README.md

---
language: en
license: apache-2.0
tags:
- vision-language-model
- visual-storytelling
- chain-of-thought
- grounded-text-generation
- cross-frame-consistency
- storytelling
- image-to-text
- contrastive-learning
- reinforcement-learning
- entity-reidentification
datasets:
- daniel3303/StoryReasoningAdversarialDPO
- daniel3303/StoryReasoning
metrics:
- precision
- recall
- bleu
- meteor
- rouge
- map
base_model:
- daniel3303/QwenStoryteller
pipeline_tag: image-to-text
model-index:
  - name: QwenStoryteller2
    results:
      - task:
          type: visual-storytelling
          name: Visual Storytelling
        dataset:
          name: StoryReasoningAdversarialDPO
          type: daniel3303/StoryReasoningAdversarialDPO
          split: test
        metrics:
          - name: Character Precision
            type: precision
            value: 0.78
          - name: Object Precision
            type: precision
            value: 0.29
          - name: Total Precision
            type: precision
            value: 0.45
          - name: mAP
            type: mean_average_precision
            value: 0.31
          - name: Character Recall
            type: recall
            value: 0.77
          - name: Object Recall
            type: recall
            value: 0.28
          - name: Total Recall
            type: recall
            value: 0.48
          - name: F1 Score
            type: f1
            value: 0.41
          - name: METEOR
            type: meteor
            value: 0.17
          - name: ROUGE-L
            type: rouge-l
            value: 0.18
          - name: BLEU-4
            type: bleu-4
            value: 0.057
          - name: Character Persistence (≥5 frames)
            type: accuracy
            value: 0.493
          - name: Object Persistence (≥5 frames)
            type: accuracy
            value: 0.213
          - name: Well-structured Stories
            type: accuracy
            value: 0.975
library_name: transformers
---

# QwenStoryteller2

QwenStoryteller2 is an improved version of QwenStoryteller, fine-tuned using contrastive reinforcement learning with Direct Preference Optimization (DPO) to achieve superior entity re-identification and visual grounding in cross-frame storytelling scenarios.

## Model Description

**Base Model:** QwenStoryteller (Qwen2.5-VL 7B)  
**Training Method:** Contrastive Reinforcement Learning with Direct Preference Optimization (LoRA rank 2048, alpha 4096)  
**Training Dataset:** [StoryReasoningAdversarialDPO](https://huggingface.co/datasets/daniel3303/StoryReasoningAdversarialDPO)

QwenStoryteller2 builds upon the original QwenStoryteller by addressing critical limitations in cross-frame entity consistency through:
- **Contrastive Learning:** Training on both real and synthetic negative story examples
- **Enhanced Entity Re-identification:** Improved tracking of characters and objects across frames
- **Better Grounding:** Superior alignment between narrative elements and visual entities
- **Reduced Hallucinations:** More reliable entity connections and fewer spurious references

The model employs a dual-component reward function that promotes appropriate entity connections in coherent sequences while discouraging incorrect connections in synthetic arrangements.

## Key Improvements Over QwenStoryteller

- **Grounding Performance:** mAP improved from 0.27 to 0.31 (+14.8%), F1 score from 0.35 to 0.41 (+17.1%)
- **Cross-frame Consistency:** Character persistence on ≥5 frames increased from 37.7% to 49.3% (+30.8%)
- **Pronoun Grounding:** Significant improvements across all pronoun types (he: 90.1%→99.1%, she: 91.1%→98.6%, they: 47.6%→68.8%)
- **Structural Quality:** Well-structured stories increased from 79.1% to 97.5% (+23.3%)
- **Entity Tracking:** Object persistence on ≥5 frames improved from 20.9% to 21.3%

## System Prompt

The model was trained with the following system prompt, and we recommend using it for optimal performance:

```
You are an AI storyteller that can analyze sequences of images and create creative narratives. 
First think step-by-step to analyze characters, objects, settings, and narrative structure. 
Then create a grounded story that maintains consistent character identity and object references across frames. 
Use <think></think> tags to show your reasoning process before writing the final story.
```

