Upload folder using huggingface_hub

README.md

@@ -0,0 +1,282 @@
+---
+license: apache-2.0
+base_model:
+  - ByteDance-Seed/BAGEL-7B-MoT
+base_model_relation: finetune
+library_name: transformers
+pipeline_tag: image-to-image
+tags:
+  - bagel
+  - image-editing
+  - image-generation
+  - interleaved-generation
+  - safetensors
+  - vbvr-pro
+---
+
+# VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning
+
+<a href="https://video-reason.com/?v=pro" target="_blank">
+    <img alt="Project Page" src="https://img.shields.io/badge/Project%20-%20Homepage-4285F4" height="20" />
+</a>
+<a href="https://github.com/Video-Reason/VBVR-Pro-Bench" target="_blank">
+    <img alt="Code" src="https://img.shields.io/badge/Evaluation_code-VBVR_Pro_Bench-100000?style=flat-square&logo=github&logoColor=white" height="20" />
+</a>
+<a href="https://github.com/Video-Reason/VBVR-Pro" target="_blank">
+    <img alt="Code" src="https://img.shields.io/badge/Training_Inferenceing-VBVR_Pro-100000?style=flat-square&logo=github&logoColor=white" height="20" />
+</a>
+<a href="https://huggingface.co/papers/2602.20159" target="_blank">
+    <img alt="arXiv" src="https://img.shields.io/badge/arXiv-VBVR_Pro-red?logo=arxiv" height="20" />
+</a>
+<a href="https://huggingface.co/datasets/Video-Reason/VBVR-Pro-SFT-Video" target="_blank">
+    <img alt="Dataset" src="https://img.shields.io/badge/%F0%9F%A4%97%20_VBVR_Pro_Dataset-Data-ffc107?color=ffc107&logoColor=white" height="20" />
+</a>
+<a href="https://huggingface.co/datasets/Video-Reason/VBVR-Pro-Bench/tree/main" target="_blank">
+    <img alt="Bench Data" src="https://img.shields.io/badge/%F0%9F%A4%97%20_VBVR_Pro_Bench-Data-ffc107?color=ffc107&logoColor=white" height="20" />
+</a>
+<a href="https://video-reason.com/pro/bench/#leaderboard" target="_blank">
+    <img alt="Leaderboard" src="https://img.shields.io/badge/%F0%9F%A4%97%20_VBVR_Pro_Bench-Leaderboard-ffc107?color=ffc107&logoColor=white" height="20" />
+</a>
+
+## Overview
+Native visual reasoning, i.e., reasoning through visual generation, has recently emerged as a promising direction for studying visual intelligence beyond language. Yet progress remains bottlenecked by the lack of scalable training tasks, reliable feedback, and controlled comparisons across generative substrates. In this work, we introduce **VBVR-Pro**, a closed-loop testbed that makes native visual reasoning through generation trainable, verifiable, optimizable, and experimentally controllable. **1) Task scaling.** VBVR-Pro turns visual reasoning into a controlled task space of *300* procedurally generated tasks. Models trained on VBVR-Pro show strong transfer beyond the proposed suite across *six* held-out visual reasoning benchmarks such as RISE-Video, MME-CoF-Pro, and BabyVision. Further analysis validates that these gains reflect visual reasoning rather than instruction-pattern fitting. **2) Verifiable rewards.** VBVR-Pro provides verifiable reward scorers for task-grounded evaluation. Through a systematic study of leading MLLMs as judges, we identify recurring failure modes of the prevalent *VLM-as-a-judge* paradigm. In contrast, the proposed scorers are grounded on verifiable task-specific rules, achieve fine-grained alignment with human judgments. Importantly, they serve as reliable reward signals for large-scale multi-task reinforcement learning and demonstrate stronger post-RL performance across visual reasoning tasks. **3) Mechanism study.** VBVR-Pro enables controlled modality studies across more than *30* image, video, and interleaved generators. Our analysis shows that video generation remains strongest for tasks requiring persistent spatiotemporal state tracking, while interleaved generation provides a compute-efficient alternative by externalizing intermediate visual states. Critically, ablations and probing confirm the presence of vision-native trajectories, that are a more crucial substrate than explicit linguistic chains of thought for visual reasoning. We release all data, models, scorers, and code to facilitate future research.
