name: diffusers
+tags:
+  - video-editing
+  - character-replacement
+  - video-to-video
+  - distillation
+  - dmd
+extra_gated_prompt: >-
+  Viggle-Animate is a Model Derivative of MiniMax H3 and is distributed under the
+  MiniMax H3 Community License Agreement, not an open-source license. Two terms in
+  particular are easy to miss.
+
+
+  **Territory.** The Agreement grants rights worldwide EXCEPT in the European Union,
+  the United Kingdom, the Republic of Korea and the United States of America. In
+  those Excluded Territories it grants you nothing, and Section V.4 forbids using,
+  reproducing, modifying, distributing or displaying the Works or their Outputs
+  there. If you are in an Excluded Territory, contact MiniMax at api@minimax.io for
+  a separate license before downloading.
+
+
+  **Acceptable use.** This model replaces the person in a video with whoever you paint
+  into its first frame. Do not use it on anyone who has not consented, and read the
+  Acceptable Use Policy in Exhibit A of the Agreement.
+extra_gated_fields:
+  Name: text
+  Organization: text
+  Country: country
+  I am not located in the EU, UK, South Korea or the USA, or I hold a separate license from MiniMax: checkbox
+  I have read and agree to the MiniMax H3 Community License Agreement and its Acceptable Use Policy: checkbox
+  I will not use this model on people who have not consented: checkbox
+---
+
+# Viggle-Animate
+
+**Edit one frame of a video. This model propagates that edit across the whole clip**, in
+4 sampling steps.
+
+![driving video and output](examples/media/before-after.png)
+
+You give it two things: the driving video, and that video's **first frame with the
+character repainted** — same camera, same pose, same background, same light, a different
+person in it. The model re-renders the clip so the new character performs the original
+motion. Everything else holds: framing, background, timing, the camera's push-in, the
+lamb sitting in the same place in her arms in every frame that it sat in his.
+
+```
+sample.py --cond driving.mp4 --ref repainted_first_frame.png --out output.mp4
+```
+
+The repainted frame is not something this model makes. It comes from an image editor —
+the one above was produced with OpenAI's `gpt-image-2` from a single instruction: replace
+the young man with an elderly woman, keep the pose, the suit, the lamb and the hillside
+identical. **This model is the second half of that pipeline, not the whole of it.**
+
+There is no text prompt at this stage. Conditioning is a frozen embedding computed once
+(see [`assets/fixed_prompt.txt`](assets/fixed_prompt.txt)), so the text encoder is never
+loaded and nothing in the output comes from a caption you write. All of the art direction
+happens in the frame you paint.
+
+## What this is
+
+A two-part derivative of [`MiniMaxAI/MiniMax-H3`](https://huggingface.co/MiniMaxAI/MiniMax-H3)'s
+`ref2va` transformer:
+
+| | |
+|---|---|
+| `transformer/` | 33.1 B parameters, bf16, 14 shards. A **full finetune** of the base `transformer_ref` on a character-replacement objective. Same architecture and tensor names as the base; different numbers |
+| `lora/` | rank 128 over 302 linear layers, 2.5 GB. A **DMD2-distilled** delta on the finetune above. This is what turns 30 sampling steps into 4 |
+
+The LoRA is a delta on the *finetuned* transformer, not on the base. Loading it onto
+stock `transformer_ref` produces garbage.
+
+## Requirements
+
+You need your own copy of the base model — this repository ships only the transformer
+and the LoRA, and loads the VAE, audio VAE and schedulers from yours at runtime.
+
+The base repository is about 269 GB, but inference here touches 11 GB of it. The
+transformer in this repository replaces the base one, and the text encoder is never
+loaded at all, so both can be skipped:
+
+```bash
+hf download MiniMaxAI/MiniMax-H3 --local-dir ./MiniMax-H3 \
+  --include "modular_model_index.json" "vae/*" "audio_vae/*" \
+            "scheduler/*" "audio_scheduler/*" \
+            "assets/ref2va.mp4"
+
+hf download Viggle/Viggle-Animate --local-dir ./Viggle-Animate
+
+pip install torch "git+https://github.com/huggingface/diffusers@d6726f3" av
+```
+
+That is 11 GB from the base model — the VAE is 9.8 GB of it — plus 65 GB here. The last
+`--include` is the demo clip [`examples/demo.sh`](examples/demo.sh) needs; drop it if you
+are bringing your own footage. Downloading the full base repository instead is fine and
+changes nothing, it just costs 259 GB you will not read.
+
+`d6726f3` is the `diffusers` commit this release was tested against. No fork and no
+patch is required: the `minimax_h3` modular pipeline upstream is sufficient.
+
+**One 80 GB card is not enough at bf16.** The transformer alone is 62 GiB resident, and
+a 480×832 / 124-frame render peaks at **80.1 GiB allocated** — over an 80 GB card's
+usable capacity before the allocator's own overhead. Either use a card with ≥ 96 GB, or
+pass `--offload`, which streams the transformer from CPU in groups of 5 blocks and holds
+about 12 GB of weights on the GPU instead of 62 — much slower, but it fits. Larger
+canvases and longer clips both push this up.
+
+## Run it
+
+```bash
+python Viggle-Animate/inference/sample.py \
+  --model-dir ./MiniMax-H3 \
+  --cond driving.mp4 \
+  --ref  character.png \
+  --out  swapped.mp4
+```
+
+Defaults are the configuration this model was evaluated at: `--steps 4`,
+`--flow-shift 3`, `--num-frames 124` (≈ 5.2 s at 24 fps), and both references laid
+out on the driving clip's own short edge. `--seed 42`. Output geometry defaults to
