Add files using upload-large-folder tool

LICENSE

@@ -0,0 +1,204 @@
+Copyright 2026 The NitroFusion2 Authors.
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2026 The NitroFusion2 Authors.
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.
+

NOTICE

@@ -0,0 +1,6 @@
+NitroFusion2
+Copyright 2026 The NitroFusion2 Authors.
+Licensed under the Apache License, Version 2.0.
+
+This repository contains the noise_refiner and context_refiner tensors extracted
+without modification from ChenDY/NitroFusion2 (transformer/ shards).

README.md

@@ -0,0 +1,107 @@
+---
+license: apache-2.0
+library_name: diffusers
+pipeline_tag: text-to-image
+tags:
+- dit
+- z-image
+- mmdit
+- refiner
+base_model: ChenDY/NitroFusion2
+---
+
+# NitroFusion2 DiT Refiners
+
+Community extraction of the two refiner stacks from [NitroFusion2](https://huggingface.co/ChenDY/NitroFusion2), revision `62d1a7000af9c46a0c4348d936bd7670ff33c1b9`.
+
+NitroFusion2 is a Z-Image-Turbo derived MMDiT. Before the 30-layer trunk runs, the model passes the noisy latent patches through a **noise refiner** and the caption features through a **context refiner**. This repository contains those two stacks and nothing else: **56 tensors, 715,746,304 parameters**, in the original `bfloat16`.
+
+| Stack | Layers | Tensors | Parameters | Modulation |
+| --- | --- | --- | --- | --- |
+| `noise_refiner` | 2 | 30 | 361,820,672 | adaLN from the timestep embedding, `tanh` gates |
+| `context_refiner` | 2 | 26 | 353,925,632 | none |
+
+Each block is a single-stream transformer: RMSNorm-preceded attention with per-head RMSNorm on q/k (head dim 128), a SwiGLU feed-forward (3840 → 10240 → 3840), and post-norm on both residual branches. The noise refiner additionally carries `adaLN_modulation`, a `Linear(256 → 15360)` that produces shift and gate parameters for attention and MLP.
+
+The 30-layer trunk (`layers.*`, 10.85 GB), the embedders, the final layer and the pad tokens are **not** included. This is a component, not a generation pipeline: it cannot produce images on its own.
+
+## Provenance
+
+| Item | Value |
+| --- | --- |
+| Source repository | `ChenDY/NitroFusion2` |
+| Source revision | `62d1a7000af9c46a0c4348d936bd7670ff33c1b9` |
+| Source file | `transformer/diffusion_pytorch_model-00001-of-00003.safetensors` |
+| Extracted scope | `noise_refiner.*` and `context_refiner.*` |
+| Tensor count | 56 |
+| Tensor bytes | 1,431,492,608 |
+| Output `model.safetensors` SHA-256 | `c871ae336b02e763a4be2018231a1b616082fd4edb04a51189b2798dcfb62284` |
+| Extraction method | HTTP range requests over two contiguous byte ranges; no full shard download |
+
+Both stacks occupy a single contiguous byte range each in the first shard, so only 1.43 GB of the 12.31 GB transformer was transferred. Every assembled range was compared byte-for-byte with the cached source ranges before the output file was hashed. Tensor values, shapes and dtypes are unchanged: no training, quantization, dtype conversion or reordering.
+
+Original checkpoint tensor names are preserved, so `load_state_dict(strict=True)` works with **no prefix stripping**. `tensor_mapping.json` records the source shard, dtype, shape and byte offsets of every tensor.
+
+## Loading
+
+The source configuration declares `diffusers 0.36.0.dev0`. `ZImageTransformerBlock` is not present in 0.35.x. Install:
+
+```bash
+pip install "diffusers==0.36.0" torch accelerate safetensors
+```
+
+```python
+import sys, torch
+from huggingface_hub import snapshot_download
+
+root = snapshot_download('gatilin/nitrofusion2-dit-refiners')
+sys.path.insert(0, root)
+from component import load_model
+
+model = load_model(root)          # original bfloat16 weights
+
+# x: [batch, seq, 3840] embedded noisy latent patches (seq must be a multiple of 32)
