Upload 21 files
Browse files- .gitattributes +1 -0
- LICENSE +202 -0
- README.md +129 -0
- modular_model_index.json +74 -0
- scheduler/scheduler_config.json +18 -0
- text_encoder/config.json +83 -0
- text_encoder/model-00001-of-00004.safetensors +3 -0
- text_encoder/model-00002-of-00004.safetensors +3 -0
- text_encoder/model-00003-of-00004.safetensors +3 -0
- text_encoder/model-00004-of-00004.safetensors +3 -0
- text_encoder/model.safetensors.index.json +434 -0
- tokenizer/chat_template.jinja +154 -0
- tokenizer/tokenizer.json +3 -0
- tokenizer/tokenizer_config.json +32 -0
- transformer/config.json +13 -0
- transformer/diffusion_pytorch_model-00001-of-00003.safetensors +3 -0
- transformer/diffusion_pytorch_model-00002-of-00003.safetensors +3 -0
- transformer/diffusion_pytorch_model-00003-of-00003.safetensors +3 -0
- transformer/diffusion_pytorch_model.safetensors.index.json +469 -0
- transformer/transformer_moss_image.py +333 -0
- vae/config.json +45 -0
- vae/diffusion_pytorch_model.safetensors +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
LICENSE
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
Apache License
|
| 3 |
+
Version 2.0, January 2004
|
| 4 |
+
http://www.apache.org/licenses/
|
| 5 |
+
|
| 6 |
+
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
| 7 |
+
|
| 8 |
+
1. Definitions.
|
| 9 |
+
|
| 10 |
+
"License" shall mean the terms and conditions for use, reproduction,
|
| 11 |
+
and distribution as defined by Sections 1 through 9 of this document.
|
| 12 |
+
|
| 13 |
+
"Licensor" shall mean the copyright owner or entity authorized by
|
| 14 |
+
the copyright owner that is granting the License.
|
| 15 |
+
|
| 16 |
+
"Legal Entity" shall mean the union of the acting entity and all
|
| 17 |
+
other entities that control, are controlled by, or are under common
|
| 18 |
+
control with that entity. For the purposes of this definition,
|
| 19 |
+
"control" means (i) the power, direct or indirect, to cause the
|
| 20 |
+
direction or management of such entity, whether by contract or
|
| 21 |
+
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
| 22 |
+
outstanding shares, or (iii) beneficial ownership of such entity.
|
| 23 |
+
|
| 24 |
+
"You" (or "Your") shall mean an individual or Legal Entity
|
| 25 |
+
exercising permissions granted by this License.
|
| 26 |
+
|
| 27 |
+
"Source" form shall mean the preferred form for making modifications,
|
| 28 |
+
including but not limited to software source code, documentation
|
| 29 |
+
source, and configuration files.
|
| 30 |
+
|
| 31 |
+
"Object" form shall mean any form resulting from mechanical
|
| 32 |
+
transformation or translation of a Source form, including but
|
| 33 |
+
not limited to compiled object code, generated documentation,
|
| 34 |
+
and conversions to other media types.
|
| 35 |
+
|
| 36 |
+
"Work" shall mean the work of authorship, whether in Source or
|
| 37 |
+
Object form, made available under the License, as indicated by a
|
| 38 |
+
copyright notice that is included in or attached to the work
|
| 39 |
+
(an example is provided in the Appendix below).
|
| 40 |
+
|
| 41 |
+
"Derivative Works" shall mean any work, whether in Source or Object
|
| 42 |
+
form, that is based on (or derived from) the Work and for which the
|
| 43 |
+
editorial revisions, annotations, elaborations, or other modifications
|
| 44 |
+
represent, as a whole, an original work of authorship. For the purposes
|
| 45 |
+
of this License, Derivative Works shall not include works that remain
|
| 46 |
+
separable from, or merely link (or bind by name) to the interfaces of,
|
| 47 |
+
the Work and Derivative Works thereof.
|
| 48 |
+
|
| 49 |
+
"Contribution" shall mean any work of authorship, including
|
| 50 |
+
the original version of the Work and any modifications or additions
|
| 51 |
+
to that Work or Derivative Works thereof, that is intentionally
|
| 52 |
+
submitted to Licensor for inclusion in the Work by the copyright owner
|
| 53 |
+
or by an individual or Legal Entity authorized to submit on behalf of
|
| 54 |
+
the copyright owner. For the purposes of this definition, "submitted"
|
| 55 |
+
means any form of electronic, verbal, or written communication sent
|
| 56 |
+
to the Licensor or its representatives, including but not limited to
|
| 57 |
+
communication on electronic mailing lists, source code control systems,
|
| 58 |
+
and issue tracking systems that are managed by, or on behalf of, the
|
| 59 |
+
Licensor for the purpose of discussing and improving the Work, but
|
| 60 |
+
excluding communication that is conspicuously marked or otherwise
|
| 61 |
+
designated in writing by the copyright owner as "Not a Contribution."
|
| 62 |
+
|
| 63 |
+
"Contributor" shall mean Licensor and any individual or Legal Entity
|
| 64 |
+
on behalf of whom a Contribution has been received by Licensor and
|
| 65 |
+
subsequently incorporated within the Work.
|
| 66 |
+
|
| 67 |
+
2. Grant of Copyright License. Subject to the terms and conditions of
|
| 68 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 69 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 70 |
+
copyright license to reproduce, prepare Derivative Works of,
|
| 71 |
+
publicly display, publicly perform, sublicense, and distribute the
|
| 72 |
+
Work and such Derivative Works in Source or Object form.
|
| 73 |
+
|
| 74 |
+
3. Grant of Patent License. Subject to the terms and conditions of
|
| 75 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 76 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 77 |
+
(except as stated in this section) patent license to make, have made,
|
| 78 |
+
use, offer to sell, sell, import, and otherwise transfer the Work,
|
| 79 |
+
where such license applies only to those patent claims licensable
|
| 80 |
+
by such Contributor that are necessarily infringed by their
|
| 81 |
+
Contribution(s) alone or by combination of their Contribution(s)
|
| 82 |
+
with the Work to which such Contribution(s) was submitted. If You
|
| 83 |
+
institute patent litigation against any entity (including a
|
| 84 |
+
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
| 85 |
+
or a Contribution incorporated within the Work constitutes direct
|
| 86 |
+
or contributory patent infringement, then any patent licenses
|
| 87 |
+
granted to You under this License for that Work shall terminate
|
| 88 |
+
as of the date such litigation is filed.
|
| 89 |
+
|
| 90 |
+
4. Redistribution. You may reproduce and distribute copies of the
|
| 91 |
+
Work or Derivative Works thereof in any medium, with or without
|
| 92 |
+
modifications, and in Source or Object form, provided that You
|
| 93 |
+
meet the following conditions:
|
| 94 |
+
|
| 95 |
+
(a) You must give any other recipients of the Work or
|
| 96 |
+
Derivative Works a copy of this License; and
|
| 97 |
+
|
| 98 |
+
(b) You must cause any modified files to carry prominent notices
|
| 99 |
+
stating that You changed the files; and
|
| 100 |
+
|
| 101 |
+
(c) You must retain, in the Source form of any Derivative Works
|
| 102 |
+
that You distribute, all copyright, patent, trademark, and
|
| 103 |
+
attribution notices from the Source form of the Work,
|
| 104 |
+
excluding those notices that do not pertain to any part of
|
| 105 |
+
the Derivative Works; and
|
| 106 |
+
|
| 107 |
+
(d) If the Work includes a "NOTICE" text file as part of its
|
| 108 |
+
distribution, then any Derivative Works that You distribute must
|
| 109 |
+
include a readable copy of the attribution notices contained
|
| 110 |
+
within such NOTICE file, excluding those notices that do not
|
| 111 |
+
pertain to any part of the Derivative Works, in at least one
|
| 112 |
+
of the following places: within a NOTICE text file distributed
|
| 113 |
+
as part of the Derivative Works; within the Source form or
|
| 114 |
+
documentation, if provided along with the Derivative Works; or,
|
| 115 |
+
within a display generated by the Derivative Works, if and
|
| 116 |
+
wherever such third-party notices normally appear. The contents
|
| 117 |
+
of the NOTICE file are for informational purposes only and
|
| 118 |
+
do not modify the License. You may add Your own attribution
|
| 119 |
+
notices within Derivative Works that You distribute, alongside
|
| 120 |
+
or as an addendum to the NOTICE text from the Work, provided
|
| 121 |
+
that such additional attribution notices cannot be construed
|
| 122 |
+
as modifying the License.
|
| 123 |
+
|
| 124 |
+
You may add Your own copyright statement to Your modifications and
|
| 125 |
+
may provide additional or different license terms and conditions
|
| 126 |
+
for use, reproduction, or distribution of Your modifications, or
|
| 127 |
+
for any such Derivative Works as a whole, provided Your use,
|
| 128 |
+
reproduction, and distribution of the Work otherwise complies with
|
| 129 |
+
the conditions stated in this License.
|
| 130 |
+
|
| 131 |
+
5. Submission of Contributions. Unless You explicitly state otherwise,
|
| 132 |
+
any Contribution intentionally submitted for inclusion in the Work
|
| 133 |
+
by You to the Licensor shall be under the terms and conditions of
|
| 134 |
+
this License, without any additional terms or conditions.
|
| 135 |
+
Notwithstanding the above, nothing herein shall supersede or modify
|
| 136 |
+
the terms of any separate license agreement you may have executed
|
| 137 |
+
with Licensor regarding such Contributions.
|
| 138 |
+
|
| 139 |
+
6. Trademarks. This License does not grant permission to use the trade
|
| 140 |
+
names, trademarks, service marks, or product names of the Licensor,
|
| 141 |
+
except as required for reasonable and customary use in describing the
|
| 142 |
+
origin of the Work and reproducing the content of the NOTICE file.
|
| 143 |
+
|
| 144 |
+
7. Disclaimer of Warranty. Unless required by applicable law or
|
| 145 |
+
agreed to in writing, Licensor provides the Work (and each
|
| 146 |
+
Contributor provides its Contributions) on an "AS IS" BASIS,
|
| 147 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
| 148 |
+
implied, including, without limitation, any warranties or conditions
|
| 149 |
+
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
| 150 |
+
PARTICULAR PURPOSE. You are solely responsible for determining the
|
| 151 |
+
appropriateness of using or redistributing the Work and assume any
|
| 152 |
+
risks associated with Your exercise of permissions under this License.
|
| 153 |
+
|
| 154 |
+
8. Limitation of Liability. In no event and under no legal theory,
|
| 155 |
+
whether in tort (including negligence), contract, or otherwise,
|
| 156 |
+
unless required by applicable law (such as deliberate and grossly
|
| 157 |
+
negligent acts) or agreed to in writing, shall any Contributor be
|
| 158 |
+
liable to You for damages, including any direct, indirect, special,
|
| 159 |
+
incidental, or consequential damages of any character arising as a
|
| 160 |
+
result of this License or out of the use or inability to use the
|
| 161 |
+
Work (including but not limited to damages for loss of goodwill,
|
| 162 |
+
work stoppage, computer failure or malfunction, or any and all
|
| 163 |
+
other commercial damages or losses), even if such Contributor
|
| 164 |
+
has been advised of the possibility of such damages.
|
| 165 |
+
|
| 166 |
+
9. Accepting Warranty or Additional Liability. While redistributing
|
| 167 |
+
the Work or Derivative Works thereof, You may choose to offer,
|
| 168 |
+
and charge a fee for, acceptance of support, warranty, indemnity,
|
| 169 |
+
or other liability obligations and/or rights consistent with this
|
| 170 |
+
License. However, in accepting such obligations, You may act only
|
| 171 |
+
on Your own behalf and on Your sole responsibility, not on behalf
|
| 172 |
+
of any other Contributor, and only if You agree to indemnify,
|
| 173 |
+
defend, and hold each Contributor harmless for any liability
|
| 174 |
+
incurred by, or claims asserted against, such Contributor by reason
|
| 175 |
+
of your accepting any such warranty or additional liability.
|
| 176 |
+
|
| 177 |
+
END OF TERMS AND CONDITIONS
|
| 178 |
+
|
| 179 |
+
APPENDIX: How to apply the Apache License to your work.
|
| 180 |
+
|
| 181 |
+
To apply the Apache License to your work, attach the following
|
| 182 |
+
boilerplate notice, with the fields enclosed by brackets "[]"
|
| 183 |
+
replaced with your own identifying information. (Don't include
|
| 184 |
+
the brackets!) The text should be enclosed in the appropriate
|
| 185 |
+
comment syntax for the file format. We also recommend that a
|
| 186 |
+
file or class name and description of purpose be included on the
|
| 187 |
+
same "printed page" as the copyright notice for easier
|
| 188 |
+
identification within third-party archives.
|
| 189 |
+
|
| 190 |
+
Copyright [yyyy] [name of copyright owner]
|
| 191 |
+
|
| 192 |
+
Licensed under the Apache License, Version 2.0 (the "License");
|
| 193 |
+
you may not use this file except in compliance with the License.
|
| 194 |
+
You may obtain a copy of the License at
|
| 195 |
+
|
| 196 |
+
http://www.apache.org/licenses/LICENSE-2.0
|
| 197 |
+
|
| 198 |
+
Unless required by applicable law or agreed to in writing, software
|
| 199 |
+
distributed under the License is distributed on an "AS IS" BASIS,
|
| 200 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
| 201 |
+
See the License for the specific language governing permissions and
|
| 202 |
+
limitations under the License.
|
README.md
ADDED
|
@@ -0,0 +1,129 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: apache-2.0
|
| 5 |
+
base_model:
|
| 6 |
+
- OpenMOSS-Team/MOSS-Image-SFT
|
| 7 |
+
pipeline_tag: text-to-image
|
| 8 |
+
library_name: diffusers
|
| 9 |
+
tags:
|
| 10 |
+
- image-generation
|
| 11 |
+
- moss-image
|
| 12 |
+
- modular-diffusers
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
# MOSS-Image SFT
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
This is [OpenMOSS-Team/MOSS-Image-SFT](https://huggingface.co/OpenMOSS-Team/MOSS-Image-SFT) converted to the standard diffusers layout, so it runs as a Modular Diffusers pipeline with the custom blocks below instead of the custom pipeline shipped in the original repository. The weights are the original bf16 ones.