+
+The models are presented in the paper [VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning](https://huggingface.co/papers/2602.20159).
+
+## Models Zoo
+<table border="1" cellspacing="0" cellpadding="4" style="border-collapse: collapse; width: 100%;">
+  <thead>
+    <tr>
+      <th width="260" style="min-width: 260px;">Model</th>
+      <th>Base Architecture</th>
+      <th>Other Remarks</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr style="border-top: 4px solid #6b7280; background-color: #e5e7eb;"><th colspan="3" align="left">Image Generation Models</th></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-BAGEL"><strong>VBVR-Pro-BAGEL</strong></a></td><td>BAGEL-7B-MoT</td><td>Complete model</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-FLUX2-dev">VBVR-Pro-FLUX2-dev</a></td><td>FLUX.2-dev</td><td>Complete model, Diffusers format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-FLUX2-dev-diffsynth">VBVR-Pro-FLUX2-dev-diffsynth</a></td><td>FLUX.2-dev</td><td>LoRA model, DiffSynth format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Qwen-Image-Edit">VBVR-Pro-Qwen-Image-Edit</a></td><td>Qwen-Image-Edit-2511</td><td>Complete model, Diffusers format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Qwen-Image-Edit-diffsynth">VBVR-Pro-Qwen-Image-Edit-diffsynth</a></td><td>Qwen-Image-Edit-2511</td><td>LoRA model, DiffSynth format</td></tr>
+    <tr style="border-top: 4px solid #6b7280; background-color: #e5e7eb;"><th colspan="3" align="left">Interleaved Image Generation Models</th></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-ThinkMorph">VBVR-Pro-ThinkMorph</a></td><td>ThinkMorph-7B</td><td>Complete model</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-SenseNova-U1">VBVR-Pro-SenseNova-U1</a></td><td>SenseNova-U1-8B-MoT</td><td>Complete model</td></tr>
+    <tr style="border-top: 4px solid #6b7280; background-color: #e5e7eb;"><th colspan="3" align="left">Video Generation Models</th></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-LTX2.3">VBVR-Pro-LTX2.3</a></td><td>LTX-Video-2.3</td><td>Complete model, Diffusers format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-LTX2.3-diffsynth">VBVR-Pro-LTX2.3-diffsynth</a></td><td>LTX-Video-2.3</td><td>LoRA model, DiffSynth format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Wan2.1-I2V-14B">VBVR-Pro-Wan2.1-I2V-14B</a></td><td>Wan2.1-I2V-14B-720P</td><td>Complete model, Diffusers format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Wan2.1-I2V-14B-diffsynth">VBVR-Pro-Wan2.1-I2V-14B-diffsynth</a></td><td>Wan2.1-I2V-14B-720P</td><td>LoRA model, DiffSynth format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Wan2.2-I2V-A14B">VBVR-Pro-Wan2.2-I2V-A14B</a></td><td>Wan2.2-I2V-A14B</td><td>Complete model, Diffusers format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Wan2.2-I2V-A14B-diffsynth">VBVR-Pro-Wan2.2-I2V-A14B-diffsynth</a></td><td>Wan2.2-I2V-A14B</td><td>LoRA model, DiffSynth format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Wan2.2-TI2V-5B">VBVR-Pro-Wan2.2-TI2V-5B</a></td><td>Wan2.2-TI2V-5B</td><td>Complete model, Diffusers format</td></tr>
+    <tr><td><a href="https://huggingface.co/Video-Reason/VBVR-Pro-Wan2.2-TI2V-5B-diffsynth">VBVR-Pro-Wan2.2-TI2V-5B-diffsynth</a></td><td>Wan2.2-TI2V-5B</td><td>LoRA model, DiffSynth format</td></tr>
+  </tbody>
+</table>
+
+## Release Information
+
+VBVR-Pro BAGEL is an image-editing and interleaved keyframe-generation fine-tune of [ByteDance-Seed/BAGEL-7B-MoT](https://huggingface.co/ByteDance-Seed/BAGEL-7B-MoT). This export uses training step **38,000** from `260717_NEO_VBVR_PRO_5e-6`.