+the driving clip's own resolution, which must be a multiple of 32 on both axes.
+
+### Preparing the reference
+
+The reference is the driving clip's first frame, repainted. Extract it, edit it, keep
+the geometry:
+
+```bash
+ffmpeg -i driving.mp4 -frames:v 1 first_frame.png
+# hand first_frame.png to an image editor, ask for the character swap, save the
+# result at exactly the same resolution as driving.mp4
+```
+
+What the editor is asked for matters more than which editor it is. Name the change, then
+pin down everything that must not change — pose, hand positions, held props, camera
+framing, background, lighting. The video model follows the painted frame closely, so a
+reference where the editor quietly reframed the shot or moved an arm will produce a clip
+that fights the driving motion.
+
+[`examples/demo.sh`](examples/demo.sh) reproduces the clip at the top of this page. It
+builds the driving footage from a video that came with the base model and pairs it with
+the repainted frame in [`examples/media/`](examples/media), so it needs no media from
+you and has a right answer you can check against `examples/media/output.mp4`.
+
+## Speed
+
+Measured on one B200, 480×832, 124 frames at 24 fps, bf16, no compile:
+
+| | |
+|---|---|
+| Sampling | **13.5 s** |
+| Forward passes | **3** (`--steps 4` is a 4-point sigma grid, so 3 NFE) |
+| Per forward | ~4.5 s |
+| VAE decode and mux | ~11 s |
+| Weight load, once per process | ~25 s |
+| Peak memory | 80.1 GiB |
+
+`--attn-backend`-style tuning is not wired into this script; the numbers above are the
+plain PyTorch path.
+
+## Known limitations
+
+These are measured, not hypothetical.
+
+- **A character photograph is not a reference, and the failure is silent.** If you pass
+  a standalone portrait of your character — a different scene, a different pose, a
+  different background — the model does not swap anyone. It reconstructs the driving
+  clip almost unchanged and reports no error. We rendered four such clips before
+  noticing. The reference must be the driving clip's own first frame, repainted; that
+  is what the model was trained on, and when the reference and the driving video
+  disagree about the scene, the driving video wins.
+- **Everything you want in the output has to survive the paint step.** The video model
+  will not invent a costume detail, a prop or a face the repainted frame does not show.
+  Judge the reference frame before you spend a render on it.
+- **It renders sharper than its own teacher** — Laplacian variance runs 16–50 % above
+  the 30-step model it was distilled from. Some of that reads as detail and some reads
+  as over-sharpening, and it gets worse with more sampling steps, not better.
+- **More steps is not monotonically better.** `--steps 4` is the operating point, not
+  a degraded version of `--steps 12`. If 4 steps looks wrong, more steps is unlikely
+  to be the fix.
+- **A single still carries limited information.** Props, accessories, garment volume
+  and prints that are visible in the reference are not always carried into the render,
+  and detail behind the character at frame 0 has to be guessed as the body moves off it.
+- **Faces are driven by the reference and the driving motion only** — there is no
+  dedicated facial-landmark channel, so tight close-up dialogue is the weakest case.
+- **It was trained on one subject filling a portrait frame.** Wide shots, several people
+  in frame, and landscape framing are all out of distribution; the demo above is a 2:3
+  crop of a landscape source for exactly that reason.
+- **The driving clip's audio is dropped** at input. The model emits an audio track of
+  its own; the fixed prompt asks for silence.
+- **Frame count is not free.** 124 frames is the evaluated window; longer renders
+  need proportionally more memory.
+
+## License
+
+**This is not an open-source release.** Viggle-Animate is a Model Derivative of
+MiniMax H3 under Section I.11 of the
+[MiniMax H3 Community License Agreement](LICENSE) — the Agreement names distillation
+explicitly — and is distributed under that same Agreement. Your obligations, in
+summary and without substituting for reading it:
+
+- **The Agreement does not cover the EU, the UK, South Korea or the USA** (I.5, II,
+  V.4). In those territories it grants no rights; contact api@minimax.io.
+- If you redistribute, pass along the Agreement, keep the [`NOTICE`](NOTICE) file, and
+  mark your modified files ([III.1–4](LICENSE)). See [`MODIFICATIONS.md`](MODIFICATIONS.md)
+  for ours.
+- Commercial products above 20 M USD annual revenue need prior written authorization
+  from MiniMax, and must display "MiniMax H3" in the product UI (IV.1, IV.2).
+- The Acceptable Use Policy in Exhibit A applies to everything you generate.
+
+The base model's text encoder is Qwen3-VL-32B (Apache 2.0), but it is neither used nor
+redistributed here.
+
+## Intended use and misuse
+
+This model exists to put a consenting performer into footage they did not shoot. It
+will just as readily put someone into footage they never agreed to appear in. Note
+where that decision is actually made: **the identity comes from the frame you paint**,
+so an image editor's safeguards are upstream of this model and none of them are in it.
+This model does not verify identity or consent, and cannot. **Do not run it on people who have not agreed to
+it.** Section V.5 of the Agreement requires anyone offering this as a service to build
+and maintain safeguards against exactly that, and to act on reports of abuse. Label
+what you generate as AI-generated.
+
+Powered by MiniMax H3.