+# attn_mask: [batch, seq] bool, True for real tokens
+# position_ids: [seq, 3] int, from the upstream create_coordinate_grid layout
+# adaln_input: [batch, 256] timestep embedding
+with torch.inference_mode():
+    refined = model.forward_noise(x, attn_mask, position_ids, adaln_input)
+
+# cap_feats: [batch, cap_seq, 3840] embedded caption features
+with torch.inference_mode():
+    refined_cap = model.forward_context(cap_feats, attn_mask, position_ids)
+```
+
+`load_model(root, dtype=torch.float32)` runs the same weights in float32.
+
+**Input adapters are not included.** `forward_noise` expects features already passed through the source model's `all_x_embedder` (`Linear(64 → 3840)` on 2×2 patches of 16-channel latents) and adaLN conditioning from `t_embedder`; `forward_context` expects features from `cap_embedder` (`RMSNorm(2560)` then `Linear(2560 → 3840)`). Take those from the source repository. The RoPE helper is included as `model.rope_embedder`, configured with the source `axes_dims` `[32, 48, 48]`, `axes_lens` `[1536, 512, 512]` and `rope_theta` 256.
+
+Position ids follow the upstream layout: caption tokens occupy `(1..cap_len, 0, 0)`, and image tokens start at `(cap_len + 1, 0, 0)` — the image grid is offset by the padded caption length.
+
+## Validation
+
+Measured on CPU with `diffusers==0.36.0` and `torch==2.14.1`. Full results are in `validation.json`.
+
+| Check | Result |
+| --- | --- |
+| Strict load | 56/56 tensors, no missing or unexpected keys, no prefix stripping |
+| Parameters | 715,746,304, all `torch.bfloat16` |
+| `noise_refiner` input | three real COCO128 images → authentic Z-Image VAE → latent `[1, 16, 32, 32]` → 2×2 patches `[256, 64]` |
+| `noise_refiner` output | `[1, 256, 3840]`, all finite on all three images |
+| `context_refiner` output | `[1, 64, 3840]`, all finite |
+| Determinism | repeated identical input, max absolute error exactly `0.0` on both stacks |
+
+The `noise_refiner` path uses **real image data end to end**. The VAE is the local Flux VAE, verified byte-identical to `Tongyi-MAI/Z-Image-Turbo`'s `vae/diffusion_pytorch_model.safetensors` (SHA-256 `f5b59a26851551b67ae1fe58d32e76486e1e812def4696a4bea97f16604d40a3`). Latents are scaled with the source `shift_factor` 0.1159 and `scaling_factor` 0.3611, noised at flow-matching `t = 0.5`, then passed through the authentic `all_x_embedder` and `t_embedder` extracted from the same pinned revision. RoPE uses the upstream `create_coordinate_grid` layout.
+
+**Two documented assumptions.** The official Qwen3 text encoder was not downloaded, so (1) `context_refiner` receives seeded synthetic caption features rather than real text-encoder output — they still pass through the authentic `cap_embedder`; and (2) the padded caption length is fixed at 64, which sets the offset of the image position ids. Neither stack is claimed to reproduce full-pipeline outputs.
+
+## Limitations
+
+- This is a component-level extraction. Without the trunk, final layer, VAE and text encoder there is no image generation capability here.
+- The refiners are designed to run before the trunk; their outputs are intermediate features, not images or final latents.
+- `bfloat16` CPU execution is not claimed to be bit-identical to the original GPU inference path.
+- COCO128 images used for validation may overlap the model's training data; they are not an independent generalization benchmark.
+
+## License
+
+Apache License 2.0, inherited from the source repository. `LICENSE` is the upstream Apache 2.0 text and `NOTICE` records attribution. No official source file was modified; `component.py`, `component_config.json`, `extraction.json`, `tensor_mapping.json`, `validation.json` and this README are added by the extractor.