+
+Given an initial image and a text instruction, the model generates the next edited image. It was trained for the VBVR-Bench Pro image setting with visual generation enabled, a maximum latent side of 64, and the BAGEL Mixture-of-Transformer-Experts language backbone.
+
+### Checkpoint format
+
+The training checkpoint stores only the 797 trainable tensors in FP32. For this release, those tensors were merged with the 426 frozen ViT and connector tensors from the base BAGEL EMA checkpoint, then the complete 1,223-tensor model was converted to BF16. The FLUX autoencoder is provided separately as `ae.safetensors`, following the layout of [SenseNova-SI-1.1-BAGEL-7B-MoT](https://huggingface.co/sensenova/SenseNova-SI-1.1-BAGEL-7B-MoT).
+
+Training-only optimizer shards and scheduler state are not included.
+
+| File | Purpose |
+| --- | --- |
+| `model.safetensors` | Complete merged BAGEL model in BF16 |
+| `ae.safetensors` | FLUX autoencoder used by BAGEL |
+| `model.safetensors.index.json` | Weight-to-file map for the model and autoencoder |
+| `config.json` | Integrated BAGEL architecture configuration |
+| `llm_config.json`, `vit_config.json` | Language and vision encoder configurations |
+| `tokenizer.json`, `tokenizer_config.json`, `vocab.json`, `merges.txt` | Qwen2 tokenizer assets |
+| `generation_config.json`, `preprocessor_config.json` | Text-generation and image-preprocessing defaults |
+
+In this release, we present all models presented in paper 
+[**VBVR-Pro-Trained-Models**](https://huggingface.co/collections/Video-Reason/VBVR-Pro),
+[**VBVR-Pro-Dataset-Video**](https://huggingface.co/datasets/Video-Reason/VBVR-Pro-SFT-Video),
+[**VBVR-Pro-Dataset-Image**](https://huggingface.co/datasets/Video-Reason/VBVR-Pro-SFT-Image),
+[**VBVR-Pro-Bench**](https://huggingface.co/datasets/Video-Reason/VBVR-Pro-Bench),
+[**VBVR-Pro-Code**](https://github.com/Video-Reason/VBVR-Pro) and 
+[**VBVR-Bench-Leaderboard**](https://video-reason.com/pro/bench/#leaderboard).
+
+## VBVR-Pro Benchmark Results
+<table border="1" cellspacing="0" cellpadding="4" style="border-collapse: collapse; width: 100%; font-size: 12px;">
+  <thead>
+    <tr>
+      <th rowspan="2" width="260" style="min-width: 260px;">Models</th>
+      <th rowspan="2">Overall</th>
+      <th colspan="6">In-Domain by Category</th>
+      <th colspan="6">Out-of-Domain by Category</th>
+    </tr>
+    <tr>
+      <th>Avg.</th><th>Abst.</th><th>Know.</th><th>Perc.</th><th>Spat.</th><th>Trans.</th>
+      <th>Avg.</th><th>Abst.</th><th>Know.</th><th>Perc.</th><th>Spat.</th><th>Trans.</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr style="border-top: 4px solid #6b7280; background-color: #e5e7eb;"><th colspan="14" align="left">Image Generation Models</th></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Proprietary Models</th></tr>
+    <tr><td>Qwen-Image-2.0</td><td><u>0.313</u></td><td><u>0.248</u></td><td><u>0.269</u></td><td><u>0.196</u></td><td><u>0.225</u></td><td><u>0.170</u></td><td><u>0.132</u></td><td><u>0.378</u></td><td><u>0.341</u></td><td><u>0.235</u></td><td><u>0.391</u></td><td><u>0.384</u></td><td><u>0.080</u></td></tr>