|
| 20 |
+
|
| 21 |
+
Note: You need diffusers from `main` (0.41.0.dev0 or later).
|
| 22 |
+
|
| 23 |
+
## Usage
|
| 24 |
+
|
| 25 |
+
You can find ready-to-use scripts in the [diffusers-recipes](https://github.com/asomoza/diffusers-recipes/blob/main/models/moss_image/README.md) repository.
|
| 26 |
+
|
| 27 |
+
MOSS-Image runs as a Modular Diffusers pipeline built from the [MOSS-Image custom blocks](https://huggingface.co/OzzyGT/moss_image_custom_blocks), which load the transformer code shipped in this repository (hence `trust_remote_code` and `trust_components_code`). The VAE runs in float32, like the original.
|
| 28 |
+
|
| 29 |
+
**Note:** both flags run Python code downloaded from the Hub: `trust_remote_code=True` runs the code of the blocks repository, and `trust_components_code=True` runs the code shipped in the `components_repo`. Only use them with repositories you trust, and review their code before passing different ones.
|
| 30 |
+
|
| 31 |
+
With every component on the GPU, generating at 2048x1152 needs more than 32 GB of VRAM. For smaller GPUs, see the offloading recipes in diffusers-recipes or use one of the [SDNQ quantized versions](https://huggingface.co/OzzyGT/MOSS_Image_SFT_sdnq_dynamic_8bit).
|
| 32 |
+
|
| 33 |
+
For prompting, recommended settings and the layout format, see the [original model card](https://huggingface.co/OpenMOSS-Team/MOSS-Image-SFT) and the [user guide](https://github.com/OpenMOSS/MOSS-Image/blob/main/USERGUIDE.md#layout-prompts-experimental). The sample below uses a layout prompt: boxes are `[y_min, x_min, y_max, x_max]` in pixels with the origin at the center of the canvas.
|
| 34 |
+
|
| 35 |
+
Note: the samples use 28 steps, which is also the default of the custom blocks, because it already gives good results in about half the time. The official recommendation is 50 steps or more, so raise `num_inference_steps` if you want the best quality.
|
| 36 |
+
|
| 37 |
+
## Sample image
|
| 38 |
+
|
| 39 |
+
The image above was generated with the following prompt and settings (seed `7`):
|
| 40 |
+
|
| 41 |
+
```python
|
| 42 |
+
import json
|
| 43 |
+
|
| 44 |
+
import torch
|
| 45 |
+
from diffusers import ModularPipelineBlocks
|
| 46 |
+
|
| 47 |
+
|
| 48 |
+
blocks = ModularPipelineBlocks.from_pretrained(
|
| 49 |
+
"OzzyGT/moss_image_custom_blocks",
|
| 50 |
+
trust_remote_code=True,
|
| 51 |
+
components_repo="OzzyGT/MOSS_Image_SFT_diffusers",
|
| 52 |
+
trust_components_code=True,
|
| 53 |
+
)
|
| 54 |
+
pipe = blocks.init_pipeline()
|
| 55 |
+
pipe.load_components(dtype={"default": torch.bfloat16, "vae": torch.float32})
|
| 56 |
+
pipe.to("cuda")
|
| 57 |
+
|
| 58 |
+
layout = {
|
| 59 |
+
"summary": {
|
| 60 |
+
"meta": ["photo"],
|
| 61 |
+
"canvas": "single",
|
| 62 |
+
"caption": (
|
| 63 |
+
"A photograph of the inside of a small bicycle repair workshop, shot on a Sony A7 IV with a 35mm f/1.8 "
|
| 64 |
+
"lens at f/2.8, 1/125 s, ISO 1600, handheld at eye level, wide 16:9 frame. Low late-afternoon sunlight "
|
| 65 |
+
"comes through a large front window on the right in hard shafts with dust in the air, the rest of the "
|
| 66 |
+
"room is lit by bare fluorescent tubes on the ceiling. Shallow depth of field with the back wall slightly "
|
| 67 |
+
"out of focus, visible sensor grain. A red road bicycle is clamped in a repair stand in the center while "
|
| 68 |
+
"an adult mechanic in a grease-stained navy work apron adjusts its rear derailleur. A painted wooden sign "
|
| 69 |
+
'on the back wall reads "SPOKE & CHAIN". A pegboard of hand tools hangs on the left wall above an A-frame '
|
| 70 |
+
"chalkboard with a price list, and an adult black cat sleeps curled up on a metal stool by the window. "
|
| 71 |
+
"Oil stains on a grey concrete floor, spare tires hanging from hooks, scuffed white walls."
|
| 72 |
+
),
|
| 73 |
+
},
|
| 74 |
+
"items": [
|
| 75 |
+
{
|
| 76 |
+
"label": "text",
|
| 77 |
+
"caption": "A long painted wooden sign on the back wall, dark green with cream serif capital letters.",
|
| 78 |
+
"text": "SPOKE & CHAIN",
|
| 79 |
+
"bbox": [-496, -464, -346, 236],
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"label": "object",
|
| 83 |
+
"caption": "A perforated pegboard with wrenches, screwdrivers, tire levers and a chain whip on hooks.",
|
| 84 |
+
"bbox": [-396, -984, 124, -544],
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"label": "object",
|
| 88 |
+
"caption": "A red steel road bicycle with drop handlebars, clamped by its seat post in a grey repair stand.",
|
| 89 |
+
"bbox": [-196, -504, 329, 371],
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"label": "object",
|
| 93 |
+
"caption": (
|
| 94 |
+
"An adult man in his forties with short grey hair, a grey t-shirt and a grease-stained navy work "
|
| 95 |
+
"apron, standing and turning a small screwdriver on the rear derailleur."
|
| 96 |
+
),
|
| 97 |
+
"bbox": [-371, 396, 504, 736],
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"label": "text",
|
| 101 |
+
"caption": "A black A-frame chalkboard on the floor with a handwritten price list in white chalk.",
|
| 102 |
+
"text": "FLAT REPAIR $12\nFULL TUNE-UP $45\nOPEN 8-6",
|
| 103 |
+
"bbox": [184, -964, 524, -604],
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"label": "object",
|
| 107 |
+
"caption": "An adult black cat sleeping curled up on the seat of a metal shop stool in a patch of sunlight.",
|
| 108 |
+
"bbox": [104, 766, 209, 981],
|
| 109 |
+
},
|
| 110 |
+
],
|
| 111 |
+
"bbox_coord_space": "abs_centralized",
|
| 112 |
+
"bbox_format": "yxyx",
|
| 113 |
+
}
|
| 114 |
+
prompt = json.dumps(layout, ensure_ascii=False)
|
| 115 |
+
|
| 116 |
+
image = pipe(
|
| 117 |
+
prompt=prompt,
|
| 118 |
+
num_inference_steps=28,
|
| 119 |
+
height=1152,
|
| 120 |
+
width=2048,
|
| 121 |
+
generator=torch.Generator("cuda").manual_seed(7),
|
| 122 |
+
output="images",
|
| 123 |
+
)[0]
|
| 124 |
+
image.save("moss_image.png")
|
| 125 |
+
```
|
| 126 |
+
|
| 127 |
+
## License
|
| 128 |
+
|
| 129 |
+
This model is released under the [Apache-2.0 license](LICENSE), the same as the original.
|
modular_model_index.json
ADDED
|
@@ -0,0 +1,74 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "ModularPipeline",
|
| 3 |
+
"_blocks_class_name": "MOSSImageBlocks",
|
| 4 |
+
"scheduler": [
|
| 5 |
+
"diffusers",
|
| 6 |
+
"FlowMatchEulerDiscreteScheduler",
|
| 7 |
+
{
|
| 8 |
+
"pretrained_model_name_or_path": null,
|
| 9 |
+
"revision": null,
|
| 10 |
+
"subfolder": "scheduler",
|
| 11 |
+
"type_hint": [
|
| 12 |
+
"diffusers",
|
| 13 |
+
"FlowMatchEulerDiscreteScheduler"
|
| 14 |
+
],
|
| 15 |
+
"variant": null
|
| 16 |
+
}
|
| 17 |
+
],
|
| 18 |
+
"text_encoder": [
|
| 19 |
+
"transformers",
|
| 20 |
+
"Qwen3_5TextModel",
|
| 21 |
+
{
|
| 22 |
+
"pretrained_model_name_or_path": null,
|
| 23 |
+
"revision": null,
|
| 24 |
+
"subfolder": "text_encoder",
|
| 25 |
+
"type_hint": [
|
| 26 |
+
"transformers",
|
| 27 |
+
"Qwen3_5TextModel"
|
| 28 |
+
],
|
| 29 |
+
"variant": null
|
| 30 |
+
}
|
| 31 |
+
],
|
| 32 |
+
"tokenizer": [
|
| 33 |
+
"transformers",
|
| 34 |
+
"Qwen2Tokenizer",
|
| 35 |
+
{
|
| 36 |
+
"pretrained_model_name_or_path": null,
|
| 37 |
+
"revision": null,
|
| 38 |
+
"subfolder": "tokenizer",
|
| 39 |
+
"type_hint": [
|
| 40 |
+
"transformers",
|
| 41 |
+
"Qwen2Tokenizer"
|
| 42 |
+
],
|
| 43 |
+
"variant": null
|
| 44 |
+
}
|
| 45 |
+
],
|
| 46 |
+
"transformer": [
|
| 47 |
+
"transformer_moss_image",
|
| 48 |
+
"MOSSImageTransformer2DModel",
|
| 49 |
+
{
|
| 50 |
+
"pretrained_model_name_or_path": null,
|
| 51 |
+
"revision": null,
|
| 52 |
+
"subfolder": "transformer",
|
| 53 |
+
"type_hint": [
|
| 54 |
+
"transformer_moss_image",
|
| 55 |
+
"MOSSImageTransformer2DModel"
|
| 56 |
+
],
|
| 57 |
+
"variant": null
|
| 58 |
+
}
|
| 59 |
+
],
|
| 60 |
+
"vae": [
|
| 61 |
+
"diffusers",
|
| 62 |
+
"AutoencoderKLFlux2",
|
| 63 |
+
{
|
| 64 |
+
"pretrained_model_name_or_path": null,
|
| 65 |
+
"revision": null,
|
| 66 |
+
"subfolder": "vae",
|
| 67 |
+
"type_hint": [
|
| 68 |
+
"diffusers",
|
| 69 |
+
"AutoencoderKLFlux2"
|
| 70 |
+
],
|
| 71 |
+
"variant": null
|
| 72 |
+
}
|
| 73 |
+
]
|
| 74 |
+
}
|
scheduler/scheduler_config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "FlowMatchEulerDiscreteScheduler",
|
| 3 |
+
"_diffusers_version": "0.41.0.dev0",
|
| 4 |
+
"base_image_seq_len": 256,
|
| 5 |
+
"base_shift": 0.5,
|
| 6 |
+
"invert_sigmas": false,
|
| 7 |
+
"max_image_seq_len": 4096,
|
| 8 |
+
"max_shift": 1.15,
|
| 9 |
+
"num_train_timesteps": 1000,
|
| 10 |
+
"shift": 5.0,
|
| 11 |
+
"shift_terminal": null,
|
| 12 |
+
"stochastic_sampling": false,
|
| 13 |
+
"time_shift_type": "exponential",
|
| 14 |
+
"use_beta_sigmas": false,
|
| 15 |
+
"use_dynamic_shifting": false,
|
| 16 |
+
"use_exponential_sigmas": false,
|
| 17 |
+
"use_karras_sigmas": false
|
| 18 |
+
}
|
text_encoder/config.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3_5TextModel"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"attn_output_gate": true,
|
| 8 |
+
"bos_token_id": null,
|
| 9 |
+
"dtype": "bfloat16",
|
| 10 |
+
"eos_token_id": 248044,
|
| 11 |
+
"full_attention_interval": 4,
|
| 12 |
+
"head_dim": 256,
|
| 13 |
+
"hidden_act": "silu",
|
| 14 |
+
"hidden_size": 4096,
|
| 15 |
+
"initializer_range": 0.02,
|
| 16 |
+
"intermediate_size": 12288,
|
| 17 |
+
"layer_types": [
|
| 18 |
+
"linear_attention",
|
| 19 |
+
"linear_attention",
|
| 20 |
+
"linear_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"linear_attention",
|
| 23 |
+
"linear_attention",
|
| 24 |
+
"linear_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"linear_attention",
|
| 27 |
+
"linear_attention",
|
| 28 |
+
"linear_attention",
|
| 29 |
+
"full_attention",
|
| 30 |
+
"linear_attention",
|
| 31 |
+
"linear_attention",
|
| 32 |
+
"linear_attention",
|
| 33 |
+
"full_attention",
|
| 34 |
+
"linear_attention",
|
| 35 |
+
"linear_attention",
|
| 36 |
+
"linear_attention",
|
| 37 |
+
"full_attention",
|
| 38 |
+
"linear_attention",
|
| 39 |
+
"linear_attention",
|
| 40 |
+
"linear_attention",
|
| 41 |
+
"full_attention",
|
| 42 |
+
"linear_attention",
|
| 43 |
+
"linear_attention",
|
| 44 |
+
"linear_attention",
|
| 45 |
+
"full_attention",
|
| 46 |
+
"linear_attention",
|
| 47 |
+
"linear_attention",
|
| 48 |
+
"linear_attention",
|
| 49 |
+
"full_attention"
|
| 50 |
+
],
|
| 51 |
+
"linear_conv_kernel_dim": 4,
|
| 52 |
+
"linear_key_head_dim": 128,
|
| 53 |
+
"linear_num_key_heads": 16,
|
| 54 |
+
"linear_num_value_heads": 32,
|
| 55 |
+
"linear_value_head_dim": 128,
|
| 56 |
+
"mamba_ssm_dtype": "float32",
|
| 57 |
+
"max_position_embeddings": 262144,
|
| 58 |
+
"mlp_only_layers": [],
|
| 59 |
+
"model_type": "qwen3_5_text",
|
| 60 |
+
"mtp_num_hidden_layers": 1,
|
| 61 |
+
"mtp_use_dedicated_embeddings": false,
|
| 62 |
+
"num_attention_heads": 16,
|
| 63 |
+
"num_hidden_layers": 32,
|
| 64 |
+
"num_key_value_heads": 4,
|
| 65 |
+
"pad_token_id": null,
|
| 66 |
+
"partial_rotary_factor": 0.25,
|
| 67 |
+
"rms_norm_eps": 1e-06,
|
| 68 |
+
"rope_parameters": {
|
| 69 |
+
"mrope_interleaved": true,
|
| 70 |
+
"mrope_section": [
|
| 71 |
+
11,
|
| 72 |
+
11,
|
| 73 |
+
10
|
| 74 |
+
],
|
| 75 |
+
"partial_rotary_factor": 0.25,
|
| 76 |
+
"rope_theta": 10000000,
|
| 77 |
+
"rope_type": "default"
|
| 78 |
+
},
|
| 79 |
+
"tie_word_embeddings": false,
|
| 80 |
+
"transformers_version": "5.14.1",
|
| 81 |
+
"use_cache": true,
|
| 82 |
+
"vocab_size": 248320
|
| 83 |
+
}
|
text_encoder/model-00001-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74f209d5299e87bbcc7de2c12e3a4b49cb56a06cd0bed6ae9a60fd5432dd6c67
|
| 3 |
+
size 4973901040
|
text_encoder/model-00002-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:365021e5a60c859a936e23b879695f4c7bce2cc008541afe173efdbde0500bab
|
| 3 |
+
size 4988361936
|
text_encoder/model-00003-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91a8bd6ee80a0f126925c278c4e01ef2ffa8410490a369c3d619e6d350757d0e
|
| 3 |
+
size 4954199896
|
text_encoder/model-00004-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45d65389f300e39d4f56840c480ad791d9c6c49d3fc044962d83785b388b7192
|
| 3 |
+
size 956953032
|
text_encoder/model.safetensors.index.json
ADDED
|
@@ -0,0 +1,434 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_parameters": 7936684544,
|
| 4 |
+
"total_size": 15873369088
|
| 5 |
+
},
|
| 6 |
+
"weight_map": {
|
| 7 |
+
"embed_tokens.weight": "model-00001-of-00004.safetensors",
|
| 8 |
+
"layers.0.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 9 |
+
"layers.0.linear_attn.A_log": "model-00001-of-00004.safetensors",
|
| 10 |
+
"layers.0.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
|
| 11 |
+
"layers.0.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
|
| 12 |
+
"layers.0.linear_attn.in_proj_a.weight": "model-00001-of-00004.safetensors",
|
| 13 |
+
"layers.0.linear_attn.in_proj_b.weight": "model-00001-of-00004.safetensors",
|
| 14 |
+
"layers.0.linear_attn.in_proj_qkv.weight": "model-00001-of-00004.safetensors",
|
| 15 |
+
"layers.0.linear_attn.in_proj_z.weight": "model-00001-of-00004.safetensors",
|
| 16 |
+
"layers.0.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
|
| 17 |
+
"layers.0.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
|
| 18 |
+
"layers.0.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 19 |
+
"layers.0.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 20 |
+
"layers.0.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 21 |
+
"layers.0.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 22 |
+
"layers.1.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 23 |
+
"layers.1.linear_attn.A_log": "model-00001-of-00004.safetensors",
|
| 24 |
+
"layers.1.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
|
| 25 |
+
"layers.1.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
|
| 26 |
+
"layers.1.linear_attn.in_proj_a.weight": "model-00001-of-00004.safetensors",
|
| 27 |
+
"layers.1.linear_attn.in_proj_b.weight": "model-00001-of-00004.safetensors",
|
| 28 |
+
"layers.1.linear_attn.in_proj_qkv.weight": "model-00001-of-00004.safetensors",
|
| 29 |
+
"layers.1.linear_attn.in_proj_z.weight": "model-00001-of-00004.safetensors",
|
| 30 |
+
"layers.1.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
|
| 31 |
+
"layers.1.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
|
| 32 |
+
"layers.1.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 33 |
+
"layers.1.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 34 |
+
"layers.1.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 35 |
+
"layers.1.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 36 |
+
"layers.10.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 37 |
+
"layers.10.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 38 |
+
"layers.10.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 39 |
+
"layers.10.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 40 |
+
"layers.10.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 41 |
+
"layers.10.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 42 |
+
"layers.10.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 43 |
+
"layers.10.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 44 |
+
"layers.10.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 45 |
+
"layers.10.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 46 |
+
"layers.10.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 47 |
+
"layers.10.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 48 |
+
"layers.10.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 49 |
+
"layers.10.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 50 |
+
"layers.11.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 51 |
+
"layers.11.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 52 |
+
"layers.11.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 53 |
+
"layers.11.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 54 |
+
"layers.11.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 55 |
+
"layers.11.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
|
| 56 |
+
"layers.11.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 57 |
+
"layers.11.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 58 |
+
"layers.11.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
|
| 59 |
+
"layers.11.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 60 |
+
"layers.11.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 61 |
+
"layers.12.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 62 |
+
"layers.12.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 63 |
+
"layers.12.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 64 |
+
"layers.12.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 65 |
+
"layers.12.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 66 |
+
"layers.12.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 67 |
+
"layers.12.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 68 |
+
"layers.12.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 69 |
+
"layers.12.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 70 |
+
"layers.12.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 71 |
+
"layers.12.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 72 |
+
"layers.12.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 73 |
+
"layers.12.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 74 |
+
"layers.12.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 75 |
+
"layers.13.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 76 |
+
"layers.13.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 77 |
+
"layers.13.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 78 |
+
"layers.13.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 79 |
+
"layers.13.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 80 |
+
"layers.13.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 81 |
+
"layers.13.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 82 |
+
"layers.13.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 83 |
+
"layers.13.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 84 |
+
"layers.13.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 85 |
+
"layers.13.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 86 |
+
"layers.13.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 87 |
+
"layers.13.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 88 |
+
"layers.13.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 89 |
+
"layers.14.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 90 |
+
"layers.14.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 91 |
+
"layers.14.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 92 |
+
"layers.14.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 93 |
+
"layers.14.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 94 |