assets/fixed_embed_fwd_anyframe.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0ae90929caf7b790f5d0de599e868cc6c179f3e969577627971c9d78936a058
+size 3714413

assets/fixed_prompt.txt

@@ -0,0 +1,16 @@
+<Video 1>: <0.2 seconds><1.2 seconds><2.2 seconds><3.2 seconds><4.2 seconds><5.0 seconds><Picture 1>: subject_definitions:
+<Video 1> is the source video for the editing task.
+<Picture 1> is one frame of the target video.
+
+summary:
+[video editing + character replacement] The target video is an edited version of <Video 1> in which every performer is replaced by a different person. <Picture 1> is one frame of the target video: it already shows the replacement, together with the background, camera framing and lighting it happens in. Every other frame of the target video shows those same people in that same place, following the motion, timing and camera of <Video 1>.
+
+retention_analysis:
+<Video 1> (source video editing): fully_preserved - the camera framing, the background, the lighting, and the motion and timing of every performance are copied frame for frame.
+<Picture 1> (appears in [Shot 1]): reference - one frame of the target video, pixel for pixel. The people in it, their faces, hair, skin tone, build and clothing, and the background, the framing and the lighting are all taken from <Picture 1> and held unchanged from the first frame to the last.
+
+detailed_description:
+[Shot 1] The people of <Picture 1> perform exactly the motion of the corresponding performers in <Video 1>, in the same framing, on the same background, under the same lighting. Nothing outside the people changes. The camera framing never changes through the end of the video.
+
+overall_soundscape:
+No music and no speech.

examples/demo.sh

@@ -0,0 +1,43 @@
+#!/usr/bin/env bash
+# Reproduce the demo on the model card, and in doing so check your install.
+#
+# This repository bundles no driving footage. It builds the driving clip from a demo
+# video that ships with the base model, using the exact crop documented below, and
+# pairs it with the repainted first frame in media/reference.png.
+#
+# The result should match media/output.mp4. On the same GPU model we get it back
+# bit-identical; on different hardware bf16 kernel scheduling shifts things, and a mean
+# absolute difference around 1.5/255 is normal. What matters is that it is the same
+# elderly woman holding the same black lamb. If it comes back as the young man from the
+# source video instead, the LoRA did not load. If it comes back as noise, the weights
+# are wrong.
+#
+#   ./examples/demo.sh /path/to/MiniMax-H3
+#
+# About a minute on a B200, most of it loading weights.
+
+set -euo pipefail
+
+MODEL_DIR="${1:?usage: demo.sh /path/to/MiniMax-H3}"
+HERE="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+OUT="${OUT:-$HERE/demo_out}"
+mkdir -p "$OUT"
+
+SRC="$MODEL_DIR/assets/ref2va.mp4"
+[ -f "$SRC" ] || { echo "missing $SRC -- see the download command on the model card"; exit 1; }
+
+# The source is 1344x768 and exactly 124 frames, which is the sampler's window. Crop a
+# 512x768 portrait window around the figure; it stays in frame for the whole push-in, so
+# no scaling and no padding are needed. media/reference.png is this crop's first frame,
+# repainted.
+ffmpeg -y -loglevel error -i "$SRC" \
+  -vf "crop=512:768:389:0" -frames:v 124 -an "$OUT/driving.mp4"
+
+python "$HERE/../inference/sample.py" \
+  --model-dir "$MODEL_DIR" \
+  --cond "$OUT/driving.mp4" \
+  --ref  "$HERE/media/reference.png" \
+  --out  "$OUT/output.mp4"
+
+echo
+echo "wrote $OUT/output.mp4 -- compare against $HERE/media/output.mp4"

examples/media/output.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b89a33a902cd92a1068307b701205029682bb5e5db0326775e3490767f1df04
+size 880425