+    <tr><td>Seedream-5.0-Pro</td><td><strong>0.557</strong></td><td><strong>0.485</strong></td><td><strong>0.518</strong></td><td><strong>0.312</strong></td><td><strong>0.509</strong></td><td><strong>0.401</strong></td><td><strong>0.217</strong></td><td><strong>0.629</strong></td><td><strong>0.507</strong></td><td><strong>0.455</strong></td><td><strong>0.661</strong></td><td><strong>0.559</strong></td><td><strong>0.202</strong></td></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Open-source Models</th></tr>
+    <tr><td>BAGEL-7B-MoT</td><td>0.089</td><td><u>0.066</u></td><td>0.039</td><td><u>0.085</u></td><td>0.067</td><td>0.046</td><td>0.027</td><td>0.111</td><td><strong>0.201</strong></td><td>0.031</td><td>0.073</td><td>0.028</td><td><strong>0.121</strong></td></tr>
+    <tr><td>FLUX.2-dev</td><td><strong>0.157</strong></td><td><strong>0.108</strong></td><td><u>0.088</u></td><td><strong>0.109</strong></td><td><u>0.072</u></td><td><u>0.100</u></td><td><strong>0.066</strong></td><td><strong>0.206</strong></td><td><u>0.197</u></td><td><strong>0.165</strong></td><td><strong>0.184</strong></td><td><strong>0.241</strong></td><td>0.077</td></tr>
+    <tr><td>Qwen-Image-Edit</td><td><u>0.134</u></td><td><strong>0.108</strong></td><td><strong>0.092</strong></td><td>0.082</td><td><strong>0.100</strong></td><td><strong>0.109</strong></td><td><u>0.056</u></td><td><u>0.159</u></td><td>0.176</td><td><u>0.063</u></td><td><u>0.141</u></td><td><u>0.182</u></td><td><u>0.082</u></td></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Strong Baselines</th></tr>
+    <tr><td>VBVR-Pro-BAGEL</td><td>0.172</td><td>0.168</td><td>0.199</td><td>0.105</td><td>0.110</td><td>0.213</td><td>0.055</td><td>0.176</td><td>0.254</td><td>0.104</td><td>0.148</td><td>0.015</td><td><u>0.145</u></td></tr>
+    <tr><td>VBVR-Pro-FLUX.2</td><td><strong>0.407</strong></td><td><strong>0.484</strong></td><td><strong>0.483</strong></td><td><strong>0.323</strong></td><td><strong>0.367</strong></td><td><strong>0.449</strong></td><td><strong>0.336</strong></td><td><strong>0.330</strong></td><td><strong>0.361</strong></td><td><strong>0.272</strong></td><td><strong>0.255</strong></td><td><strong>0.454</strong></td><td>0.128</td></tr>
+    <tr><td>VBVR-Pro-Qwen-Image</td><td><u>0.322</u></td><td><u>0.332</u></td><td><u>0.298</u></td><td><u>0.217</u></td><td><u>0.193</u></td><td><u>0.431</u></td><td><u>0.222</u></td><td><u>0.311</u></td><td><u>0.341</u></td><td><u>0.239</u></td><td><u>0.233</u></td><td><u>0.413</u></td><td><strong>0.181</strong></td></tr>
+    <tr style="border-top: 4px solid #6b7280; background-color: #e5e7eb;"><th colspan="14" align="left">Interleaved Image Generation Models</th></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Proprietary Models</th></tr>
+    <tr><td>GPT-Image-2</td><td><u>0.507</u></td><td><u>0.428</u></td><td><u>0.456</u></td><td><u>0.318</u></td><td><u>0.428</u></td><td><u>0.206</u></td><td><strong>0.300</strong></td><td><u>0.587</u></td><td><u>0.398</u></td><td><u>0.413</u></td><td><u>0.633</u></td><td><u>0.480</u></td><td><strong>0.303</strong></td></tr>