+
"layers.14.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 95 |
+
"layers.14.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 96 |
+
"layers.14.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 97 |
+
"layers.14.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 98 |
+
"layers.14.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 99 |
+
"layers.14.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 100 |
+
"layers.14.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 101 |
+
"layers.14.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 102 |
+
"layers.14.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 103 |
+
"layers.15.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 104 |
+
"layers.15.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 105 |
+
"layers.15.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 106 |
+
"layers.15.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 107 |
+
"layers.15.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 108 |
+
"layers.15.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
|
| 109 |
+
"layers.15.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 110 |
+
"layers.15.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 111 |
+
"layers.15.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
|
| 112 |
+
"layers.15.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 113 |
+
"layers.15.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 114 |
+
"layers.16.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 115 |
+
"layers.16.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 116 |
+
"layers.16.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 117 |
+
"layers.16.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 118 |
+
"layers.16.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 119 |
+
"layers.16.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 120 |
+
"layers.16.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 121 |
+
"layers.16.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 122 |
+
"layers.16.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 123 |
+
"layers.16.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 124 |
+
"layers.16.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 125 |
+
"layers.16.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 126 |
+
"layers.16.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 127 |
+
"layers.16.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 128 |
+
"layers.17.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 129 |
+
"layers.17.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 130 |
+
"layers.17.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 131 |
+
"layers.17.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 132 |
+
"layers.17.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 133 |
+
"layers.17.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 134 |
+
"layers.17.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 135 |
+
"layers.17.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 136 |
+
"layers.17.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 137 |
+
"layers.17.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 138 |
+
"layers.17.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 139 |
+
"layers.17.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 140 |
+
"layers.17.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 141 |
+
"layers.17.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 142 |
+
"layers.18.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 143 |
+
"layers.18.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 144 |
+
"layers.18.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 145 |
+
"layers.18.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 146 |
+
"layers.18.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 147 |
+
"layers.18.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 148 |
+
"layers.18.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 149 |
+
"layers.18.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 150 |
+
"layers.18.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 151 |
+
"layers.18.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 152 |
+
"layers.18.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 153 |
+
"layers.18.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 154 |
+
"layers.18.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 155 |
+
"layers.18.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 156 |
+
"layers.19.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 157 |
+
"layers.19.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 158 |
+
"layers.19.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 159 |
+
"layers.19.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 160 |
+
"layers.19.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 161 |
+
"layers.19.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
|
| 162 |
+
"layers.19.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 163 |
+
"layers.19.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 164 |
+
"layers.19.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
|
| 165 |
+
"layers.19.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 166 |
+
"layers.19.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 167 |
+
"layers.2.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 168 |
+
"layers.2.linear_attn.A_log": "model-00001-of-00004.safetensors",
|
| 169 |
+
"layers.2.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
|
| 170 |
+
"layers.2.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
|
| 171 |
+
"layers.2.linear_attn.in_proj_a.weight": "model-00001-of-00004.safetensors",
|
| 172 |
+
"layers.2.linear_attn.in_proj_b.weight": "model-00001-of-00004.safetensors",
|
| 173 |
+
"layers.2.linear_attn.in_proj_qkv.weight": "model-00001-of-00004.safetensors",
|
| 174 |
+
"layers.2.linear_attn.in_proj_z.weight": "model-00001-of-00004.safetensors",
|
| 175 |
+
"layers.2.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
|
| 176 |
+
"layers.2.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
|
| 177 |
+
"layers.2.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 178 |
+
"layers.2.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 179 |
+
"layers.2.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 180 |
+
"layers.2.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 181 |
+
"layers.20.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 182 |
+
"layers.20.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 183 |
+
"layers.20.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 184 |
+
"layers.20.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 185 |
+
"layers.20.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 186 |
+
"layers.20.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 187 |
+
"layers.20.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 188 |
+
"layers.20.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 189 |
+
"layers.20.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 190 |
+
"layers.20.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 191 |
+
"layers.20.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 192 |
+
"layers.20.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 193 |
+
"layers.20.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 194 |
+
"layers.20.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 195 |
+
"layers.21.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 196 |
+
"layers.21.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 197 |
+
"layers.21.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 198 |
+
"layers.21.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 199 |
+
"layers.21.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 200 |
+
"layers.21.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 201 |
+
"layers.21.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 202 |
+
"layers.21.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 203 |
+
"layers.21.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 204 |
+
"layers.21.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 205 |
+
"layers.21.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 206 |
+
"layers.21.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 207 |
+
"layers.21.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 208 |
+
"layers.21.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 209 |
+
"layers.22.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 210 |
+
"layers.22.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 211 |
+
"layers.22.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 212 |
+
"layers.22.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 213 |
+
"layers.22.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 214 |
+
"layers.22.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 215 |
+
"layers.22.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 216 |
+
"layers.22.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 217 |
+
"layers.22.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 218 |
+
"layers.22.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 219 |
+
"layers.22.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 220 |
+
"layers.22.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 221 |
+
"layers.22.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 222 |
+
"layers.22.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 223 |
+
"layers.23.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 224 |
+
"layers.23.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 225 |
+
"layers.23.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 226 |
+
"layers.23.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 227 |
+
"layers.23.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 228 |
+
"layers.23.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
|
| 229 |
+
"layers.23.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 230 |
+
"layers.23.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 231 |
+
"layers.23.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
|
| 232 |
+
"layers.23.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 233 |
+
"layers.23.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 234 |
+
"layers.24.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 235 |
+
"layers.24.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 236 |
+
"layers.24.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 237 |
+
"layers.24.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 238 |
+
"layers.24.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 239 |
+
"layers.24.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 240 |
+
"layers.24.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 241 |
+
"layers.24.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 242 |
+
"layers.24.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 243 |
+
"layers.24.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 244 |
+
"layers.24.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 245 |
+
"layers.24.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 246 |
+
"layers.24.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 247 |
+
"layers.24.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 248 |
+
"layers.25.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 249 |
+
"layers.25.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 250 |
+
"layers.25.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 251 |
+
"layers.25.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 252 |
+
"layers.25.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 253 |
+
"layers.25.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 254 |
+
"layers.25.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 255 |
+
"layers.25.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 256 |
+
"layers.25.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 257 |
+
"layers.25.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 258 |
+
"layers.25.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 259 |
+
"layers.25.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 260 |
+
"layers.25.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 261 |
+
"layers.25.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 262 |
+
"layers.26.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 263 |
+
"layers.26.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 264 |
+
"layers.26.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 265 |
+
"layers.26.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 266 |
+
"layers.26.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 267 |
+
"layers.26.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 268 |
+
"layers.26.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 269 |
+
"layers.26.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 270 |
+
"layers.26.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 271 |
+
"layers.26.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 272 |
+
"layers.26.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 273 |
+
"layers.26.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 274 |
+
"layers.26.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 275 |
+
"layers.26.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 276 |
+
"layers.27.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 277 |
+
"layers.27.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 278 |
+
"layers.27.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 279 |
+
"layers.27.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 280 |
+
"layers.27.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 281 |
+
"layers.27.self_attn.k_norm.weight": "model-00003-of-00004.safetensors",
|
| 282 |
+
"layers.27.self_attn.k_proj.weight": "model-00003-of-00004.safetensors",
|
| 283 |
+
"layers.27.self_attn.o_proj.weight": "model-00003-of-00004.safetensors",
|
| 284 |
+
"layers.27.self_attn.q_norm.weight": "model-00003-of-00004.safetensors",
|
| 285 |
+
"layers.27.self_attn.q_proj.weight": "model-00003-of-00004.safetensors",
|
| 286 |
+
"layers.27.self_attn.v_proj.weight": "model-00003-of-00004.safetensors",
|
| 287 |
+
"layers.28.input_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 288 |
+
"layers.28.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 289 |
+
"layers.28.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 290 |
+
"layers.28.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 291 |
+
"layers.28.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 292 |
+
"layers.28.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 293 |
+
"layers.28.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 294 |
+
"layers.28.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 295 |
+
"layers.28.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 296 |
+
"layers.28.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 297 |
+
"layers.28.mlp.down_proj.weight": "model-00003-of-00004.safetensors",
|
| 298 |
+
"layers.28.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 299 |
+
"layers.28.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 300 |
+
"layers.28.post_attention_layernorm.weight": "model-00003-of-00004.safetensors",
|
| 301 |
+
"layers.29.input_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 302 |
+
"layers.29.linear_attn.A_log": "model-00003-of-00004.safetensors",
|
| 303 |
+
"layers.29.linear_attn.conv1d.weight": "model-00003-of-00004.safetensors",
|
| 304 |
+
"layers.29.linear_attn.dt_bias": "model-00003-of-00004.safetensors",
|
| 305 |
+
"layers.29.linear_attn.in_proj_a.weight": "model-00003-of-00004.safetensors",
|
| 306 |
+
"layers.29.linear_attn.in_proj_b.weight": "model-00003-of-00004.safetensors",
|
| 307 |
+
"layers.29.linear_attn.in_proj_qkv.weight": "model-00003-of-00004.safetensors",
|
| 308 |
+
"layers.29.linear_attn.in_proj_z.weight": "model-00003-of-00004.safetensors",
|
| 309 |
+
"layers.29.linear_attn.norm.weight": "model-00003-of-00004.safetensors",
|
| 310 |
+
"layers.29.linear_attn.out_proj.weight": "model-00003-of-00004.safetensors",
|
| 311 |
+
"layers.29.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 312 |
+
"layers.29.mlp.gate_proj.weight": "model-00003-of-00004.safetensors",
|
| 313 |
+
"layers.29.mlp.up_proj.weight": "model-00003-of-00004.safetensors",
|
| 314 |
+
"layers.29.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 315 |
+
"layers.3.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 316 |
+
"layers.3.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 317 |
+
"layers.3.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 318 |
+
"layers.3.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 319 |
+
"layers.3.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 320 |
+
"layers.3.self_attn.k_norm.weight": "model-00001-of-00004.safetensors",
|
| 321 |
+
"layers.3.self_attn.k_proj.weight": "model-00001-of-00004.safetensors",
|
| 322 |
+
"layers.3.self_attn.o_proj.weight": "model-00001-of-00004.safetensors",
|
| 323 |
+
"layers.3.self_attn.q_norm.weight": "model-00001-of-00004.safetensors",
|
| 324 |
+
"layers.3.self_attn.q_proj.weight": "model-00001-of-00004.safetensors",
|
| 325 |
+
"layers.3.self_attn.v_proj.weight": "model-00001-of-00004.safetensors",
|
| 326 |
+
"layers.30.input_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 327 |
+
"layers.30.linear_attn.A_log": "model-00004-of-00004.safetensors",
|
| 328 |
+
"layers.30.linear_attn.conv1d.weight": "model-00004-of-00004.safetensors",
|
| 329 |
+
"layers.30.linear_attn.dt_bias": "model-00004-of-00004.safetensors",
|
| 330 |
+
"layers.30.linear_attn.in_proj_a.weight": "model-00004-of-00004.safetensors",
|
| 331 |
+
"layers.30.linear_attn.in_proj_b.weight": "model-00004-of-00004.safetensors",
|
| 332 |
+
"layers.30.linear_attn.in_proj_qkv.weight": "model-00004-of-00004.safetensors",
|
| 333 |
+
"layers.30.linear_attn.in_proj_z.weight": "model-00004-of-00004.safetensors",
|
| 334 |
+
"layers.30.linear_attn.norm.weight": "model-00004-of-00004.safetensors",
|
| 335 |
+
"layers.30.linear_attn.out_proj.weight": "model-00004-of-00004.safetensors",
|
| 336 |
+
"layers.30.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 337 |
+
"layers.30.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
|
| 338 |
+
"layers.30.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
|
| 339 |
+
"layers.30.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 340 |
+
"layers.31.input_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 341 |
+
"layers.31.mlp.down_proj.weight": "model-00004-of-00004.safetensors",
|
| 342 |
+
"layers.31.mlp.gate_proj.weight": "model-00004-of-00004.safetensors",
|
| 343 |
+