inference/sample.py

@@ -0,0 +1,123 @@
+"""Viggle-Animate: replace the performers in a video with the person in a still.
+
+    python inference/sample.py --cond driving.mp4 --ref character.png --out swapped.mp4
+
+`--cond` supplies the motion, camera framing, background and lighting; `--ref` supplies
+who is in it. Everything about the render except the people is copied from `--cond`.
+
+The text encoder is never loaded. Conditioning comes from `assets/fixed_embed_fwd_anyframe.pt`,
+a frozen 362 x 5120 tensor computed once from the fixed prompt in `assets/fixed_prompt.txt`,
+so Qwen3-VL (63 GB of the base repo) stays on disk and the text block of the packed sequence
+is 362 rows instead of several thousand. There is no per-clip prompt and no caption: nothing
+in the output comes from text you write.
+
+Needs `--model-dir` pointing at a local copy of MiniMaxAI/MiniMax-H3 for the VAE, the audio
+VAE and the schedulers. This repository ships only the transformer and the LoRA.
+"""
+
+import argparse
+import os
+import time
+
+import torch
+from diffusers import MiniMaxH3Transformer3DModel, ModularPipeline
+from diffusers.modular_pipelines.minimax_h3 import MiniMaxH3ImageReference, MiniMaxH3VideoReference
+from diffusers.modular_pipelines.minimax_h3.encoders import MiniMaxH3Ref2VATextEncoderStep
+from diffusers.utils.export_utils import encode_video
+
+HERE = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+
+parser = argparse.ArgumentParser()
+parser.add_argument("--cond", required=True, help="the video whose motion, framing and background are kept")
+parser.add_argument("--ref", required=True, help="a single still of the person to put in it")
+parser.add_argument("--out", required=True)
+parser.add_argument("--model-dir", required=True,
+                    help="a local copy of MiniMaxAI/MiniMax-H3, for the VAE / audio VAE / schedulers")
+parser.add_argument("--transformer", default=os.path.join(HERE, "transformer"))
+parser.add_argument("--lora", default=os.path.join(HERE, "lora"))
+parser.add_argument("--embed", default=os.path.join(HERE, "assets", "fixed_embed_fwd_anyframe.pt"))
+parser.add_argument("--num-frames", type=int, default=124, help="at 24 fps; 124 frames is ~5.2 s")
+parser.add_argument("--steps", type=int, default=4,
+                    help="the distilled student's operating point. More is not monotonically better: "
+                         "s4 is not a degraded s12")
+parser.add_argument("--flow-shift", type=float, default=3.0,
+                    help="the base model's released default is 12; the few-step student wants 3")
+parser.add_argument("--height", type=int, default=None, help="defaults to the conditioning clip's own height")
+parser.add_argument("--width", type=int, default=None, help="defaults to the conditioning clip's own width")
+parser.add_argument("--short-edge", type=int, default=None,
+                    help="the canvas both references are laid out on. Defaults to the conditioning clip's own "
+                         "short edge, which is what this model was evaluated at")
+parser.add_argument("--offload", action="store_true",
+                    help="stream the transformer from CPU in groups of 5 blocks: ~12 GB resident instead of 62")
+parser.add_argument("--seed", type=int, default=42)
+args = parser.parse_args()
+
+fixed = torch.load(args.embed, weights_only=False)
+
+
+def use_fixed_embeds(self, components, state):
+    block_state = self.get_block_state(state)
+    block_state.prompt_embeds = fixed["prompt_embeds"].to(components._execution_device, torch.bfloat16)
+    block_state.text_token_tags = fixed["text_token_tags"]
+    self.set_block_state(state, block_state)
+    return components, state
+
+
+MiniMaxH3Ref2VATextEncoderStep.__call__ = use_fixed_embeds
+
+# The reference order is frozen: the presentation names `<Video 1>` then `<Picture 1>`, and that order
+# advances the shared rotary clock, so it is part of the layout rather than a detail of the prompt. The
+# driving clip's own soundtrack is dropped, as it is in training.
+video = MiniMaxH3VideoReference.from_file(args.cond)
+video.audio, video.sample_rate = None, None
+
+# Passing an orientation that disagrees with the clip silently generates a transposed video, so the output
+# geometry is derived from the clip rather than typed.
+height = args.height or video.frames.shape[1]
+width = args.width or video.frames.shape[2]
+short_edge = args.short_edge or min(height, width)
+
+pipe = ModularPipeline.from_pretrained(args.model_dir, workflow="ref2va")
+# Both references are pinned to the target's own short edge. The base model's released defaults (768 for the
+# video reference, 2048 for the image) put the references on a grid the target never shares; this model was
+# finetuned and evaluated with them nested, and changing it changes the take.
+pipe.register_to_config(canvas_short_edge=short_edge,
+                        canvas_max_pixels=short_edge * max(height, width),
+                        reference_image_short_edge=short_edge)
+
+t0 = time.time()
+# `transformer_ref` is deliberately absent: our finetune replaces it outright, so loading the base copy
+# first would read 62 GB off disk only to drop it.
+pipe.load_components(names=["vae", "audio_vae", "scheduler", "audio_scheduler"],
+                     pretrained_model_name_or_path=args.model_dir, dtype=torch.bfloat16)
+pipe.transformer_ref = MiniMaxH3Transformer3DModel.from_pretrained(args.transformer, torch_dtype=torch.bfloat16)
+# `prefix=None` and the explicit `weight_name` are both required. The loader defaults to looking for a `.bin`
+# (raises) and to filtering keys for a `transformer.` prefix, which these bare keys do not have -- that
+# mismatch loads *nothing* and only warns, so the default would silently render the un-distilled model.
+pipe.transformer_ref.load_lora_adapter(args.lora, weight_name="pytorch_lora_weights.safetensors", prefix=None)
+pipe.scheduler.set_shift(args.flow_shift)
+
+if args.offload:
+    pipe.transformer_ref.enable_group_offload(
+        onload_device=torch.device("cuda"), offload_type="block_level", num_blocks_per_group=5,
+        non_blocking=True, use_stream=True, record_stream=True)
+    pipe.vae.to("cuda")
+    pipe.audio_vae.to("cuda")
+else:
+    pipe.to("cuda")
+print(f"loaded in {time.time() - t0:.0f}s; canvas {height}x{width}, references on short edge {short_edge}")
+
+t0 = time.time()
+result = pipe(
+    prompt=fixed["presentation"],
+    references=[video, MiniMaxH3ImageReference.from_file(args.ref)],
+    num_frames=args.num_frames,
+    height=height,
+    width=width,
+    num_inference_steps=args.steps,
+    generator=torch.Generator().manual_seed(args.seed),
+    output=["videos", "audio", "sampling_rate"],
+)
+encode_video(result["videos"][0], fps=24, output_path=args.out,
+             audio=result["audio"][0], audio_sample_rate=result["sampling_rate"])
+print(f"{time.time() - t0:.0f}s, peak {torch.cuda.max_memory_allocated() / 2**30:.1f} GiB -> {args.out}")