+    <tr><td>Nano Banana Pro</td><td><strong>0.564</strong></td><td><strong>0.480</strong></td><td><strong>0.518</strong></td><td><strong>0.422</strong></td><td><strong>0.512</strong></td><td><strong>0.285</strong></td><td><u>0.174</u></td><td><strong>0.648</strong></td><td><strong>0.553</strong></td><td><strong>0.499</strong></td><td><strong>0.657</strong></td><td><strong>0.585</strong></td><td><u>0.220</u></td></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Open-source Models</th></tr>
+    <tr><td>ThinkMorph-7B</td><td>0.154</td><td>0.113</td><td>0.100</td><td>0.082</td><td>0.101</td><td>0.148</td><td>0.031</td><td>0.195</td><td>0.176</td><td>0.166</td><td>0.163</td><td>0.253</td><td>0.103</td></tr>
+    <tr><td>VBVR-SenseNova-U1</td><td><u>0.408</u></td><td><u>0.469</u></td><td><u>0.356</u></td><td><u>0.313</u></td><td><u>0.373</u></td><td><strong>0.386</strong></td><td><strong>0.477</strong></td><td><u>0.347</u></td><td><u>0.291</u></td><td><u>0.317</u></td><td><u>0.275</u></td><td><u>0.480</u></td><td><u>0.238</u></td></tr>
+    <tr><td>SenseNova-U1-8B-MoT</td><td><strong>0.565</strong></td><td><strong>0.533</strong></td><td><strong>0.501</strong></td><td><strong>0.395</strong></td><td><strong>0.544</strong></td><td><u>0.355</u></td><td><u>0.349</u></td><td><strong>0.597</strong></td><td><strong>0.448</strong></td><td><strong>0.495</strong></td><td><strong>0.533</strong></td><td><strong>0.717</strong></td><td><strong>0.401</strong></td></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Strong Baselines</th></tr>
+    <tr><td>VBVR-Pro-ThinkMorph</td><td><u>0.373</u></td><td><u>0.402</u></td><td><u>0.403</u></td><td><u>0.344</u></td><td><u>0.238</u></td><td><u>0.454</u></td><td><u>0.184</u></td><td><u>0.344</u></td><td><u>0.367</u></td><td><u>0.224</u></td><td><u>0.238</u></td><td><u>0.535</u></td><td><u>0.257</u></td></tr>
+    <tr><td>VBVR-Pro-SenseNova-U1</td><td><strong>0.638</strong></td><td><strong>0.811</strong></td><td><strong>0.648</strong></td><td><strong>0.695</strong></td><td><strong>0.621</strong></td><td><strong>0.770</strong></td><td><strong>0.541</strong></td><td><strong>0.464</strong></td><td><strong>0.480</strong></td><td><strong>0.328</strong></td><td><strong>0.344</strong></td><td><strong>0.558</strong></td><td><strong>0.408</strong></td></tr>
+    <tr style="border-top: 4px solid #6b7280; background-color: #e5e7eb;"><th colspan="14" align="left">Video Generation Models</th></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Proprietary Models</th></tr>
+    <tr><td>Veo 3.1</td><td>0.309</td><td>0.312</td><td><u>0.275</u></td><td>0.299</td><td>0.252</td><td>0.267</td><td>0.157</td><td>0.305</td><td><u>0.305</u></td><td>0.233</td><td>0.252</td><td><u>0.312</u></td><td>0.219</td></tr>
+    <tr><td>Kling V3</td><td><u>0.392</u></td><td><u>0.356</u></td><td>0.213</td><td><u>0.326</u></td><td><u>0.320</u></td><td><u>0.355</u></td><td><u>0.229</u></td><td><u>0.427</u></td><td>0.294</td><td><strong>0.564</strong></td><td><u>0.375</u></td><td>0.242</td><td><u>0.412</u></td></tr>
+    <tr><td>SeedDance 2.0</td><td><strong>0.499</strong></td><td><strong>0.451</strong></td><td><strong>0.338</strong></td><td><strong>0.361</strong></td><td><strong>0.353</strong></td><td><strong>0.468</strong></td><td><strong>0.308</strong></td><td><strong>0.547</strong></td><td><strong>0.369</strong></td><td><u>0.511</u></td><td><strong>0.478</strong></td><td><strong>0.538</strong></td><td><strong>0.532</strong></td></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Open-source Models</th></tr>