"layers.31.mlp.up_proj.weight": "model-00004-of-00004.safetensors",
|
| 344 |
+
"layers.31.post_attention_layernorm.weight": "model-00004-of-00004.safetensors",
|
| 345 |
+
"layers.31.self_attn.k_norm.weight": "model-00004-of-00004.safetensors",
|
| 346 |
+
"layers.31.self_attn.k_proj.weight": "model-00004-of-00004.safetensors",
|
| 347 |
+
"layers.31.self_attn.o_proj.weight": "model-00004-of-00004.safetensors",
|
| 348 |
+
"layers.31.self_attn.q_norm.weight": "model-00004-of-00004.safetensors",
|
| 349 |
+
"layers.31.self_attn.q_proj.weight": "model-00004-of-00004.safetensors",
|
| 350 |
+
"layers.31.self_attn.v_proj.weight": "model-00004-of-00004.safetensors",
|
| 351 |
+
"layers.4.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 352 |
+
"layers.4.linear_attn.A_log": "model-00001-of-00004.safetensors",
|
| 353 |
+
"layers.4.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
|
| 354 |
+
"layers.4.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
|
| 355 |
+
"layers.4.linear_attn.in_proj_a.weight": "model-00001-of-00004.safetensors",
|
| 356 |
+
"layers.4.linear_attn.in_proj_b.weight": "model-00001-of-00004.safetensors",
|
| 357 |
+
"layers.4.linear_attn.in_proj_qkv.weight": "model-00001-of-00004.safetensors",
|
| 358 |
+
"layers.4.linear_attn.in_proj_z.weight": "model-00001-of-00004.safetensors",
|
| 359 |
+
"layers.4.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
|
| 360 |
+
"layers.4.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
|
| 361 |
+
"layers.4.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 362 |
+
"layers.4.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 363 |
+
"layers.4.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 364 |
+
"layers.4.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 365 |
+
"layers.5.input_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 366 |
+
"layers.5.linear_attn.A_log": "model-00001-of-00004.safetensors",
|
| 367 |
+
"layers.5.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
|
| 368 |
+
"layers.5.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
|
| 369 |
+
"layers.5.linear_attn.in_proj_a.weight": "model-00001-of-00004.safetensors",
|
| 370 |
+
"layers.5.linear_attn.in_proj_b.weight": "model-00001-of-00004.safetensors",
|
| 371 |
+
"layers.5.linear_attn.in_proj_qkv.weight": "model-00001-of-00004.safetensors",
|
| 372 |
+
"layers.5.linear_attn.in_proj_z.weight": "model-00001-of-00004.safetensors",
|
| 373 |
+
"layers.5.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
|
| 374 |
+
"layers.5.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
|
| 375 |
+
"layers.5.mlp.down_proj.weight": "model-00001-of-00004.safetensors",
|
| 376 |
+
"layers.5.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 377 |
+
"layers.5.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 378 |
+
"layers.5.post_attention_layernorm.weight": "model-00001-of-00004.safetensors",
|
| 379 |
+
"layers.6.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 380 |
+
"layers.6.linear_attn.A_log": "model-00001-of-00004.safetensors",
|
| 381 |
+
"layers.6.linear_attn.conv1d.weight": "model-00001-of-00004.safetensors",
|
| 382 |
+
"layers.6.linear_attn.dt_bias": "model-00001-of-00004.safetensors",
|
| 383 |
+
"layers.6.linear_attn.in_proj_a.weight": "model-00001-of-00004.safetensors",
|
| 384 |
+
"layers.6.linear_attn.in_proj_b.weight": "model-00001-of-00004.safetensors",
|
| 385 |
+
"layers.6.linear_attn.in_proj_qkv.weight": "model-00001-of-00004.safetensors",
|
| 386 |
+
"layers.6.linear_attn.in_proj_z.weight": "model-00001-of-00004.safetensors",
|
| 387 |
+
"layers.6.linear_attn.norm.weight": "model-00001-of-00004.safetensors",
|
| 388 |
+
"layers.6.linear_attn.out_proj.weight": "model-00001-of-00004.safetensors",
|
| 389 |
+
"layers.6.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 390 |
+
"layers.6.mlp.gate_proj.weight": "model-00001-of-00004.safetensors",
|
| 391 |
+
"layers.6.mlp.up_proj.weight": "model-00001-of-00004.safetensors",
|
| 392 |
+
"layers.6.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 393 |
+
"layers.7.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 394 |
+
"layers.7.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 395 |
+
"layers.7.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 396 |
+
"layers.7.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 397 |
+
"layers.7.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 398 |
+
"layers.7.self_attn.k_norm.weight": "model-00002-of-00004.safetensors",
|
| 399 |
+
"layers.7.self_attn.k_proj.weight": "model-00002-of-00004.safetensors",
|
| 400 |
+
"layers.7.self_attn.o_proj.weight": "model-00002-of-00004.safetensors",
|
| 401 |
+
"layers.7.self_attn.q_norm.weight": "model-00002-of-00004.safetensors",
|
| 402 |
+
"layers.7.self_attn.q_proj.weight": "model-00002-of-00004.safetensors",
|
| 403 |
+
"layers.7.self_attn.v_proj.weight": "model-00002-of-00004.safetensors",
|
| 404 |
+
"layers.8.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 405 |
+
"layers.8.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 406 |
+
"layers.8.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 407 |
+
"layers.8.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 408 |
+
"layers.8.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 409 |
+
"layers.8.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 410 |
+
"layers.8.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 411 |
+
"layers.8.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 412 |
+
"layers.8.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 413 |
+
"layers.8.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 414 |
+
"layers.8.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 415 |
+
"layers.8.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 416 |
+
"layers.8.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 417 |
+
"layers.8.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 418 |
+
"layers.9.input_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 419 |
+
"layers.9.linear_attn.A_log": "model-00002-of-00004.safetensors",
|
| 420 |
+
"layers.9.linear_attn.conv1d.weight": "model-00002-of-00004.safetensors",
|
| 421 |
+
"layers.9.linear_attn.dt_bias": "model-00002-of-00004.safetensors",
|
| 422 |
+
"layers.9.linear_attn.in_proj_a.weight": "model-00002-of-00004.safetensors",
|
| 423 |
+
"layers.9.linear_attn.in_proj_b.weight": "model-00002-of-00004.safetensors",
|
| 424 |
+
"layers.9.linear_attn.in_proj_qkv.weight": "model-00002-of-00004.safetensors",
|
| 425 |
+
"layers.9.linear_attn.in_proj_z.weight": "model-00002-of-00004.safetensors",
|
| 426 |
+
"layers.9.linear_attn.norm.weight": "model-00002-of-00004.safetensors",
|
| 427 |
+
"layers.9.linear_attn.out_proj.weight": "model-00002-of-00004.safetensors",
|
| 428 |
+
"layers.9.mlp.down_proj.weight": "model-00002-of-00004.safetensors",
|
| 429 |
+
"layers.9.mlp.gate_proj.weight": "model-00002-of-00004.safetensors",
|
| 430 |
+
"layers.9.mlp.up_proj.weight": "model-00002-of-00004.safetensors",
|
| 431 |
+
"layers.9.post_attention_layernorm.weight": "model-00002-of-00004.safetensors",
|
| 432 |
+
"norm.weight": "model-00004-of-00004.safetensors"
|
| 433 |
+
}
|
| 434 |
+
}
|
tokenizer/chat_template.jinja
ADDED
|
@@ -0,0 +1,154 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- set image_count = namespace(value=0) %}
|
| 2 |
+
{%- set video_count = namespace(value=0) %}
|
| 3 |
+
{%- macro render_content(content, do_vision_count, is_system_content=false) %}
|
| 4 |
+
{%- if content is string %}
|
| 5 |
+
{{- content }}
|
| 6 |
+
{%- elif content is iterable and content is not mapping %}
|
| 7 |
+
{%- for item in content %}
|
| 8 |
+
{%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
|
| 9 |
+
{%- if is_system_content %}
|
| 10 |
+
{{- raise_exception('System message cannot contain images.') }}
|
| 11 |
+
{%- endif %}
|
| 12 |
+
{%- if do_vision_count %}
|
| 13 |
+
{%- set image_count.value = image_count.value + 1 %}
|
| 14 |
+
{%- endif %}
|
| 15 |
+
{%- if add_vision_id %}
|
| 16 |
+
{{- 'Picture ' ~ image_count.value ~ ': ' }}
|
| 17 |
+
{%- endif %}
|
| 18 |
+
{{- '<|vision_start|><|image_pad|><|vision_end|>' }}
|
| 19 |
+
{%- elif 'video' in item or item.type == 'video' %}
|
| 20 |
+
{%- if is_system_content %}
|
| 21 |
+
{{- raise_exception('System message cannot contain videos.') }}
|
| 22 |
+
{%- endif %}
|
| 23 |
+
{%- if do_vision_count %}
|
| 24 |
+
{%- set video_count.value = video_count.value + 1 %}
|
| 25 |
+
{%- endif %}
|
| 26 |
+
{%- if add_vision_id %}
|
| 27 |
+
{{- 'Video ' ~ video_count.value ~ ': ' }}
|
| 28 |
+
{%- endif %}
|
| 29 |
+
{{- '<|vision_start|><|video_pad|><|vision_end|>' }}
|
| 30 |
+
{%- elif 'text' in item %}
|
| 31 |
+
{{- item.text }}
|
| 32 |
+
{%- else %}
|
| 33 |
+
{{- raise_exception('Unexpected item type in content.') }}
|
| 34 |
+
{%- endif %}
|
| 35 |
+
{%- endfor %}
|
| 36 |
+
{%- elif content is none or content is undefined %}
|
| 37 |
+
{{- '' }}
|
| 38 |
+
{%- else %}
|
| 39 |
+
{{- raise_exception('Unexpected content type.') }}
|
| 40 |
+
{%- endif %}
|
| 41 |
+
{%- endmacro %}
|
| 42 |
+
{%- if not messages %}
|
| 43 |
+
{{- raise_exception('No messages provided.') }}
|
| 44 |
+
{%- endif %}
|
| 45 |
+
{%- if tools and tools is iterable and tools is not mapping %}
|
| 46 |
+
{{- '<|im_start|>system\n' }}
|
| 47 |
+
{{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
|
| 48 |
+
{%- for tool in tools %}
|
| 49 |
+
{{- "\n" }}
|
| 50 |
+
{{- tool | tojson }}
|
| 51 |
+
{%- endfor %}
|
| 52 |
+
{{- "\n</tools>" }}
|
| 53 |
+
{{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
|
| 54 |
+
{%- if messages[0].role == 'system' %}
|
| 55 |
+
{%- set content = render_content(messages[0].content, false, true)|trim %}
|
| 56 |
+
{%- if content %}
|
| 57 |
+
{{- '\n\n' + content }}
|
| 58 |
+
{%- endif %}
|
| 59 |
+
{%- endif %}
|
| 60 |
+
{{- '<|im_end|>\n' }}
|
| 61 |
+
{%- else %}
|
| 62 |
+
{%- if messages[0].role == 'system' %}
|
| 63 |
+
{%- set content = render_content(messages[0].content, false, true)|trim %}
|
| 64 |
+
{{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
|
| 65 |
+
{%- endif %}
|
| 66 |
+
{%- endif %}
|
| 67 |
+
{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
|
| 68 |
+
{%- for message in messages[::-1] %}
|
| 69 |
+
{%- set index = (messages|length - 1) - loop.index0 %}
|
| 70 |
+
{%- if ns.multi_step_tool and message.role == "user" %}
|
| 71 |
+
{%- set content = render_content(message.content, false)|trim %}
|
| 72 |
+
{%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
|
| 73 |
+
{%- set ns.multi_step_tool = false %}
|
| 74 |
+
{%- set ns.last_query_index = index %}
|
| 75 |
+
{%- endif %}
|
| 76 |
+
{%- endif %}
|
| 77 |
+
{%- endfor %}
|
| 78 |
+
{%- if ns.multi_step_tool %}
|
| 79 |
+
{{- raise_exception('No user query found in messages.') }}
|
| 80 |
+
{%- endif %}
|
| 81 |
+
{%- for message in messages %}
|
| 82 |
+
{%- set content = render_content(message.content, true)|trim %}
|
| 83 |
+
{%- if message.role == "system" %}
|
| 84 |
+
{%- if not loop.first %}
|
| 85 |
+
{{- raise_exception('System message must be at the beginning.') }}
|
| 86 |
+
{%- endif %}
|
| 87 |
+
{%- elif message.role == "user" %}
|
| 88 |
+
{{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
|
| 89 |
+
{%- elif message.role == "assistant" %}
|
| 90 |
+
{%- set reasoning_content = '' %}
|
| 91 |
+
{%- if message.reasoning_content is string %}
|
| 92 |
+
{%- set reasoning_content = message.reasoning_content %}
|
| 93 |
+
{%- else %}
|
| 94 |
+
{%- if '</think>' in content %}
|
| 95 |
+
{%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
|
| 96 |
+
{%- set content = content.split('</think>')[-1].lstrip('\n') %}
|
| 97 |
+
{%- endif %}
|
| 98 |
+
{%- endif %}
|
| 99 |
+
{%- set reasoning_content = reasoning_content|trim %}
|
| 100 |
+
{%- if loop.index0 > ns.last_query_index %}
|
| 101 |
+
{{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
|
| 102 |
+
{%- else %}
|
| 103 |
+
{{- '<|im_start|>' + message.role + '\n' + content }}
|
| 104 |
+
{%- endif %}
|
| 105 |
+
{%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
|
| 106 |
+
{%- for tool_call in message.tool_calls %}
|
| 107 |
+
{%- if tool_call.function is defined %}
|
| 108 |
+
{%- set tool_call = tool_call.function %}
|
| 109 |
+
{%- endif %}
|
| 110 |
+
{%- if loop.first %}
|
| 111 |
+
{%- if content|trim %}
|
| 112 |
+
{{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
|
| 113 |
+
{%- else %}
|
| 114 |
+
{{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
|
| 115 |
+
{%- endif %}
|
| 116 |
+
{%- else %}
|
| 117 |
+
{{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
|
| 118 |
+
{%- endif %}
|
| 119 |
+
{%- if tool_call.arguments is defined %}
|
| 120 |
+
{%- for args_name, args_value in tool_call.arguments|items %}
|
| 121 |
+
{{- '<parameter=' + args_name + '>\n' }}
|
| 122 |
+
{%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
|
| 123 |
+
{{- args_value }}
|
| 124 |
+
{{- '\n</parameter>\n' }}
|
| 125 |
+
{%- endfor %}
|
| 126 |
+
{%- endif %}
|
| 127 |
+
{{- '</function>\n</tool_call>' }}
|
| 128 |
+
{%- endfor %}
|
| 129 |
+
{%- endif %}
|
| 130 |
+
{{- '<|im_end|>\n' }}
|
| 131 |
+
{%- elif message.role == "tool" %}
|
| 132 |
+
{%- if loop.previtem and loop.previtem.role != "tool" %}
|
| 133 |
+
{{- '<|im_start|>user' }}
|
| 134 |
+
{%- endif %}
|
| 135 |
+
{{- '\n<tool_response>\n' }}
|
| 136 |
+
{{- content }}
|
| 137 |
+
{{- '\n</tool_response>' }}
|
| 138 |
+
{%- if not loop.last and loop.nextitem.role != "tool" %}
|
| 139 |
+
{{- '<|im_end|>\n' }}
|
| 140 |
+
{%- elif loop.last %}
|
| 141 |
+
{{- '<|im_end|>\n' }}
|
| 142 |
+
{%- endif %}
|
| 143 |
+
{%- else %}
|
| 144 |
+
{{- raise_exception('Unexpected message role.') }}
|
| 145 |
+
{%- endif %}
|
| 146 |
+
{%- endfor %}
|
| 147 |
+
{%- if add_generation_prompt %}
|
| 148 |
+
{{- '<|im_start|>assistant\n' }}
|
| 149 |
+
{%- if enable_thinking is defined and enable_thinking is false %}
|
| 150 |
+
{{- '<think>\n\n</think>\n\n' }}
|
| 151 |
+
{%- else %}
|
| 152 |
+
{{- '<think>\n' }}
|
| 153 |
+
{%- endif %}
|
| 154 |
+
{%- endif %}
|
tokenizer/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06b9509352d2af50381ab2247e083b80d32d5c0aba91c272ca9ff729b6a0e523
|
| 3 |
+
size 19989325
|
tokenizer/tokenizer_config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"audio_bos_token": "<|audio_start|>",
|
| 4 |
+
"audio_eos_token": "<|audio_end|>",
|
| 5 |
+
"audio_token": "<|audio_pad|>",
|
| 6 |
+
"backend": "tokenizers",
|
| 7 |
+
"bos_token": null,
|
| 8 |
+
"clean_up_tokenization_spaces": false,
|
| 9 |
+
"eos_token": "<|im_end|>",
|
| 10 |
+
"errors": "replace",
|
| 11 |
+
"image_token": "<|image_pad|>",
|
| 12 |
+
"is_local": true,
|
| 13 |
+
"local_files_only": true,
|
| 14 |
+
"model_max_length": 262144,
|
| 15 |
+
"model_specific_special_tokens": {
|
| 16 |
+
"audio_bos_token": "<|audio_start|>",
|
| 17 |
+
"audio_eos_token": "<|audio_end|>",
|
| 18 |
+
"audio_token": "<|audio_pad|>",
|
| 19 |
+
"image_token": "<|image_pad|>",
|
| 20 |
+
"video_token": "<|video_pad|>",
|
| 21 |
+
"vision_bos_token": "<|vision_start|>",
|
| 22 |
+
"vision_eos_token": "<|vision_end|>"
|
| 23 |
+
},
|
| 24 |
+
"pad_token": "<|endoftext|>",
|
| 25 |
+
"pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
|
| 26 |
+
"split_special_tokens": false,
|
| 27 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 28 |
+
"unk_token": null,
|
| 29 |
+
"video_token": "<|video_pad|>",
|
| 30 |
+
"vision_bos_token": "<|vision_start|>",
|
| 31 |
+
"vision_eos_token": "<|vision_end|>"
|
| 32 |
+
}
|
transformer/config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "MOSSImageTransformer2DModel",
|
| 3 |
+
"_diffusers_version": "0.41.0.dev0",
|
| 4 |
+
"attention_head_dim": 128,
|
| 5 |
+
"ffn_dim": 8192,
|
| 6 |
+
"in_channels": 128,
|
| 7 |
+
"num_attention_heads": 24,
|
| 8 |
+
"num_layers": 42,
|
| 9 |
+
"num_refiner_layers": 2,
|
| 10 |
+
"rope_theta": 10000.0,
|
| 11 |
+
"text_embed_dim": 4096,
|
| 12 |
+
"timestep_embed_dim": 256
|
| 13 |
+
}
|
transformer/diffusion_pytorch_model-00001-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4fb54fb6c20195e177d80ddbfc5b72a9d0d8f4fca171acd95381d07cf3c6dcf0
|
| 3 |
+
size 4991980336
|
transformer/diffusion_pytorch_model-00002-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e2061d3dcb25e4fa3686e486f4ee85853054a0b87b996cb5c83fa571ee39ae1
|
| 3 |
+
size 4983139792
|
transformer/diffusion_pytorch_model-00003-of-00003.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1a535dc8a3fc6fc1d71027c2a643d28d791064f1a73c42b20dc9c06d6fe775a
|
| 3 |
+
size 189549160
|
transformer/diffusion_pytorch_model.safetensors.index.json
ADDED
|
@@ -0,0 +1,469 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_size": 10164617216
|
| 4 |
+
},
|
| 5 |
+
"weight_map": {
|
| 6 |
+
"block_modulation.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 7 |
+
"block_modulation.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 8 |
+
"context_embedder.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 9 |
+
"context_embedder.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 10 |
+
"context_refiner.0.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 11 |
+
"context_refiner.0.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 12 |
+
"context_refiner.0.attn.to_out.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 13 |
+
"context_refiner.0.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 14 |
+
"context_refiner.0.attn.to_qkv.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 15 |
+
"context_refiner.0.attn.to_qkv.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 16 |
+
"context_refiner.0.ff_down.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 17 |
+
"context_refiner.0.ff_down.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 18 |
+
"context_refiner.0.ff_gate.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 19 |
+
"context_refiner.0.ff_gate.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 20 |
+
"context_refiner.0.ff_up.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 21 |
+
"context_refiner.0.ff_up.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 22 |
+
"context_refiner.0.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 23 |
+
"context_refiner.0.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 24 |
+
"context_refiner.1.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 25 |
+
"context_refiner.1.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 26 |
+
"context_refiner.1.attn.to_out.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 27 |
+
"context_refiner.1.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 28 |
+
"context_refiner.1.attn.to_qkv.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 29 |
+