component.py

@@ -0,0 +1,45 @@
+"""Load the NitroFusion2 DiT refiner stacks using the official diffusers Z-Image blocks."""
+from pathlib import Path
+import json
+import torch
+from torch import nn
+from accelerate import init_empty_weights
+from safetensors.torch import load_file
+from diffusers.models.transformers.transformer_z_image import ZImageTransformerBlock, RopeEmbedder
+
+
+class NitroFusion2Refiners(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        dim = config['dim']
+        args = (dim, config['n_heads'], config['n_kv_heads'], config['norm_eps'], config['qk_norm'])
+        n = config['n_refiner_layers']
+        self.noise_refiner = nn.ModuleList([
+            ZImageTransformerBlock(1000 + i, *args, modulation=True) for i in range(n)])
+        self.context_refiner = nn.ModuleList([
+            ZImageTransformerBlock(i, *args, modulation=False) for i in range(n)])
+        self.rope_embedder = RopeEmbedder(
+            theta=config['rope_theta'], axes_dims=config['axes_dims'], axes_lens=config['axes_lens'])
+
+    def forward_noise(self, x, attn_mask, position_ids, adaln_input):
+        freqs_cis = self.rope_embedder(position_ids).unsqueeze(0)
+        for layer in self.noise_refiner:
+            x = layer(x, attn_mask, freqs_cis, adaln_input)
+        return x
+
+    def forward_context(self, cap_feats, attn_mask, position_ids):
+        freqs_cis = self.rope_embedder(position_ids).unsqueeze(0)
+        for layer in self.context_refiner:
+            cap_feats = layer(cap_feats, attn_mask, freqs_cis)
+        return cap_feats
+
+
+def load_model(root, dtype=None):
+    root = Path(root)
+    config = json.loads((root / 'config.json').read_text())
+    with init_empty_weights(include_buffers=False):
+        model = NitroFusion2Refiners(config)
+    state = load_file(str(root / 'model.safetensors'), device='cpu')
+    model.load_state_dict(state, strict=True, assign=True)
+    model = model.eval()
+    return model if dtype is None else model.to(dtype)

component_config.json

@@ -0,0 +1,14 @@
+{
+  "kind": "dit-refiner",
+  "source_repo": "ChenDY/NitroFusion2",
+  "source_revision": "62d1a7000af9c46a0c4348d936bd7670ff33c1b9",
+  "prefixes": ["noise_refiner.", "context_refiner."],
+  "block_class": "ZImageTransformerBlock",
+  "block_module": "diffusers.models.transformers.transformer_z_image",
+  "source_declared_diffusers_version": "0.36.0.dev0",
+  "validated_diffusers_version": "0.36.0",
+  "naming": "original checkpoint tensor names preserved; load_state_dict(strict=True) with no prefix stripping",
+  "excluded": "layers.* (30-layer MMDiT trunk), all_x_embedder.*, all_final_layer.*, cap_embedder.*, t_embedder.*, x_pad_token, cap_pad_token",
+  "input_adapters_required": "noise_refiner expects embedded noisy latent patches from all_x_embedder plus adaLN conditioning from t_embedder; context_refiner expects caption features from cap_embedder. Those adapters are not included in this component and must be taken from the source repository.",
+  "license": "Apache-2.0"
+}

config.json

@@ -0,0 +1,31 @@
+{
+  "_class_name": "ZImageTransformer2DModel",
+  "_diffusers_version": "0.36.0.dev0",
+  "all_f_patch_size": [
+    1
+  ],
+  "all_patch_size": [
+    2
+  ],
+  "axes_dims": [
+    32,
+    48,
+    48
+  ],
+  "axes_lens": [
+    1536,
+    512,
+    512
+  ],
+  "cap_feat_dim": 2560,
+  "dim": 3840,
+  "in_channels": 16,
+  "n_heads": 30,
+  "n_kv_heads": 30,
+  "n_layers": 30,
+  "n_refiner_layers": 2,
+  "norm_eps": 1e-05,
+  "qk_norm": true,
+  "rope_theta": 256.0,
+  "t_scale": 1000.0
+}

extraction.json

@@ -0,0 +1,15 @@
+{
+  "source_repo": "ChenDY/NitroFusion2",
+  "source_revision": "62d1a7000af9c46a0c4348d936bd7670ff33c1b9",
+  "source_shards": [
+    "diffusion_pytorch_model-00001-of-00003.safetensors"
+  ],
+  "scope": "noise_refiner.* and context_refiner.* (the two 2-layer refiner stacks of the Z-Image MMDiT; the 30-layer trunk, embedders and final layer are excluded)",
+  "naming": "original checkpoint tensor names preserved",
+  "tensor_count": 56,
+  "tensor_bytes": 1431492608,
+  "parameters": 715746304,
+  "sha256": "c871ae336b02e763a4be2018231a1b616082fd4edb04a51189b2798dcfb62284",
+  "source_range_bytes_verified": true,
+  "license": "Apache-2.0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c871ae336b02e763a4be2018231a1b616082fd4edb04a51189b2798dcfb62284
+size 1431498952