+    <tr><td>HunyuanVideo-I2V</td><td>0.054</td><td>0.054</td><td>0.023</td><td>0.064</td><td>0.015</td><td>0.084</td><td>0.032</td><td>0.053</td><td>0.088</td><td>0.014</td><td>0.028</td><td>0.062</td><td>0.055</td></tr>
+    <tr><td>CogVideoX1.5-5B-I2V</td><td>0.085</td><td>0.100</td><td>0.061</td><td>0.118</td><td>0.069</td><td>0.092</td><td>0.060</td><td>0.070</td><td>0.125</td><td>0.038</td><td>0.051</td><td>0.040</td><td>0.024</td></tr>
+    <tr><td>Wan2.1-I2V-14B</td><td>0.100</td><td>0.105</td><td>0.052</td><td>0.125</td><td>0.091</td><td>0.102</td><td>0.052</td><td>0.095</td><td>0.112</td><td>0.073</td><td>0.071</td><td>0.123</td><td>0.044</td></tr>
+    <tr><td>Wan2.2-TI2V-5B</td><td>0.094</td><td>0.066</td><td>0.029</td><td>0.073</td><td>0.050</td><td>0.083</td><td>0.031</td><td>0.122</td><td>0.156</td><td>0.052</td><td>0.106</td><td>0.063</td><td>0.099</td></tr>
+    <tr><td>Wan2.2-I2V-14B-720P</td><td><u>0.182</u></td><td><u>0.157</u></td><td><u>0.082</u></td><td><u>0.131</u></td><td><u>0.110</u></td><td><u>0.161</u></td><td><u>0.156</u></td><td><u>0.207</u></td><td><u>0.224</u></td><td><u>0.139</u></td><td><u>0.140</u></td><td><u>0.195</u></td><td><u>0.273</u></td></tr>
+    <tr><td>LTX2.3-I2AV</td><td>0.112</td><td>0.106</td><td>0.062</td><td>0.109</td><td>0.070</td><td>0.133</td><td>0.055</td><td>0.119</td><td>0.161</td><td>0.135</td><td>0.086</td><td>0.091</td><td>0.050</td></tr>
+    <tr><td>VBVR-Wan2.2</td><td><strong>0.517</strong></td><td><strong>0.548</strong></td><td><strong>0.237</strong></td><td><strong>0.499</strong></td><td><strong>0.334</strong></td><td><strong>0.566</strong></td><td><strong>0.591</strong></td><td><strong>0.486</strong></td><td><strong>0.310</strong></td><td><strong>0.343</strong></td><td><strong>0.345</strong></td><td><strong>0.732</strong></td><td><strong>0.684</strong></td></tr>
+    <tr style="background-color: #dbeafe;"><th colspan="14" align="left">Strong Baselines</th></tr>
+    <tr><td>VBVR-Pro-LTX2.3</td><td>0.425</td><td>0.527</td><td>0.409</td><td>0.510</td><td>0.346</td><td>0.460</td><td>0.390</td><td>0.324</td><td>0.381</td><td>0.108</td><td>0.201</td><td>0.477</td><td>0.386</td></tr>
+    <tr><td>VBVR-Pro-Wan2.1-I2V-14B</td><td><u>0.562</u></td><td><u>0.730</u></td><td><u>0.617</u></td><td><u>0.580</u></td><td><u>0.452</u></td><td><u>0.676</u></td><td><u>0.623</u></td><td><u>0.395</u></td><td><u>0.410</u></td><td><u>0.305</u></td><td><u>0.230</u></td><td><u>0.617</u></td><td><u>0.439</u></td></tr>
+    <tr><td>VBVR-Pro-Wan2.2-TI2V-5B</td><td>0.470</td><td>0.641</td><td>0.528</td><td>0.556</td><td>0.373</td><td>0.565</td><td>0.557</td><td>0.300</td><td>0.333</td><td>0.127</td><td>0.161</td><td>0.505</td><td>0.409</td></tr>
+    <tr><td>VBVR-Pro-Wan2.2-I2V-14B</td><td><strong>0.670</strong></td><td><strong>0.808</strong></td><td><strong>0.632</strong></td><td><strong>0.685</strong></td><td><strong>0.556</strong></td><td><strong>0.751</strong></td><td><strong>0.636</strong></td><td><strong>0.532</strong></td><td><strong>0.479</strong></td><td><strong>0.418</strong></td><td><strong>0.350</strong></td><td><strong>0.679</strong></td><td><strong>0.690</strong></td></tr>