"context_refiner.1.attn.to_qkv.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 30 |
+
"context_refiner.1.ff_down.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 31 |
+
"context_refiner.1.ff_down.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 32 |
+
"context_refiner.1.ff_gate.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 33 |
+
"context_refiner.1.ff_gate.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 34 |
+
"context_refiner.1.ff_up.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 35 |
+
"context_refiner.1.ff_up.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 36 |
+
"context_refiner.1.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 37 |
+
"context_refiner.1.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 38 |
+
"norm_out.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 39 |
+
"norm_out_modulation.bias": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 40 |
+
"norm_out_modulation.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 41 |
+
"proj_out.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 42 |
+
"time_embedder.linear_1.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 43 |
+
"time_embedder.linear_1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 44 |
+
"time_embedder.linear_2.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 45 |
+
"time_embedder.linear_2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 46 |
+
"transformer_blocks.0.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 47 |
+
"transformer_blocks.0.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 48 |
+
"transformer_blocks.0.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 49 |
+
"transformer_blocks.0.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 50 |
+
"transformer_blocks.0.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 51 |
+
"transformer_blocks.0.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 52 |
+
"transformer_blocks.0.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 53 |
+
"transformer_blocks.0.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 54 |
+
"transformer_blocks.0.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 55 |
+
"transformer_blocks.0.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 56 |
+
"transformer_blocks.1.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 57 |
+
"transformer_blocks.1.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 58 |
+
"transformer_blocks.1.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 59 |
+
"transformer_blocks.1.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 60 |
+
"transformer_blocks.1.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 61 |
+
"transformer_blocks.1.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 62 |
+
"transformer_blocks.1.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 63 |
+
"transformer_blocks.1.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 64 |
+
"transformer_blocks.1.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 65 |
+
"transformer_blocks.1.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 66 |
+
"transformer_blocks.10.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 67 |
+
"transformer_blocks.10.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 68 |
+
"transformer_blocks.10.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 69 |
+
"transformer_blocks.10.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 70 |
+
"transformer_blocks.10.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 71 |
+
"transformer_blocks.10.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 72 |
+
"transformer_blocks.10.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 73 |
+
"transformer_blocks.10.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 74 |
+
"transformer_blocks.10.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 75 |
+
"transformer_blocks.10.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 76 |
+
"transformer_blocks.11.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 77 |
+
"transformer_blocks.11.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 78 |
+
"transformer_blocks.11.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 79 |
+
"transformer_blocks.11.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 80 |
+
"transformer_blocks.11.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 81 |
+
"transformer_blocks.11.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 82 |
+
"transformer_blocks.11.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 83 |
+
"transformer_blocks.11.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 84 |
+
"transformer_blocks.11.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 85 |
+
"transformer_blocks.11.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 86 |
+
"transformer_blocks.12.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 87 |
+
"transformer_blocks.12.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 88 |
+
"transformer_blocks.12.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 89 |
+
"transformer_blocks.12.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 90 |
+
"transformer_blocks.12.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 91 |
+
"transformer_blocks.12.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 92 |
+
"transformer_blocks.12.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 93 |
+
"transformer_blocks.12.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 94 |
+
"transformer_blocks.12.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 95 |
+
"transformer_blocks.12.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 96 |
+
"transformer_blocks.13.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 97 |
+
"transformer_blocks.13.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 98 |
+
"transformer_blocks.13.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 99 |
+
"transformer_blocks.13.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 100 |
+
"transformer_blocks.13.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 101 |
+
"transformer_blocks.13.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 102 |
+
"transformer_blocks.13.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 103 |
+
"transformer_blocks.13.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 104 |
+
"transformer_blocks.13.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 105 |
+
"transformer_blocks.13.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 106 |
+
"transformer_blocks.14.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 107 |
+
"transformer_blocks.14.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 108 |
+
"transformer_blocks.14.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 109 |
+
"transformer_blocks.14.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 110 |
+
"transformer_blocks.14.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 111 |
+
"transformer_blocks.14.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 112 |
+
"transformer_blocks.14.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 113 |
+
"transformer_blocks.14.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 114 |
+
"transformer_blocks.14.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 115 |
+
"transformer_blocks.14.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 116 |
+
"transformer_blocks.15.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 117 |
+
"transformer_blocks.15.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 118 |
+
"transformer_blocks.15.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 119 |
+
"transformer_blocks.15.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 120 |
+
"transformer_blocks.15.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 121 |
+
"transformer_blocks.15.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 122 |
+
"transformer_blocks.15.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 123 |
+
"transformer_blocks.15.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 124 |
+
"transformer_blocks.15.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 125 |
+
"transformer_blocks.15.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 126 |
+
"transformer_blocks.16.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 127 |
+
"transformer_blocks.16.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 128 |
+
"transformer_blocks.16.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 129 |
+
"transformer_blocks.16.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 130 |
+
"transformer_blocks.16.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 131 |
+
"transformer_blocks.16.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 132 |
+
"transformer_blocks.16.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 133 |
+
"transformer_blocks.16.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 134 |
+
"transformer_blocks.16.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 135 |
+
"transformer_blocks.16.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 136 |
+
"transformer_blocks.17.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 137 |
+
"transformer_blocks.17.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 138 |
+
"transformer_blocks.17.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 139 |
+
"transformer_blocks.17.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 140 |
+
"transformer_blocks.17.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 141 |
+
"transformer_blocks.17.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 142 |
+
"transformer_blocks.17.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 143 |
+
"transformer_blocks.17.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 144 |
+
"transformer_blocks.17.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 145 |
+
"transformer_blocks.17.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 146 |
+
"transformer_blocks.18.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 147 |
+
"transformer_blocks.18.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 148 |
+
"transformer_blocks.18.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 149 |
+
"transformer_blocks.18.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 150 |
+
"transformer_blocks.18.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 151 |
+
"transformer_blocks.18.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 152 |
+
"transformer_blocks.18.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 153 |
+
"transformer_blocks.18.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 154 |
+
"transformer_blocks.18.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 155 |
+
"transformer_blocks.18.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 156 |
+
"transformer_blocks.19.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 157 |
+
"transformer_blocks.19.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 158 |
+
"transformer_blocks.19.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 159 |
+
"transformer_blocks.19.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 160 |
+
"transformer_blocks.19.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 161 |
+
"transformer_blocks.19.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 162 |
+
"transformer_blocks.19.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 163 |
+
"transformer_blocks.19.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 164 |
+
"transformer_blocks.19.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 165 |
+
"transformer_blocks.19.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 166 |
+
"transformer_blocks.2.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 167 |
+
"transformer_blocks.2.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 168 |
+
"transformer_blocks.2.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 169 |
+
"transformer_blocks.2.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 170 |
+
"transformer_blocks.2.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 171 |
+
"transformer_blocks.2.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 172 |
+
"transformer_blocks.2.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 173 |
+
"transformer_blocks.2.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 174 |
+
"transformer_blocks.2.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 175 |
+
"transformer_blocks.2.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 176 |
+
"transformer_blocks.20.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 177 |
+
"transformer_blocks.20.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 178 |
+
"transformer_blocks.20.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 179 |
+
"transformer_blocks.20.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 180 |
+
"transformer_blocks.20.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 181 |
+
"transformer_blocks.20.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 182 |
+
"transformer_blocks.20.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 183 |
+
"transformer_blocks.20.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 184 |
+
"transformer_blocks.20.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 185 |
+
"transformer_blocks.20.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 186 |
+
"transformer_blocks.21.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 187 |
+
"transformer_blocks.21.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 188 |
+
"transformer_blocks.21.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 189 |
+
"transformer_blocks.21.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 190 |
+
"transformer_blocks.21.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 191 |
+
"transformer_blocks.21.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 192 |
+
"transformer_blocks.21.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 193 |
+
"transformer_blocks.21.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 194 |
+
"transformer_blocks.21.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 195 |
+
"transformer_blocks.21.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 196 |
+
"transformer_blocks.22.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 197 |
+
"transformer_blocks.22.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 198 |
+
"transformer_blocks.22.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 199 |
+
"transformer_blocks.22.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 200 |
+
"transformer_blocks.22.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 201 |
+
"transformer_blocks.22.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 202 |
+
"transformer_blocks.22.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 203 |
+
"transformer_blocks.22.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 204 |
+
"transformer_blocks.22.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 205 |
+
"transformer_blocks.22.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 206 |
+
"transformer_blocks.23.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 207 |
+
"transformer_blocks.23.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 208 |
+
"transformer_blocks.23.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 209 |
+
"transformer_blocks.23.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 210 |
+
"transformer_blocks.23.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 211 |
+
"transformer_blocks.23.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 212 |
+
"transformer_blocks.23.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 213 |
+
"transformer_blocks.23.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 214 |
+
"transformer_blocks.23.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 215 |
+
"transformer_blocks.23.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 216 |
+
"transformer_blocks.24.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 217 |
+
"transformer_blocks.24.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 218 |
+
"transformer_blocks.24.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 219 |
+
"transformer_blocks.24.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 220 |
+
"transformer_blocks.24.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 221 |
+
"transformer_blocks.24.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 222 |
+
"transformer_blocks.24.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 223 |
+
"transformer_blocks.24.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 224 |
+
"transformer_blocks.24.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 225 |
+
"transformer_blocks.24.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 226 |
+
"transformer_blocks.25.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 227 |
+
"transformer_blocks.25.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 228 |
+
"transformer_blocks.25.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 229 |
+
"transformer_blocks.25.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 230 |
+
"transformer_blocks.25.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 231 |
+
"transformer_blocks.25.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 232 |
+
"transformer_blocks.25.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 233 |
+
"transformer_blocks.25.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 234 |
+
"transformer_blocks.25.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 235 |
+
"transformer_blocks.25.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 236 |
+
"transformer_blocks.26.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 237 |
+
"transformer_blocks.26.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 238 |
+
"transformer_blocks.26.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 239 |
+
"transformer_blocks.26.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 240 |
+
"transformer_blocks.26.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 241 |
+
"transformer_blocks.26.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 242 |
+
"transformer_blocks.26.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 243 |
+
"transformer_blocks.26.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 244 |
+
"transformer_blocks.26.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 245 |
+
"transformer_blocks.26.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 246 |
+
"transformer_blocks.27.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 247 |
+
"transformer_blocks.27.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 248 |
+
"transformer_blocks.27.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 249 |
+
"transformer_blocks.27.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 250 |
+
"transformer_blocks.27.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 251 |
+