lora/pytorch_lora_weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bcdd60965d95e45a0b06f7d27435f21cbc2c8bd4f26f7b1cd8e66af7122c4af8
+size 2666439896

requirements.txt

@@ -0,0 +1,7 @@
+# Tested with these. Nothing here is pinned tightly except diffusers, which is pinned to a
+# commit rather than a release: the `minimax_h3` modular pipeline is newer than any tag.
+torch==2.9.1
+git+https://github.com/huggingface/diffusers@d6726f3
+transformers==4.57.3
+safetensors==0.7.0
+av==16.1.0

transformer/config.json

@@ -0,0 +1,26 @@
+{
+  "_class_name": "MiniMaxH3Transformer3DModel",
+  "_diffusers_version": "0.36.0.dev0",
+  "num_attention_heads": 56,
+  "attention_head_dim": 128,
+  "hidden_size": 5376,
+  "num_layers": 50,
+  "num_refiner_layers": 2,
+  "ffn_dim": 14336,
+  "in_channels": 24,
+  "audio_in_channels": 32,
+  "patch_size": [
+    1,
+    2,
+    2
+  ],
+  "text_dim": 5120,
+  "freq_dim": 256,
+  "time_embed_hidden_dim": 5376,
+  "time_embed_dim": 2688,
+  "rope_freq_dim": 16,
+  "rope_theta": 10000.0,
+  "norm_eps": 1e-05,
+  "qk_norm_eps": 1e-05,
+  "final_norm_eps": 1e-05
+}

transformer/diffusion_pytorch_model-00001-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fb30791125c5814baaefb5bac1e3de2cb9595e027866ffa3a1d1c946789ef427
+size 4945656288

transformer/diffusion_pytorch_model-00002-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58d254c72eb9637b86b11be13f2d648e4dd7c37faddf0fd09f7808faef83e6a4
+size 4490213792

transformer/diffusion_pytorch_model-00003-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd8410138703641d2f6475f5650aaaa6e2c335b12a992b88aab4cb6465666f3d
+size 4701942704

transformer/diffusion_pytorch_model-00004-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c42d981e9fa75379f14622855f045f2dee79f75b91ced7eeb0247bd4aab1cd95
+size 4933368952

transformer/diffusion_pytorch_model-00005-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2b6fb580aef372b651083570c7c25edfb9dce065b49a3f605499b31109d0f9d2
+size 4567284240

transformer/diffusion_pytorch_model-00006-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3183d396d2fc2d695145c98988742849658561e7e2f69ce8c3a78d414a363a0
+size 4701942704

transformer/diffusion_pytorch_model-00007-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89fb79d363342bbf09735a0ddd8ae6bb782cfe829f1c530c99fb40b47fafa19c
+size 4933368952

transformer/diffusion_pytorch_model-00008-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b2613f24f59acb2822ededd73f609343eaa65a29c7e6769d825838f159a2a17
+size 4567284240

transformer/diffusion_pytorch_model-00009-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82f6272df4fdf01264ba0172896c3ef0962581346fc23f4d18e54b534b1ae051
+size 4701942704

transformer/diffusion_pytorch_model-00010-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a21d6103074bfd21c6e6843e959722204ea669d612eaed70f97f5966b8620e5a
+size 4933368952

transformer/diffusion_pytorch_model-00011-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a2b57d4cccf22ef041b4e9675ec3827a6048fc1a1e3cf19fe021bd07841918d
+size 4567284240

transformer/diffusion_pytorch_model-00012-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2cf2825b56586f34bb79df4ef287a0ef8f72a2b61917042858cfaf150a50ad0b
+size 4701942704

transformer/diffusion_pytorch_model-00013-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:291fded8f271e6b639cbf40b561fd109fdcaa398024d77bd7d3f8abdbdd8f487
+size 4933368952

transformer/diffusion_pytorch_model-00014-of-00014.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:145d6c9dfc17738427eb2e7fd794a47e6397d181d8e5c798286ca9134f840955
+size 4567090544