+  </tbody>
+</table>
+
+## Quick Start
+
+### Method 1: Standalone BAGEL inference
+
+#### Requirements
+
+- Linux and a CUDA-capable GPU
+- Python 3.10 or newer
+- PyTorch 2.5 or newer
+- Transformers 4.49
+- Accelerate, Safetensors, Pillow, and FlashAttention
+- The [BAGEL codebase](https://github.com/bytedance-seed/BAGEL)
+
+Install the official BAGEL environment:
+
+```bash
+git clone https://github.com/bytedance-seed/BAGEL.git
+cd BAGEL
+conda create -n bagel python=3.10 -y
+conda activate bagel
+pip install -r requirements.txt
+pip install flash_attn==2.5.8 --no-build-isolation
+```
+
+Download this repository:
+
+```python
+from huggingface_hub import snapshot_download
+
+snapshot_download(
+    repo_id="Video-Reason/VBVR-Pro-BAGEL",
+    local_dir="models/VBVR-Pro-BAGEL",
+    allow_patterns=["*.json", "*.safetensors", "*.md", "*.txt"],
+)
+```
+
+#### Inference
+
+This export uses the original BAGEL model implementation. Construct `Bagel` with `visual_gen=True`, `visual_und=True`, `max_latent_size=64`, and `layer_module="Qwen2MoTDecoderLayer"`; load `model.safetensors` with Accelerate or Safetensors; and load the autoencoder from `ae.safetensors`.
+
+The evaluation path used for this checkpoint was:
+
+```bash
+bash inference/run_bagel_vbvr_pro_eval.sh \
+  In-Domain_50 \
+  /path/to/VBVR-Pro-BAGEL/model.safetensors \
+  vigen 8 25
+```
+
+The launcher invokes `inference/bagel_vbvr_pro.py`. It builds the BAGEL architecture from the base assets, supplies this fine-tuned checkpoint, and calls `InterleaveInferencer.interleave_inference` with a text instruction followed by the initial image.
+
+#### Evaluation defaults
+
+| Setting | Value |
+| --- | --- |
+| Input | `prompt.txt` and `first_frame.png` |
+| Output | `frame_1.png` |
+| Denoising steps | 50 |
+| Text CFG scale | 4.0 |
+| Image CFG scale | 2.0 |
+| CFG interval | `[0.0, 1.0]` |
+| Timestep shift | 3.0 |
+| CFG renormalization | `text_channel` |
+| Seed | 1 |
+
+When using the legacy BAGEL `app.py`, which looks specifically for `ema.safetensors`, create a local alias after downloading:
+
+```bash
+ln -s model.safetensors models/VBVR-Pro-BAGEL/ema.safetensors
+python app.py --model_path models/VBVR-Pro-BAGEL
+```
+
+The alias is not part of the Hugging Face release because `model.safetensors` is the canonical weight filename used by the reference release layout.
+
+### Method 2: Unified VBVR-Pro inference
+
+Clone [Video-Reason/VBVR-Pro](https://github.com/Video-Reason/VBVR-Pro) and
+create its unified inference environment:
+
+```bash
+git clone https://github.com/Video-Reason/VBVR-Pro.git
+cd VBVR-Pro/
+uv sync --extra cu124 # or one of [cu118|cu121|cu124|cu126|cu128|cu129]
+source .venv/bin/activate
+```
+
+Then run the unified inference script:
+
+```bash
+python example.py \
+  --model_path Video-Reason/VBVR-Pro-BAGEL \
+  --image_paths first_frame.png \
+  --prompt "Move the object to the requested destination." \
+  --output output.png
+```
+
+## Notes
+
+- The checkpoint is intended for BF16 CUDA inference. A high-memory GPU is recommended for direct single-GPU loading.