"transformer_blocks.27.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 252 |
+
"transformer_blocks.27.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 253 |
+
"transformer_blocks.27.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 254 |
+
"transformer_blocks.27.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 255 |
+
"transformer_blocks.27.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 256 |
+
"transformer_blocks.28.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 257 |
+
"transformer_blocks.28.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 258 |
+
"transformer_blocks.28.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 259 |
+
"transformer_blocks.28.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 260 |
+
"transformer_blocks.28.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 261 |
+
"transformer_blocks.28.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 262 |
+
"transformer_blocks.28.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 263 |
+
"transformer_blocks.28.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 264 |
+
"transformer_blocks.28.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 265 |
+
"transformer_blocks.28.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 266 |
+
"transformer_blocks.29.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 267 |
+
"transformer_blocks.29.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 268 |
+
"transformer_blocks.29.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 269 |
+
"transformer_blocks.29.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 270 |
+
"transformer_blocks.29.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 271 |
+
"transformer_blocks.29.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 272 |
+
"transformer_blocks.29.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 273 |
+
"transformer_blocks.29.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 274 |
+
"transformer_blocks.29.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 275 |
+
"transformer_blocks.29.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 276 |
+
"transformer_blocks.3.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 277 |
+
"transformer_blocks.3.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 278 |
+
"transformer_blocks.3.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 279 |
+
"transformer_blocks.3.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 280 |
+
"transformer_blocks.3.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 281 |
+
"transformer_blocks.3.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 282 |
+
"transformer_blocks.3.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 283 |
+
"transformer_blocks.3.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 284 |
+
"transformer_blocks.3.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 285 |
+
"transformer_blocks.3.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 286 |
+
"transformer_blocks.30.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 287 |
+
"transformer_blocks.30.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 288 |
+
"transformer_blocks.30.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 289 |
+
"transformer_blocks.30.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 290 |
+
"transformer_blocks.30.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 291 |
+
"transformer_blocks.30.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 292 |
+
"transformer_blocks.30.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 293 |
+
"transformer_blocks.30.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 294 |
+
"transformer_blocks.30.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 295 |
+
"transformer_blocks.30.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 296 |
+
"transformer_blocks.31.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 297 |
+
"transformer_blocks.31.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 298 |
+
"transformer_blocks.31.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 299 |
+
"transformer_blocks.31.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 300 |
+
"transformer_blocks.31.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 301 |
+
"transformer_blocks.31.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 302 |
+
"transformer_blocks.31.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 303 |
+
"transformer_blocks.31.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 304 |
+
"transformer_blocks.31.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 305 |
+
"transformer_blocks.31.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 306 |
+
"transformer_blocks.32.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 307 |
+
"transformer_blocks.32.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 308 |
+
"transformer_blocks.32.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 309 |
+
"transformer_blocks.32.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 310 |
+
"transformer_blocks.32.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 311 |
+
"transformer_blocks.32.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 312 |
+
"transformer_blocks.32.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 313 |
+
"transformer_blocks.32.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 314 |
+
"transformer_blocks.32.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 315 |
+
"transformer_blocks.32.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 316 |
+
"transformer_blocks.33.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 317 |
+
"transformer_blocks.33.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 318 |
+
"transformer_blocks.33.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 319 |
+
"transformer_blocks.33.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 320 |
+
"transformer_blocks.33.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 321 |
+
"transformer_blocks.33.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 322 |
+
"transformer_blocks.33.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 323 |
+
"transformer_blocks.33.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 324 |
+
"transformer_blocks.33.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 325 |
+
"transformer_blocks.33.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 326 |
+
"transformer_blocks.34.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 327 |
+
"transformer_blocks.34.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 328 |
+
"transformer_blocks.34.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 329 |
+
"transformer_blocks.34.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 330 |
+
"transformer_blocks.34.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 331 |
+
"transformer_blocks.34.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 332 |
+
"transformer_blocks.34.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 333 |
+
"transformer_blocks.34.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 334 |
+
"transformer_blocks.34.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 335 |
+
"transformer_blocks.34.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 336 |
+
"transformer_blocks.35.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 337 |
+
"transformer_blocks.35.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 338 |
+
"transformer_blocks.35.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 339 |
+
"transformer_blocks.35.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 340 |
+
"transformer_blocks.35.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 341 |
+
"transformer_blocks.35.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 342 |
+
"transformer_blocks.35.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 343 |
+
"transformer_blocks.35.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 344 |
+
"transformer_blocks.35.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 345 |
+
"transformer_blocks.35.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 346 |
+
"transformer_blocks.36.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 347 |
+
"transformer_blocks.36.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 348 |
+
"transformer_blocks.36.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 349 |
+
"transformer_blocks.36.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 350 |
+
"transformer_blocks.36.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 351 |
+
"transformer_blocks.36.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 352 |
+
"transformer_blocks.36.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 353 |
+
"transformer_blocks.36.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 354 |
+
"transformer_blocks.36.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 355 |
+
"transformer_blocks.36.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 356 |
+
"transformer_blocks.37.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 357 |
+
"transformer_blocks.37.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 358 |
+
"transformer_blocks.37.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 359 |
+
"transformer_blocks.37.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 360 |
+
"transformer_blocks.37.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 361 |
+
"transformer_blocks.37.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 362 |
+
"transformer_blocks.37.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 363 |
+
"transformer_blocks.37.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 364 |
+
"transformer_blocks.37.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 365 |
+
"transformer_blocks.37.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 366 |
+
"transformer_blocks.38.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 367 |
+
"transformer_blocks.38.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 368 |
+
"transformer_blocks.38.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 369 |
+
"transformer_blocks.38.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 370 |
+
"transformer_blocks.38.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 371 |
+
"transformer_blocks.38.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 372 |
+
"transformer_blocks.38.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 373 |
+
"transformer_blocks.38.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 374 |
+
"transformer_blocks.38.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 375 |
+
"transformer_blocks.38.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 376 |
+
"transformer_blocks.39.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 377 |
+
"transformer_blocks.39.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 378 |
+
"transformer_blocks.39.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 379 |
+
"transformer_blocks.39.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 380 |
+
"transformer_blocks.39.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 381 |
+
"transformer_blocks.39.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 382 |
+
"transformer_blocks.39.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 383 |
+
"transformer_blocks.39.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 384 |
+
"transformer_blocks.39.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 385 |
+
"transformer_blocks.39.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 386 |
+
"transformer_blocks.4.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 387 |
+
"transformer_blocks.4.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 388 |
+
"transformer_blocks.4.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 389 |
+
"transformer_blocks.4.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 390 |
+
"transformer_blocks.4.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 391 |
+
"transformer_blocks.4.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 392 |
+
"transformer_blocks.4.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 393 |
+
"transformer_blocks.4.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 394 |
+
"transformer_blocks.4.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 395 |
+
"transformer_blocks.4.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 396 |
+
"transformer_blocks.40.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 397 |
+
"transformer_blocks.40.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 398 |
+
"transformer_blocks.40.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 399 |
+
"transformer_blocks.40.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 400 |
+
"transformer_blocks.40.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 401 |
+
"transformer_blocks.40.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 402 |
+
"transformer_blocks.40.ff_in.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 403 |
+
"transformer_blocks.40.ff_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 404 |
+
"transformer_blocks.40.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 405 |
+
"transformer_blocks.40.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 406 |
+
"transformer_blocks.41.attn.norm_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 407 |
+
"transformer_blocks.41.attn.norm_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 408 |
+
"transformer_blocks.41.attn.to_k.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 409 |
+
"transformer_blocks.41.attn.to_out.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 410 |
+
"transformer_blocks.41.attn.to_q.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 411 |
+
"transformer_blocks.41.attn.to_v.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 412 |
+
"transformer_blocks.41.ff_in.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 413 |
+
"transformer_blocks.41.ff_out.weight": "diffusion_pytorch_model-00003-of-00003.safetensors",
|
| 414 |
+
"transformer_blocks.41.norm1.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 415 |
+
"transformer_blocks.41.norm2.weight": "diffusion_pytorch_model-00002-of-00003.safetensors",
|
| 416 |
+
"transformer_blocks.5.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 417 |
+
"transformer_blocks.5.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 418 |
+
"transformer_blocks.5.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 419 |
+
"transformer_blocks.5.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 420 |
+
"transformer_blocks.5.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 421 |
+
"transformer_blocks.5.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 422 |
+
"transformer_blocks.5.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 423 |
+
"transformer_blocks.5.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 424 |
+
"transformer_blocks.5.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 425 |
+
"transformer_blocks.5.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 426 |
+
"transformer_blocks.6.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 427 |
+
"transformer_blocks.6.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 428 |
+
"transformer_blocks.6.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 429 |
+
"transformer_blocks.6.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 430 |
+
"transformer_blocks.6.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 431 |
+
"transformer_blocks.6.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 432 |
+
"transformer_blocks.6.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 433 |
+
"transformer_blocks.6.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 434 |
+
"transformer_blocks.6.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 435 |
+
"transformer_blocks.6.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 436 |
+
"transformer_blocks.7.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 437 |
+
"transformer_blocks.7.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 438 |
+
"transformer_blocks.7.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 439 |
+
"transformer_blocks.7.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 440 |
+
"transformer_blocks.7.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 441 |
+
"transformer_blocks.7.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 442 |
+
"transformer_blocks.7.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 443 |
+
"transformer_blocks.7.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 444 |
+
"transformer_blocks.7.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 445 |
+
"transformer_blocks.7.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 446 |
+
"transformer_blocks.8.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 447 |
+
"transformer_blocks.8.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 448 |
+
"transformer_blocks.8.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 449 |
+
"transformer_blocks.8.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 450 |
+
"transformer_blocks.8.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 451 |
+
"transformer_blocks.8.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 452 |
+
"transformer_blocks.8.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 453 |
+
"transformer_blocks.8.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 454 |
+
"transformer_blocks.8.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 455 |
+
"transformer_blocks.8.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 456 |
+
"transformer_blocks.9.attn.norm_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 457 |
+
"transformer_blocks.9.attn.norm_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 458 |
+
"transformer_blocks.9.attn.to_k.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 459 |
+
"transformer_blocks.9.attn.to_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 460 |
+
"transformer_blocks.9.attn.to_q.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 461 |
+
"transformer_blocks.9.attn.to_v.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 462 |
+
"transformer_blocks.9.ff_in.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 463 |
+
"transformer_blocks.9.ff_out.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 464 |
+
"transformer_blocks.9.norm1.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 465 |
+
"transformer_blocks.9.norm2.weight": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 466 |
+
"x_embedder.bias": "diffusion_pytorch_model-00001-of-00003.safetensors",
|
| 467 |
+
"x_embedder.weight": "diffusion_pytorch_model-00001-of-00003.safetensors"
|
| 468 |
+
}
|
| 469 |
+
}
|
transformer/transformer_moss_image.py
ADDED
|
@@ -0,0 +1,333 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Copyright 2026 OpenMOSS and The HuggingFace Team. All rights reserved.