transformer/diffusion_pytorch_model.safetensors.index.json

@@ -0,0 +1,645 @@
+{
+  "metadata": {
+    "total_size": 66245985792
+  },
+  "weight_map": {
+    "audio_proj_in.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "audio_proj_in.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "audio_proj_out.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "audio_proj_out.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "context_embedder.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "context_embedder.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "norm_out.linear.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "norm_out.linear.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "norm_out.norm.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "proj_in.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "proj_in.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "proj_out.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "proj_out.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "time_embedder.linear_1.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "time_embedder.linear_1.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "time_embedder.linear_2.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "time_embedder.linear_2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.final_norm.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.attn.to_out.0.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.ff.net.0.proj.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.ff.net.2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.norm1.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.0.norm2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.attn.to_out.0.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.ff.net.0.proj.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.ff.net.2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.norm1.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "token_refiner.refiner_blocks.1.norm2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.adaln_proj.linear.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.adaln_proj.linear.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.attn.to_out.0.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.ff.net.0.proj.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.ff.net.2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.norm1.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.0.norm2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.adaln_proj.linear.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.adaln_proj.linear.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.attn.to_out.0.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.ff.net.0.proj.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.ff.net.2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.norm1.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.1.norm2.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.10.adaln_proj.linear.bias": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.10.adaln_proj.linear.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.10.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.10.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.10.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.10.attn.to_out.0.weight": "diffusion_pytorch_model-00001-of-00014.safetensors",
+    "transformer_blocks.10.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.10.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.10.ff.net.0.proj.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.10.ff.net.2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.10.norm1.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.10.norm2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.adaln_proj.linear.bias": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.adaln_proj.linear.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.attn.to_out.0.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.ff.net.0.proj.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.ff.net.2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.norm1.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.11.norm2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.adaln_proj.linear.bias": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.adaln_proj.linear.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.attn.to_out.0.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.ff.net.0.proj.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.ff.net.2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.norm1.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.12.norm2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.adaln_proj.linear.bias": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.adaln_proj.linear.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.attn.to_out.0.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.ff.net.0.proj.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.ff.net.2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.norm1.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.13.norm2.weight": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.14.adaln_proj.linear.bias": "diffusion_pytorch_model-00002-of-00014.safetensors",
+    "transformer_blocks.14.adaln_proj.linear.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.attn.norm_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.attn.norm_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.attn.to_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.attn.to_out.0.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.attn.to_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.attn.to_v.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.ff.net.0.proj.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.ff.net.2.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.norm1.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.14.norm2.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.adaln_proj.linear.bias": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.adaln_proj.linear.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.attn.norm_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.attn.norm_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.attn.to_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.attn.to_out.0.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.attn.to_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.attn.to_v.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.ff.net.0.proj.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.ff.net.2.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.norm1.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.15.norm2.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.adaln_proj.linear.bias": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.adaln_proj.linear.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.attn.norm_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.attn.norm_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.attn.to_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.attn.to_out.0.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.attn.to_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.attn.to_v.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.ff.net.0.proj.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.ff.net.2.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.norm1.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.16.norm2.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.adaln_proj.linear.bias": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.adaln_proj.linear.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.attn.norm_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.attn.norm_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.attn.to_k.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.attn.to_out.0.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.attn.to_q.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.attn.to_v.weight": "diffusion_pytorch_model-00003-of-00014.safetensors",
+    "transformer_blocks.17.ff.net.0.proj.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.17.ff.net.2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.17.norm1.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.17.norm2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.adaln_proj.linear.bias": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.adaln_proj.linear.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.attn.norm_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.attn.norm_q.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.attn.to_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.attn.to_out.0.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.attn.to_q.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.attn.to_v.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.ff.net.0.proj.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.ff.net.2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.norm1.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.18.norm2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.adaln_proj.linear.bias": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.adaln_proj.linear.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.attn.norm_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.attn.norm_q.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.attn.to_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.attn.to_out.0.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.attn.to_q.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.attn.to_v.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.ff.net.0.proj.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.ff.net.2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.norm1.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.19.norm2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.adaln_proj.linear.bias": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.adaln_proj.linear.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.attn.norm_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.attn.norm_q.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.attn.to_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.attn.to_out.0.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.attn.to_q.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.attn.to_v.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.ff.net.0.proj.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.ff.net.2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.norm1.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.2.norm2.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.20.adaln_proj.linear.bias": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.20.adaln_proj.linear.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.20.attn.norm_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.20.attn.norm_q.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.20.attn.to_k.weight": "diffusion_pytorch_model-00004-of-00014.safetensors",
+    "transformer_blocks.20.attn.to_out.0.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.20.attn.to_q.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.20.attn.to_v.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.20.ff.net.0.proj.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.20.ff.net.2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.20.norm1.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.20.norm2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.adaln_proj.linear.bias": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.adaln_proj.linear.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.attn.norm_k.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.attn.norm_q.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.attn.to_k.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.attn.to_out.0.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.attn.to_q.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.attn.to_v.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.ff.net.0.proj.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.ff.net.2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.norm1.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.21.norm2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.adaln_proj.linear.bias": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.adaln_proj.linear.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.attn.norm_k.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.attn.norm_q.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.attn.to_k.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.attn.to_out.0.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.attn.to_q.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.attn.to_v.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.ff.net.0.proj.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.ff.net.2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.norm1.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.22.norm2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.adaln_proj.linear.bias": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.adaln_proj.linear.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.attn.norm_k.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.attn.norm_q.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.attn.to_k.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.attn.to_out.0.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.attn.to_q.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.attn.to_v.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.ff.net.0.proj.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.ff.net.2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.norm1.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.23.norm2.weight": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.24.adaln_proj.linear.bias": "diffusion_pytorch_model-00005-of-00014.safetensors",