+- The released weights are a complete inference checkpoint; the original BAGEL EMA file is not needed after the merge.
+- Higher output resolutions and additional generated images increase runtime and memory use substantially.
+- Generated content should be reviewed before use in downstream applications.
+
+## License
+
+This model is released under the Apache License 2.0, following the base BAGEL model. BAGEL is fine-tuned from Qwen2.5-7B-Instruct and SigLIP and uses the FLUX.1-schnell autoencoder; consult the upstream repositories for their notices.
+
+## Citation
+
+```bibtex
+@article{deng2025bagel,
+  title   = {Emerging Properties in Unified Multimodal Pretraining},
+  author  = {Deng, Chaorui and Zhu, Deyao and Li, Kunchang and Gou, Chenhui and Li, Feng and Wang, Zeyu and Zhong, Shu and Yu, Weihao and Nie, Xiaonan and Song, Ziang and Shi, Guang and Fan, Haoqi},
+  journal = {arXiv preprint arXiv:2505.14683},
+  year    = {2025}
+}
+```

ae.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afc8e28272cd15db3919bacdb6918ce9c1ed22e96cb12c4d5ed0fba823529e38
+size 335304388

config.json

@@ -0,0 +1,60 @@
+{
+  "_name_or_path": "VBVR-Pro-BAGEL",
+  "architectures": [
+    "Bagel"
+  ],
+  "model_type": "bagel",
+  "visual_gen": true,
+  "visual_und": true,
+  "llm_config": {
+    "architectures": [
+      "Qwen2ForCausalLM"
+    ],
+    "attention_dropout": 0.0,
+    "bos_token_id": 151643,
+    "eos_token_id": 151645,
+    "hidden_act": "silu",
+    "hidden_size": 3584,
+    "initializer_range": 0.02,
+    "intermediate_size": 18944,
+    "layer_module": "Qwen2MoTDecoderLayer",
+    "max_position_embeddings": 32768,
+    "max_window_layers": 28,
+    "model_type": "qwen2",
+    "num_attention_heads": 28,
+    "num_hidden_layers": 28,
+    "num_key_value_heads": 4,
+    "qk_norm": true,
+    "rms_norm_eps": 1e-06,
+    "rope_theta": 1000000.0,
+    "sliding_window": 131072,
+    "tie_word_embeddings": false,
+    "torch_dtype": "bfloat16",
+    "transformers_version": "4.49.0",
+    "use_cache": true,
+    "use_sliding_window": false,
+    "vocab_size": 152064
+  },
+  "vit_config": {
+    "hidden_size": 1152,
+    "image_size": 980,
+    "intermediate_size": 4304,
+    "model_type": "siglip_vision_model",
+    "num_attention_heads": 16,
+    "num_channels": 3,
+    "num_hidden_layers": 27,
+    "patch_size": 14
+  },
+  "vae_config": {
+    "downsample": 8,
+    "z_channels": 16
+  },
+  "connector_act": "gelu_pytorch_tanh",
+  "interpolate_pos": false,
+  "latent_patch_size": 2,
+  "max_latent_size": 64,
+  "timestep_shift": 1.0,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.49.0",
+  "vit_max_num_patch_per_side": 70
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "pad_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_p": 0.8,
+  "top_k": 20,
+  "transformers_version": "4.37.0"
+}

llm_config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.43.1",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60662ee31111ab045b52f77739893c65d378b90b07a7af9c697f29428c7c4c8b
+size 29214685368

preprocessor_config.json

@@ -0,0 +1,24 @@
+{
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "image_processor_type": "SiglipImageProcessor",
+  "image_std": [
+    0.5,
+    0.5,
+    0.5
+  ],
+  "processor_class": "BagelProcessor",
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 980,
+    "width": 980
+  }
+}

tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

vit_config.json

@@ -0,0 +1,9 @@
+{
+    "hidden_size": 1152,
+    "image_size": 980,
+    "intermediate_size": 4304,
+    "model_type": "siglip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 27,
+    "patch_size": 14
+}