|
| 2 |
+
#
|
| 3 |
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
| 4 |
+
# you may not use this file except in compliance with the License.
|
| 5 |
+
# You may obtain a copy of the License at
|
| 6 |
+
#
|
| 7 |
+
# http://www.apache.org/licenses/LICENSE-2.0
|
| 8 |
+
#
|
| 9 |
+
# Unless required by applicable law or agreed to in writing, software
|
| 10 |
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
| 11 |
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
| 12 |
+
# See the License for the specific language governing permissions and
|
| 13 |
+
# limitations under the License.
|
| 14 |
+
|
| 15 |
+
from typing import Any
|
| 16 |
+
|
| 17 |
+
import torch
|
| 18 |
+
import torch.nn as nn
|
| 19 |
+
|
| 20 |
+
from diffusers.configuration_utils import ConfigMixin, register_to_config
|
| 21 |
+
from diffusers.loaders import PeftAdapterMixin
|
| 22 |
+
from diffusers.models.attention import AttentionMixin, AttentionModuleMixin
|
| 23 |
+
from diffusers.models.attention_dispatch import dispatch_attention_fn
|
| 24 |
+
from diffusers.models.embeddings import TimestepEmbedding
|
| 25 |
+
from diffusers.models.modeling_outputs import Transformer2DModelOutput
|
| 26 |
+
from diffusers.models.modeling_utils import ModelMixin
|
| 27 |
+
from diffusers.models.normalization import RMSNorm
|
| 28 |
+
from diffusers.utils import apply_lora_scale
|
| 29 |
+
from diffusers.utils.torch_utils import maybe_adjust_dtype_for_device
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
def situ(hidden_states: torch.Tensor, gate: torch.Tensor) -> torch.Tensor:
|
| 33 |
+
"""Gated activation of the MOSS-Image feed-forward layers, with the gate and the value soft-clamped by tanh."""
|
| 34 |
+
return (4.0 * torch.tanh(gate / 4.0) * torch.sigmoid(gate)) * (25.0 * torch.tanh(hidden_states / 25.0))
|
| 35 |
+
|
| 36 |
+
|
| 37 |
+
def apply_rotary_emb(hidden_states: torch.Tensor, freqs_cos: torch.Tensor, freqs_sin: torch.Tensor) -> torch.Tensor:
|
| 38 |
+
"""Rotate-half RoPE, computed in the dtype of `hidden_states` like the original model."""
|
| 39 |
+
x1, x2 = hidden_states.chunk(2, dim=-1)
|
| 40 |
+
return hidden_states * freqs_cos + torch.cat((-x2, x1), dim=-1) * freqs_sin
|
| 41 |
+
|
| 42 |
+
|
| 43 |
+
class MOSSImageAttnProcessor:
|
| 44 |
+
_attention_backend = None
|
| 45 |
+
_parallel_config = None
|
| 46 |
+
|
| 47 |
+
def __call__(
|
| 48 |
+
self,
|
| 49 |
+
attn: "MOSSImageAttention",
|
| 50 |
+
hidden_states: torch.Tensor,
|
| 51 |
+
freqs_cos: torch.Tensor,
|
| 52 |
+
freqs_sin: torch.Tensor,
|
| 53 |
+
) -> torch.Tensor:
|
| 54 |
+
if attn.to_qkv is not None:
|
| 55 |
+
query, key, value = attn.to_qkv(hidden_states).chunk(3, dim=-1)
|
| 56 |
+
else:
|
| 57 |
+
query = attn.to_q(hidden_states)
|
| 58 |
+
key = attn.to_k(hidden_states)
|
| 59 |
+
value = attn.to_v(hidden_states)
|
| 60 |
+
|
| 61 |
+
query = attn.norm_q(query.unflatten(-1, (attn.heads, -1)))
|
| 62 |
+
key = attn.norm_k(key.unflatten(-1, (attn.heads, -1)))
|
| 63 |
+
value = value.unflatten(-1, (attn.heads, -1))
|
| 64 |
+
|
| 65 |
+
query = apply_rotary_emb(query, freqs_cos, freqs_sin)
|
| 66 |
+
key = apply_rotary_emb(key, freqs_cos, freqs_sin)
|
| 67 |
+
|
| 68 |
+
hidden_states = dispatch_attention_fn(
|
| 69 |
+
query.to(value.dtype),
|
| 70 |
+
key.to(value.dtype),
|
| 71 |
+
value,
|
| 72 |
+
backend=self._attention_backend,
|
| 73 |
+
parallel_config=self._parallel_config,
|
| 74 |
+
)
|
| 75 |
+
hidden_states = hidden_states.flatten(2, 3)
|
| 76 |
+
return attn.to_out(hidden_states)
|
| 77 |
+
|
| 78 |
+
|
| 79 |
+
class MOSSImageAttention(nn.Module, AttentionModuleMixin):
|
| 80 |
+
_default_processor_cls = MOSSImageAttnProcessor
|
| 81 |
+
_available_processors = [MOSSImageAttnProcessor]
|
| 82 |
+
|
| 83 |
+
def __init__(self, dim: int, heads: int, bias: bool, fused_qkv: bool, eps: float = 1e-6):
|
| 84 |
+
super().__init__()
|
| 85 |
+
self.heads = heads
|
| 86 |
+
head_dim = dim // heads
|
| 87 |
+
|
| 88 |
+
# The text refiner blocks use one fused, biased qkv projection; the main blocks use three unbiased ones.
|
| 89 |
+
if fused_qkv:
|
| 90 |
+
self.to_qkv = nn.Linear(dim, 3 * dim, bias=bias)
|
| 91 |
+
self.to_q = self.to_k = self.to_v = None
|
| 92 |
+
else:
|
| 93 |
+
self.to_qkv = None
|
| 94 |
+
self.to_q = nn.Linear(dim, dim, bias=bias)
|
| 95 |
+
self.to_k = nn.Linear(dim, dim, bias=bias)
|
| 96 |
+
self.to_v = nn.Linear(dim, dim, bias=bias)
|
| 97 |
+
self.norm_q = RMSNorm(head_dim, eps=eps)
|
| 98 |
+
self.norm_k = RMSNorm(head_dim, eps=eps)
|
| 99 |
+
self.to_out = nn.Linear(dim, dim, bias=bias)
|
| 100 |
+
|
| 101 |
+
self.set_processor(self._default_processor_cls())
|
| 102 |
+
|
| 103 |
+
def forward(self, hidden_states: torch.Tensor, freqs_cos: torch.Tensor, freqs_sin: torch.Tensor) -> torch.Tensor:
|
| 104 |
+
return self.processor(self, hidden_states, freqs_cos, freqs_sin)
|
| 105 |
+
|
| 106 |
+
|
| 107 |
+
class MOSSImageTextRefinerBlock(nn.Module):
|
| 108 |
+
"""Unmodulated transformer block that refines the text-encoder states before the joint blocks."""
|
| 109 |
+
|
| 110 |
+
def __init__(self, dim: int, ffn_dim: int, num_attention_heads: int, eps: float = 1e-6):
|
| 111 |
+
super().__init__()
|
| 112 |
+
self.norm1 = RMSNorm(dim, eps=eps)
|
| 113 |
+
self.attn = MOSSImageAttention(dim, num_attention_heads, bias=True, fused_qkv=True, eps=eps)
|
| 114 |
+
self.norm2 = RMSNorm(dim, eps=eps)
|
| 115 |
+
self.ff_gate = nn.Linear(dim, ffn_dim)
|
| 116 |
+
self.ff_up = nn.Linear(dim, ffn_dim)
|
| 117 |
+
self.ff_down = nn.Linear(ffn_dim, dim)
|
| 118 |
+
|
| 119 |
+
def forward(self, hidden_states: torch.Tensor, freqs_cos: torch.Tensor, freqs_sin: torch.Tensor) -> torch.Tensor:
|
| 120 |
+
hidden_states = hidden_states + self.attn(self.norm1(hidden_states), freqs_cos, freqs_sin)
|
| 121 |
+
norm_hidden_states = self.norm2(hidden_states)
|
| 122 |
+
ff_output = self.ff_down(situ(self.ff_up(norm_hidden_states), self.ff_gate(norm_hidden_states)))
|
| 123 |
+
return hidden_states + ff_output
|
| 124 |
+
|
| 125 |
+
|
| 126 |
+
class MOSSImageTransformerBlock(nn.Module):
|
| 127 |
+
"""Single-stream block over the concatenated image and text tokens, modulated by the shared timestep embedding."""
|
| 128 |
+
|
| 129 |
+
def __init__(self, dim: int, ffn_dim: int, num_attention_heads: int, eps: float = 1e-6):
|
| 130 |
+
super().__init__()
|
| 131 |
+
self.norm1 = RMSNorm(dim, eps=eps)
|
| 132 |
+
self.attn = MOSSImageAttention(dim, num_attention_heads, bias=False, fused_qkv=False, eps=eps)
|
| 133 |
+
self.norm2 = RMSNorm(dim, eps=eps)
|
| 134 |
+
self.ff_in = nn.Linear(dim, 2 * ffn_dim, bias=False)
|
| 135 |
+
self.ff_out = nn.Linear(ffn_dim, dim, bias=False)
|
| 136 |
+
|
| 137 |
+
def forward(
|
| 138 |
+
self,
|
| 139 |
+
hidden_states: torch.Tensor,
|
| 140 |
+
modulation: torch.Tensor,
|
| 141 |
+
freqs_cos: torch.Tensor,
|
| 142 |
+
freqs_sin: torch.Tensor,
|
| 143 |
+
) -> torch.Tensor:
|
| 144 |
+
shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = modulation.unbind(1)
|
| 145 |
+
|
| 146 |
+
norm_hidden_states = self.norm1(hidden_states) * (1 + scale_msa[:, None]) + shift_msa[:, None]
|
| 147 |
+
attn_output = self.attn(norm_hidden_states, freqs_cos, freqs_sin)
|
| 148 |
+
hidden_states = hidden_states + attn_output * gate_msa[:, None]
|
| 149 |
+
|
| 150 |
+
norm_hidden_states = self.norm2(hidden_states) * (1 + scale_mlp[:, None]) + shift_mlp[:, None]
|
| 151 |
+
ff_hidden_states, ff_gate = self.ff_in(norm_hidden_states).chunk(2, dim=-1)
|
| 152 |
+
ff_output = self.ff_out(situ(ff_hidden_states, ff_gate))
|
| 153 |
+
return hidden_states + ff_output * gate_mlp[:, None]
|
| 154 |
+
|
| 155 |
+
|
| 156 |
+
class MOSSImageRotaryPosEmbed(nn.Module):
|
| 157 |
+
"""4-axis RoPE over `(unused, frame, height, width)` coordinates.
|
| 158 |
+
|
| 159 |
+
Image tokens sit at frame `text_seq_len` on a grid centered at zero; text token `i` sits at frame `i`. Each axis
|
| 160 |
+
uses `head_dim // 8` frequencies, built in float32 on the CPU like the original model.