+    "transformer_blocks.24.adaln_proj.linear.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.attn.norm_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.attn.norm_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.attn.to_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.attn.to_out.0.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.attn.to_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.attn.to_v.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.ff.net.0.proj.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.ff.net.2.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.norm1.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.24.norm2.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.adaln_proj.linear.bias": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.adaln_proj.linear.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.attn.norm_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.attn.norm_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.attn.to_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.attn.to_out.0.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.attn.to_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.attn.to_v.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.ff.net.0.proj.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.ff.net.2.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.norm1.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.25.norm2.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.adaln_proj.linear.bias": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.adaln_proj.linear.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.attn.norm_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.attn.norm_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.attn.to_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.attn.to_out.0.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.attn.to_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.attn.to_v.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.ff.net.0.proj.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.ff.net.2.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.norm1.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.26.norm2.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.adaln_proj.linear.bias": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.adaln_proj.linear.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.attn.norm_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.attn.norm_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.attn.to_k.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.attn.to_out.0.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.attn.to_q.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.attn.to_v.weight": "diffusion_pytorch_model-00006-of-00014.safetensors",
+    "transformer_blocks.27.ff.net.0.proj.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.27.ff.net.2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.27.norm1.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.27.norm2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.adaln_proj.linear.bias": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.adaln_proj.linear.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.attn.norm_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.attn.norm_q.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.attn.to_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.attn.to_out.0.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.attn.to_q.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.attn.to_v.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.ff.net.0.proj.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.ff.net.2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.norm1.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.28.norm2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.adaln_proj.linear.bias": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.adaln_proj.linear.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.attn.norm_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.attn.norm_q.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.attn.to_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.attn.to_out.0.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.attn.to_q.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.attn.to_v.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.ff.net.0.proj.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.ff.net.2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.norm1.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.29.norm2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.adaln_proj.linear.bias": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.adaln_proj.linear.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.attn.norm_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.attn.norm_q.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.attn.to_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.attn.to_out.0.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.attn.to_q.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.attn.to_v.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.ff.net.0.proj.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.ff.net.2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.norm1.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.3.norm2.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.30.adaln_proj.linear.bias": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.30.adaln_proj.linear.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.30.attn.norm_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.30.attn.norm_q.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.30.attn.to_k.weight": "diffusion_pytorch_model-00007-of-00014.safetensors",
+    "transformer_blocks.30.attn.to_out.0.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.30.attn.to_q.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.30.attn.to_v.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.30.ff.net.0.proj.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.30.ff.net.2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.30.norm1.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.30.norm2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.adaln_proj.linear.bias": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.adaln_proj.linear.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.attn.norm_k.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.attn.norm_q.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.attn.to_k.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.attn.to_out.0.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.attn.to_q.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.attn.to_v.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.ff.net.0.proj.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.ff.net.2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.norm1.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.31.norm2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.adaln_proj.linear.bias": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.adaln_proj.linear.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.attn.norm_k.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.attn.norm_q.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.attn.to_k.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.attn.to_out.0.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.attn.to_q.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.attn.to_v.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.ff.net.0.proj.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.ff.net.2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.norm1.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.32.norm2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.adaln_proj.linear.bias": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.adaln_proj.linear.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.attn.norm_k.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.attn.norm_q.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.attn.to_k.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.attn.to_out.0.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.attn.to_q.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.attn.to_v.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.ff.net.0.proj.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.ff.net.2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.norm1.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.33.norm2.weight": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.34.adaln_proj.linear.bias": "diffusion_pytorch_model-00008-of-00014.safetensors",
+    "transformer_blocks.34.adaln_proj.linear.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.attn.norm_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.attn.norm_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.attn.to_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.attn.to_out.0.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.attn.to_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.attn.to_v.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.ff.net.0.proj.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.ff.net.2.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.norm1.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.34.norm2.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.adaln_proj.linear.bias": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.adaln_proj.linear.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.attn.norm_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.attn.norm_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.attn.to_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.attn.to_out.0.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.attn.to_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.attn.to_v.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.ff.net.0.proj.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.ff.net.2.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.norm1.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.35.norm2.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.adaln_proj.linear.bias": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.adaln_proj.linear.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.attn.norm_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.attn.norm_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.attn.to_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.attn.to_out.0.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.attn.to_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.attn.to_v.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.ff.net.0.proj.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.ff.net.2.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.norm1.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.36.norm2.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.adaln_proj.linear.bias": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.adaln_proj.linear.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.attn.norm_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.attn.norm_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.attn.to_k.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.attn.to_out.0.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.attn.to_q.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.attn.to_v.weight": "diffusion_pytorch_model-00009-of-00014.safetensors",