tensor_mapping.json

@@ -0,0 +1,648 @@
+{
+  "context_refiner.0.attention.norm_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      22645888,
+      22646144
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "context_refiner.0.attention.norm_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      22646144,
+      22646400
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "context_refiner.0.attention.to_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      22646400,
+      52137600
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.0.attention.to_out.0.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      52137600,
+      81628800
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.0.attention.to_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      81628800,
+      111120000
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.0.attention.to_v.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      111120000,
+      140611200
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.0.attention_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      140611200,
+      140618880
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "context_refiner.0.attention_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      140618880,
+      140626560
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "context_refiner.0.feed_forward.w1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      140626560,
+      219269760
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "context_refiner.0.feed_forward.w2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      219269760,
+      297912960
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      10240
+    ]
+  },
+  "context_refiner.0.feed_forward.w3.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      297912960,
+      376556160
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "context_refiner.0.ffn_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      376556160,
+      376563840
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "context_refiner.0.ffn_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      376563840,
+      376571520
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "context_refiner.1.attention.norm_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      376571520,
+      376571776
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "context_refiner.1.attention.norm_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      376571776,
+      376572032
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "context_refiner.1.attention.to_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      376572032,
+      406063232
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.1.attention.to_out.0.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      406063232,
+      435554432
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.1.attention.to_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      435554432,
+      465045632
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.1.attention.to_v.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      465045632,
+      494536832
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "context_refiner.1.attention_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      494536832,
+      494544512
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "context_refiner.1.attention_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      494544512,
+      494552192
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "context_refiner.1.feed_forward.w1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      494552192,
+      573195392
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "context_refiner.1.feed_forward.w2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      573195392,
+      651838592
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      10240
+    ]
+  },
+  "context_refiner.1.feed_forward.w3.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      651838592,
+      730481792
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "context_refiner.1.ffn_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      730481792,
+      730489472
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "context_refiner.1.ffn_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      730489472,
+      730497152
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.0.adaLN_modulation.0.bias": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4262134912,
+      4262165632
+    ],
+    "dtype": "BF16",
+    "shape": [
+      15360
+    ]
+  },
+  "noise_refiner.0.adaLN_modulation.0.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4262165632,
+      4270029952
+    ],
+    "dtype": "BF16",
+    "shape": [
+      15360,
+      256
+    ]
+  },
+  "noise_refiner.0.attention.norm_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4270029952,
+      4270030208
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "noise_refiner.0.attention.norm_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4270030208,
+      4270030464
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "noise_refiner.0.attention.to_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4270030464,
+      4299521664
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.0.attention.to_out.0.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4299521664,
+      4329012864
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.0.attention.to_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4329012864,
+      4358504064
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.0.attention.to_v.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4358504064,
+      4387995264
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.0.attention_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4387995264,
+      4388002944
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.0.attention_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4388002944,
+      4388010624
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.0.feed_forward.w1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4388010624,
+      4466653824
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "noise_refiner.0.feed_forward.w2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4466653824,
+      4545297024
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      10240
+    ]
+  },
+  "noise_refiner.0.feed_forward.w3.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4545297024,
+      4623940224
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "noise_refiner.0.ffn_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4623940224,
+      4623947904
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.0.ffn_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4623947904,
+      4623955584
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.1.adaLN_modulation.0.bias": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4623955584,
+      4623986304
+    ],
+    "dtype": "BF16",
+    "shape": [
+      15360
+    ]
+  },
+  "noise_refiner.1.adaLN_modulation.0.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4623986304,
+      4631850624
+    ],
+    "dtype": "BF16",
+    "shape": [
+      15360,
+      256
+    ]
+  },
+  "noise_refiner.1.attention.norm_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4631850624,
+      4631850880
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "noise_refiner.1.attention.norm_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4631850880,
+      4631851136
+    ],
+    "dtype": "BF16",
+    "shape": [
+      128
+    ]
+  },
+  "noise_refiner.1.attention.to_k.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4631851136,
+      4661342336
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.1.attention.to_out.0.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4661342336,
+      4690833536
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.1.attention.to_q.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4690833536,
+      4720324736
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.1.attention.to_v.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4720324736,
+      4749815936
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      3840
+    ]
+  },
+  "noise_refiner.1.attention_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4749815936,
+      4749823616
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.1.attention_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4749823616,
+      4749831296
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.1.feed_forward.w1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4749831296,
+      4828474496
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "noise_refiner.1.feed_forward.w2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4828474496,
+      4907117696
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840,
+      10240
+    ]
+  },
+  "noise_refiner.1.feed_forward.w3.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4907117696,
+      4985760896
+    ],
+    "dtype": "BF16",
+    "shape": [
+      10240,
+      3840
+    ]
+  },
+  "noise_refiner.1.ffn_norm1.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4985760896,
+      4985768576
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  },
+  "noise_refiner.1.ffn_norm2.weight": {
+    "source_shard": "diffusion_pytorch_model-00001-of-00003.safetensors",
+    "source_data_offsets": [
+      4985768576,
+      4985776256
+    ],
+    "dtype": "BF16",
+    "shape": [
+      3840
+    ]
+  }
+}