|
| 161 |
+
"""
|
| 162 |
+
|
| 163 |
+
def __init__(self, head_dim: int, theta: float):
|
| 164 |
+
super().__init__()
|
| 165 |
+
self.num_bands = head_dim // 8
|
| 166 |
+
self.theta = theta
|
| 167 |
+
|
| 168 |
+
def forward(self, height: int, width: int, text_seq_len: int, device: torch.device):
|
| 169 |
+
exponent = torch.arange(self.num_bands, dtype=torch.float32) / self.num_bands
|
| 170 |
+
bands = (1 / (self.theta**exponent)).to(device)
|
| 171 |
+
|
| 172 |
+
grid_h, grid_w = torch.meshgrid(
|
| 173 |
+
torch.arange(height, device=device), torch.arange(width, device=device), indexing="ij"
|
| 174 |
+
)
|
| 175 |
+
image_ids = torch.stack(
|
| 176 |
+
(torch.zeros_like(grid_h), torch.full_like(grid_h, text_seq_len), grid_h, grid_w), dim=-1
|
| 177 |
+
)
|
| 178 |
+
image_ids = image_ids.flatten(0, 1).float()
|
| 179 |
+
image_ids[:, 2] -= (height - 1) / 2
|
| 180 |
+
image_ids[:, 3] -= (width - 1) / 2
|
| 181 |
+
text_ids = torch.zeros(text_seq_len, 4, device=device, dtype=torch.float32)
|
| 182 |
+
text_ids[:, 1] = torch.arange(text_seq_len, device=device)
|
| 183 |
+
|
| 184 |
+
phase = (torch.cat((image_ids, text_ids), dim=0).unsqueeze(-1) * bands).flatten(1)
|
| 185 |
+
phase = torch.cat((phase, phase), dim=-1)
|
| 186 |
+
# (1, seq_len, 1, head_dim), broadcast over the attention heads
|
| 187 |
+
return phase.cos()[None, :, None], phase.sin()[None, :, None]
|
| 188 |
+
|
| 189 |
+
|
| 190 |
+
class MOSSImageTransformer2DModel(ModelMixin, ConfigMixin, PeftAdapterMixin, AttentionMixin):
|
| 191 |
+
r"""
|
| 192 |
+
The MOSS-Image diffusion transformer.
|
| 193 |
+
|
| 194 |
+
Image tokens are the FLUX.2 VAE latents (patchified to 128 channels at 1/16 of the image resolution). The text
|
| 195 |
+
states pass through `num_refiner_layers` unmodulated blocks, then image and text tokens go through `num_layers`
|
| 196 |
+
single-stream blocks, all modulated by one projection of the timestep embedding. Each sample runs at its own text
|
| 197 |
+
length, so prompts are never padded.
|
| 198 |
+
|
| 199 |
+
Args:
|
| 200 |
+
in_channels (`int`, defaults to `128`): Channels of the patchified latents.
|
| 201 |
+
num_layers (`int`, defaults to `42`): Number of single-stream transformer blocks.
|
| 202 |
+
num_refiner_layers (`int`, defaults to `2`): Number of text refiner blocks.
|
| 203 |
+
num_attention_heads (`int`, defaults to `24`): Number of attention heads.
|
| 204 |
+
attention_head_dim (`int`, defaults to `128`): Dimension of each attention head.
|
| 205 |
+
ffn_dim (`int`, defaults to `8192`): Hidden dimension of the feed-forward layers.
|
| 206 |
+
text_embed_dim (`int`, defaults to `4096`): Dimension of the text-encoder states.
|
| 207 |
+
timestep_embed_dim (`int`, defaults to `256`): Dimension of the sinusoidal timestep embedding.
|
| 208 |
+
rope_theta (`float`, defaults to `10000.0`): Base of the RoPE frequencies.
|
| 209 |
+
"""
|
| 210 |
+
|
| 211 |
+
_supports_gradient_checkpointing = True
|
| 212 |
+
_no_split_modules = ["MOSSImageTransformerBlock", "MOSSImageTextRefinerBlock"]
|
| 213 |
+
_repeated_blocks = ["MOSSImageTransformerBlock", "MOSSImageTextRefinerBlock"]
|
| 214 |
+
_skip_layerwise_casting_patterns = ["norm"]
|
| 215 |
+
|
| 216 |
+
@register_to_config
|
| 217 |
+
def __init__(
|
| 218 |
+
self,
|
| 219 |
+
in_channels: int = 128,
|
| 220 |
+
num_layers: int = 42,
|
| 221 |
+
num_refiner_layers: int = 2,
|
| 222 |
+
num_attention_heads: int = 24,
|
| 223 |
+
attention_head_dim: int = 128,
|
| 224 |
+
ffn_dim: int = 8192,
|
| 225 |
+
text_embed_dim: int = 4096,
|
| 226 |
+
timestep_embed_dim: int = 256,
|
| 227 |
+
rope_theta: float = 10000.0,
|
| 228 |
+
):
|
| 229 |
+
super().__init__()
|
| 230 |
+
inner_dim = num_attention_heads * attention_head_dim
|
| 231 |
+
|
| 232 |
+
self.rope = MOSSImageRotaryPosEmbed(attention_head_dim, rope_theta)
|
| 233 |
+
self.x_embedder = nn.Linear(in_channels, inner_dim)
|
| 234 |
+
self.context_embedder = nn.Linear(text_embed_dim, inner_dim)
|
| 235 |
+
self.context_refiner = nn.ModuleList(
|
| 236 |
+
[MOSSImageTextRefinerBlock(inner_dim, ffn_dim, num_attention_heads) for _ in range(num_refiner_layers)]
|
| 237 |
+
)
|
| 238 |
+
self.time_embedder = TimestepEmbedding(timestep_embed_dim, inner_dim)
|
| 239 |
+
self.block_modulation = nn.Linear(inner_dim, 6 * inner_dim)
|
| 240 |
+
self.transformer_blocks = nn.ModuleList(
|
| 241 |
+
[MOSSImageTransformerBlock(inner_dim, ffn_dim, num_attention_heads) for _ in range(num_layers)]
|
| 242 |
+
)
|
| 243 |
+
self.norm_out = RMSNorm(inner_dim, eps=1e-6)
|
| 244 |
+
self.norm_out_modulation = nn.Linear(inner_dim, 2 * inner_dim)
|
| 245 |
+
self.proj_out = nn.Linear(inner_dim, in_channels, bias=False)
|
| 246 |
+
|
| 247 |
+
self.gradient_checkpointing = False
|
| 248 |
+
|
| 249 |
+
def timestep_embedding(self, timestep: torch.Tensor) -> torch.Tensor:
|
| 250 |
+
# cos first, then sin; float64 where the device supports it, like the original model
|
| 251 |
+
half = self.config.timestep_embed_dim // 2
|
| 252 |
+
freqs_dtype = maybe_adjust_dtype_for_device(torch.float64, timestep.device)
|
| 253 |
+
freqs = 10000 ** (-torch.arange(half, device=timestep.device, dtype=freqs_dtype) / half)
|
| 254 |
+
phase = timestep.to(freqs_dtype)[:, None] * freqs
|
| 255 |
+
return torch.cat((phase.cos(), phase.sin()), dim=-1).float()
|
| 256 |
+
|
| 257 |
+
@apply_lora_scale("attention_kwargs")
|
| 258 |
+
def forward(
|
| 259 |
+
self,
|
| 260 |
+
hidden_states: torch.Tensor,
|
| 261 |
+
timestep: torch.Tensor,
|
| 262 |
+
encoder_hidden_states: list[torch.Tensor],
|
| 263 |
+
skip_layers: list[int] | None = None,
|
| 264 |
+
attention_kwargs: dict[str, Any] | None = None,
|
| 265 |
+
return_dict: bool = True,
|
| 266 |
+
) -> torch.Tensor | Transformer2DModelOutput:
|
| 267 |
+
r"""
|
| 268 |
+
Args:
|
| 269 |
+
hidden_states (`torch.Tensor`):
|
| 270 |
+
Patchified latents of shape `(batch_size, in_channels, height, width)`.
|
| 271 |
+
timestep (`torch.Tensor`):
|
| 272 |
+
Timesteps on the `[0, 1000]` scale, a scalar or one per sample.
|
| 273 |
+
encoder_hidden_states (`list[torch.Tensor]`):
|
| 274 |
+
Unpadded text-encoder states, one `(text_seq_len, text_embed_dim)` tensor per sample.
|
| 275 |
+
skip_layers (`list[int]`, *optional*):
|
| 276 |
+
Indices of the transformer blocks to skip.
|
| 277 |
+
attention_kwargs (`dict`, *optional*):
|
| 278 |
+
Extra arguments; `scale` sets the LoRA scale.
|
| 279 |
+
return_dict (`bool`, defaults to `True`):
|
| 280 |
+
Whether to return a [`~models.modeling_outputs.Transformer2DModelOutput`] instead of a tuple.
|
| 281 |
+
|
| 282 |
+
Returns:
|
| 283 |
+
[`~models.modeling_outputs.Transformer2DModelOutput`] or `tuple`: The predicted flow of shape
|
| 284 |
+
`(batch_size, in_channels, height, width)`, as `sample` or as the first element of the tuple.
|
| 285 |
+
"""
|
| 286 |
+
batch_size, _, height, width = hidden_states.shape
|
| 287 |
+
if len(encoder_hidden_states) != batch_size:
|
| 288 |
+
raise ValueError(
|
| 289 |
+
f"`encoder_hidden_states` holds {len(encoder_hidden_states)} tensors for a batch of {batch_size}"
|
| 290 |
+
)
|
| 291 |
+
device = hidden_states.device
|
| 292 |
+
image_seq_len = height * width
|
| 293 |
+
|
| 294 |
+
timestep = torch.as_tensor(timestep, device=device).reshape(-1).expand(batch_size)
|
| 295 |
+
temb = self.time_embedder(self.timestep_embedding(timestep).to(hidden_states.dtype))
|
| 296 |
+
modulation = self.block_modulation(temb).view(batch_size, 6, -1)
|
| 297 |
+
shift_out, scale_out = self.norm_out_modulation(temb).view(batch_size, 2, -1).unbind(1)
|
| 298 |
+
|
| 299 |
+
# Image positions depend on each sample's text length, so samples run one at a time
|
| 300 |
+
outputs = []
|
| 301 |
+
for index in range(batch_size):
|
| 302 |
+
text_seq_len = encoder_hidden_states[index].shape[0]
|
| 303 |
+
freqs_cos, freqs_sin = self.rope(height, width, text_seq_len, device)
|
| 304 |
+
freqs_cos, freqs_sin = freqs_cos.to(hidden_states.dtype), freqs_sin.to(hidden_states.dtype)
|
| 305 |
+
text_freqs = (freqs_cos[:, image_seq_len:], freqs_sin[:, image_seq_len:])
|
| 306 |
+
|
| 307 |
+
image_tokens = self.x_embedder(hidden_states[index : index + 1].flatten(2).transpose(1, 2).contiguous())
|
| 308 |
+
text_tokens = self.context_embedder(encoder_hidden_states[index][None])
|
| 309 |
+
for block in self.context_refiner:
|
| 310 |
+
if torch.is_grad_enabled() and self.gradient_checkpointing:
|
| 311 |
+
text_tokens = self._gradient_checkpointing_func(block, text_tokens, *text_freqs)
|
| 312 |
+
else:
|
| 313 |
+
text_tokens = block(text_tokens, *text_freqs)
|
| 314 |
+
|
| 315 |
+
tokens = torch.cat((image_tokens, text_tokens), dim=1)
|
| 316 |
+
sample_modulation = modulation[index : index + 1]
|
| 317 |
+
for block_index, block in enumerate(self.transformer_blocks):
|
| 318 |
+
if skip_layers is not None and block_index in skip_layers:
|
| 319 |
+
continue
|
| 320 |
+
if torch.is_grad_enabled() and self.gradient_checkpointing:
|
| 321 |
+
tokens = self._gradient_checkpointing_func(block, tokens, sample_modulation, freqs_cos, freqs_sin)
|
| 322 |
+
else:
|
| 323 |
+
tokens = block(tokens, sample_modulation, freqs_cos, freqs_sin)
|
| 324 |
+
|
| 325 |
+
image_tokens = self.norm_out(tokens[:, :image_seq_len])
|
| 326 |
+
image_tokens = image_tokens * (1 + scale_out[index : index + 1, None]) + shift_out[index : index + 1, None]
|
| 327 |
+
image_tokens = self.proj_out(image_tokens)
|
| 328 |
+
outputs.append(image_tokens.transpose(1, 2).reshape(1, -1, height, width))
|
| 329 |
+
|
| 330 |
+
output = torch.cat(outputs, dim=0)
|
| 331 |
+
if not return_dict:
|
| 332 |
+
return (output,)
|
| 333 |
+
return Transformer2DModelOutput(sample=output)
|
vae/config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_class_name": "AutoencoderKLFlux2",
|
| 3 |
+
"_diffusers_version": "0.38.0",
|
| 4 |
+
"act_fn": "silu",
|
| 5 |
+
"batch_norm_eps": 0.0001,
|
| 6 |
+
"batch_norm_momentum": 0.1,
|
| 7 |
+
"block_out_channels": [
|
| 8 |
+
128,
|
| 9 |
+
256,
|
| 10 |
+
512,
|
| 11 |
+
512
|
| 12 |
+
],
|
| 13 |
+
"decoder_block_out_channels": [
|
| 14 |
+
96,
|
| 15 |
+
192,
|
| 16 |
+
384,
|
| 17 |
+
384
|
| 18 |
+
],
|
| 19 |
+
"down_block_types": [
|
| 20 |
+
"DownEncoderBlock2D",
|
| 21 |
+
"DownEncoderBlock2D",
|
| 22 |
+
"DownEncoderBlock2D",
|
| 23 |
+
"DownEncoderBlock2D"
|
| 24 |
+
],
|
| 25 |
+
"force_upcast": true,
|
| 26 |
+
"in_channels": 3,
|
| 27 |
+
"latent_channels": 32,
|
| 28 |
+
"layers_per_block": 2,
|
| 29 |
+
"mid_block_add_attention": true,
|
| 30 |
+
"norm_num_groups": 32,
|
| 31 |
+
"out_channels": 3,
|
| 32 |
+
"patch_size": [
|
| 33 |
+
2,
|
| 34 |
+
2
|
| 35 |
+
],
|
| 36 |
+
"sample_size": 1024,
|
| 37 |
+
"up_block_types": [
|
| 38 |
+
"UpDecoderBlock2D",
|
| 39 |
+
"UpDecoderBlock2D",
|
| 40 |
+
"UpDecoderBlock2D",
|
| 41 |
+
"UpDecoderBlock2D"
|
| 42 |
+
],
|
| 43 |
+
"use_post_quant_conv": true,
|
| 44 |
+
"use_quant_conv": true
|
| 45 |
+
}
|
vae/diffusion_pytorch_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3bdf94dc67ec69b30ea9fe1fa8340085679cae64d019f7c7e4a9921f1a7f17a
|
| 3 |
+
size 249521372
|