+    "transformer_blocks.37.ff.net.0.proj.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.37.ff.net.2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.37.norm1.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.37.norm2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.adaln_proj.linear.bias": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.adaln_proj.linear.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.attn.norm_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.attn.norm_q.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.attn.to_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.attn.to_out.0.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.attn.to_q.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.attn.to_v.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.ff.net.0.proj.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.ff.net.2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.norm1.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.38.norm2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.adaln_proj.linear.bias": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.adaln_proj.linear.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.attn.norm_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.attn.norm_q.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.attn.to_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.attn.to_out.0.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.attn.to_q.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.attn.to_v.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.ff.net.0.proj.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.ff.net.2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.norm1.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.39.norm2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.adaln_proj.linear.bias": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.adaln_proj.linear.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.attn.norm_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.attn.norm_q.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.attn.to_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.attn.to_out.0.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.attn.to_q.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.attn.to_v.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.ff.net.0.proj.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.ff.net.2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.norm1.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.4.norm2.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.40.adaln_proj.linear.bias": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.40.adaln_proj.linear.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.40.attn.norm_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.40.attn.norm_q.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.40.attn.to_k.weight": "diffusion_pytorch_model-00010-of-00014.safetensors",
+    "transformer_blocks.40.attn.to_out.0.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.40.attn.to_q.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.40.attn.to_v.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.40.ff.net.0.proj.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.40.ff.net.2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.40.norm1.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.40.norm2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.adaln_proj.linear.bias": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.adaln_proj.linear.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.attn.norm_k.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.attn.norm_q.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.attn.to_k.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.attn.to_out.0.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.attn.to_q.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.attn.to_v.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.ff.net.0.proj.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.ff.net.2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.norm1.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.41.norm2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.adaln_proj.linear.bias": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.adaln_proj.linear.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.attn.norm_k.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.attn.norm_q.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.attn.to_k.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.attn.to_out.0.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.attn.to_q.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.attn.to_v.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.ff.net.0.proj.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.ff.net.2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.norm1.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.42.norm2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.adaln_proj.linear.bias": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.adaln_proj.linear.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.attn.norm_k.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.attn.norm_q.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.attn.to_k.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.attn.to_out.0.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.attn.to_q.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.attn.to_v.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.ff.net.0.proj.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.ff.net.2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.norm1.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.43.norm2.weight": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.44.adaln_proj.linear.bias": "diffusion_pytorch_model-00011-of-00014.safetensors",
+    "transformer_blocks.44.adaln_proj.linear.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.attn.norm_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.attn.norm_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.attn.to_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.attn.to_out.0.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.attn.to_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.attn.to_v.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.ff.net.0.proj.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.ff.net.2.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.norm1.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.44.norm2.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.adaln_proj.linear.bias": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.adaln_proj.linear.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.attn.norm_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.attn.norm_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.attn.to_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.attn.to_out.0.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.attn.to_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.attn.to_v.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.ff.net.0.proj.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.ff.net.2.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.norm1.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.45.norm2.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.adaln_proj.linear.bias": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.adaln_proj.linear.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.attn.norm_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.attn.norm_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.attn.to_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.attn.to_out.0.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.attn.to_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.attn.to_v.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.ff.net.0.proj.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.ff.net.2.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.norm1.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.46.norm2.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.adaln_proj.linear.bias": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.adaln_proj.linear.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.attn.norm_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.attn.norm_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.attn.to_k.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.attn.to_out.0.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.attn.to_q.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.attn.to_v.weight": "diffusion_pytorch_model-00012-of-00014.safetensors",
+    "transformer_blocks.47.ff.net.0.proj.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.47.ff.net.2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.47.norm1.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.47.norm2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.adaln_proj.linear.bias": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.adaln_proj.linear.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.attn.norm_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.attn.norm_q.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.attn.to_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.attn.to_out.0.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.attn.to_q.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.attn.to_v.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.ff.net.0.proj.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.ff.net.2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.norm1.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.48.norm2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.adaln_proj.linear.bias": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.adaln_proj.linear.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.attn.norm_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.attn.norm_q.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.attn.to_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.attn.to_out.0.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.attn.to_q.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.attn.to_v.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.ff.net.0.proj.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.ff.net.2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.norm1.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.49.norm2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.adaln_proj.linear.bias": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.adaln_proj.linear.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.attn.norm_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.attn.norm_q.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.attn.to_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.attn.to_out.0.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.attn.to_q.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.attn.to_v.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.ff.net.0.proj.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.ff.net.2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.norm1.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.5.norm2.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.6.adaln_proj.linear.bias": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.6.adaln_proj.linear.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.6.attn.norm_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.6.attn.norm_q.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.6.attn.to_k.weight": "diffusion_pytorch_model-00013-of-00014.safetensors",
+    "transformer_blocks.6.attn.to_out.0.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.6.attn.to_q.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.6.attn.to_v.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.6.ff.net.0.proj.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.6.ff.net.2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.6.norm1.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.6.norm2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.adaln_proj.linear.bias": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.adaln_proj.linear.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.attn.norm_k.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.attn.norm_q.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.attn.to_k.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.attn.to_out.0.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.attn.to_q.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.attn.to_v.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.ff.net.0.proj.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.ff.net.2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.norm1.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.7.norm2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.adaln_proj.linear.bias": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.adaln_proj.linear.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.attn.norm_k.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.attn.norm_q.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.attn.to_k.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.attn.to_out.0.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.attn.to_q.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.attn.to_v.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.ff.net.0.proj.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.ff.net.2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.norm1.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.8.norm2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.adaln_proj.linear.bias": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.adaln_proj.linear.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.attn.norm_k.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.attn.norm_q.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.attn.to_k.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.attn.to_out.0.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.attn.to_q.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.attn.to_v.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.ff.net.0.proj.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.ff.net.2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.norm1.weight": "diffusion_pytorch_model-00014-of-00014.safetensors",
+    "transformer_blocks.9.norm2.weight": "diffusion_pytorch_model-00014-of-00014.safetensors"
+  }
+}