validation.json

@@ -0,0 +1,114 @@
+{
+  "strict_load": true,
+  "finite": true,
+  "diffusers_version": "0.36.0",
+  "source_declared_diffusers_version": "0.36.0.dev0",
+  "torch_version": "2.14.1",
+  "device": "cpu",
+  "dtype": [
+    "torch.bfloat16"
+  ],
+  "parameters": 715746304,
+  "tensor_count": 56,
+  "noise_refiner_validation": {
+    "input": "three real COCO128 images encoded by the authentic Z-Image VAE, patchified and embedded with the authentic all_x_embedder, noised at flow-matching t=0.5, real t_embedder adaLN conditioning and upstream create_coordinate_grid RoPE",
+    "vae_sha256": "f5b59a26851551b67ae1fe58d32e76486e1e812def4696a4bea97f16604d40a3",
+    "vae_identical_to_z_image_turbo": true,
+    "assumed_caption_length": 64,
+    "assumption_note": "image position ids are offset by the padded caption length upstream; the official Qwen3 text encoder is out of scope, so a fixed 64-token caption length is assumed",
+    "per_image": [
+      {
+        "image": "000000000009.png",
+        "source_size": [
+          256,
+          256
+        ],
+        "latent_shape": [
+          1,
+          16,
+          32,
+          32
+        ],
+        "patch_shape": [
+          256,
+          64
+        ],
+        "noise_refiner_output_shape": [
+          1,
+          256,
+          3840
+        ],
+        "timestep": 0.5,
+        "finite": true,
+        "repeat_max_abs_error": 0.0,
+        "output_abs_mean": 0.1828508973121643
+      },
+      {
+        "image": "000000000030.png",
+        "source_size": [
+          256,
+          256
+        ],
+        "latent_shape": [
+          1,
+          16,
+          32,
+          32
+        ],
+        "patch_shape": [
+          256,
+          64
+        ],
+        "noise_refiner_output_shape": [
+          1,
+          256,
+          3840
+        ],
+        "timestep": 0.5,
+        "finite": true,
+        "repeat_max_abs_error": 0.0,
+        "output_abs_mean": 0.17951683700084686
+      },
+      {
+        "image": "000000000034.png",
+        "source_size": [
+          256,
+          256
+        ],
+        "latent_shape": [
+          1,
+          16,
+          32,
+          32
+        ],
+        "patch_shape": [
+          256,
+          64
+        ],
+        "noise_refiner_output_shape": [
+          1,
+          256,
+          3840
+        ],
+        "timestep": 0.5,
+        "finite": true,
+        "repeat_max_abs_error": 0.0,
+        "output_abs_mean": 0.16232071816921234
+      }
+    ]
+  },
+  "context_refiner_validation": {
+    "input": "synthetic Qwen3-shaped caption features (seeded), passed through the authentic cap_embedder",
+    "cap_len": 64,
+    "output_shape": [
+      1,
+      64,
+      3840
+    ],
+    "finite": true,
+    "repeat_max_abs_error": 0.0,
+    "output_abs_mean": 0.5141564011573792
+  },
+  "elapsed_seconds": 20.54,
+  "limitations": "Component-level CPU test only. The 30-layer DiT trunk, final layer, VAE decoder and text encoder are excluded, so this does not evaluate image generation quality or end-to-end parity with the full NitroFusion2 pipeline. context_refiner inputs are synthetic because the official Qwen3 text encoder was not downloaded. bfloat16 CPU execution is not claimed to match GPU inference. COCO128 may overlap training data."
+}