## Usage

```python
from transformers import Qwen2_5_VLForConditionalGeneration, AutoProcessor
from qwen_vl_utils import process_vision_info
import torch
from PIL import Image

# Load the model
model = Qwen2_5_VLForConditionalGeneration.from_pretrained(
    "daniel3303/QwenStoryteller2", torch_dtype="auto", device_map="auto"
)

# Load processor
processor = AutoProcessor.from_pretrained("daniel3303/QwenStoryteller2")

# Load images
images = [
    Image.open("image1.jpg"),
    Image.open("image2.jpg"),
    Image.open("image3.jpg"),
    Image.open("image4.jpg"),
    Image.open("image5.jpg")
]

# Create image content list
image_content = []
for img in images:
    image_content.append({
        "type": "image",
        "image": img,
    })

# Add text prompt at the end
image_content.append({"type": "text", "text": "Generate a story based on these images."})

# Create messages with system prompt
messages = [
    {
        "role": "system", 
        "content": "You are an AI storyteller that can analyze sequences of images and create creative narratives. First think step-by-step to analyze characters, objects, settings, and narrative structure. Then create a grounded story that maintains consistent character identity and object references across frames. Use <think></think> tags to show your reasoning process before writing the final story."
    },
    {
        "role": "user",
        "content": image_content,
    }
]

# Preparation for inference
text = processor.apply_chat_template(
    messages, tokenize=False, add_generation_prompt=True
)
image_inputs, video_inputs = process_vision_info(messages)
inputs = processor(
    text=[text],
    images=image_inputs,
    videos=video_inputs,
    padding=True,
    return_tensors="pt",
)
inputs = inputs.to(model.device)

# Inference: Generation of the output
generated_ids = model.generate(
    **inputs, 
    max_new_tokens=4096,
    do_sample=True,
    temperature=0.7,
    top_p=0.9
)
generated_ids_trimmed = [
    out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
story = processor.batch_decode(
    generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
)[0]

print(story)
```

### Using vLLM for faster inference

For significantly faster inference, you can use vLLM to serve the model:

```bash
# Install vLLM
pip install vllm

# Serve the model with vLLM
vllm serve daniel3303/QwenStoryteller2
```

## Training Methodology

### Contrastive Learning Framework

QwenStoryteller2 was trained using a novel contrastive reinforcement learning approach:

1. **Synthetic Story Generation:** Extended the StoryReasoning dataset with 4,178 synthetic stories created by sampling images from different movies to create incoherent sequences
2. **Dual-Component Reward Function:** Combined entity re-identification (R_reid) and grounding (R_ground) rewards with structural validation
3. **Direct Preference Optimization:** Used offline preference pairs generated from the reward function to train the model

### Reward Function Components

- **Entity Re-identification Reward:** Tracks character and object persistence across frames, promoting connections in real stories while penalizing them in synthetic ones
- **Grounding Reward:** Evaluates pronoun and proper noun grounding to visual entities
- **Structure Validation:** Ensures generated outputs maintain required format and consistency

### Training Configuration

- **Method:** Direct Preference Optimization (DPO) with LoRA fine-tuning
- **LoRA Parameters:** Rank 2048, alpha 4096
- **Optimizer:** AdamW with learning rate 5×10⁻⁶
- **Batch Size:** 8
- **Epochs:** 3
- **Temperature Parameter (β):** 0.1

## Performance Metrics

| Metric | QwenStoryteller | QwenStoryteller2 | Improvement |
|--------|-----------------|------------------|-------------|
| Character Precision | 0.83 | 0.78 | -6.0% |
| Object Precision | 0.46 | 0.29 | -37.0% |
| Total Precision | 0.57 | 0.45 | -21.1% |
| mAP | 0.27 | 0.31 | +14.8% |
| Character Recall | 0.62 | 0.77 | +24.2% |
| Object Recall | 0.25 | 0.28 | +12.0% |
| Total Recall | 0.40 | 0.48 | +20.0% |
| F1 Score | 0.35 | 0.41 | +17.1% |
| METEOR | 0.14 | 0.17 | +21.4% |
| ROUGE-L | 0.16 | 0.18 | +12.5% |
| BLEU-4 | 0.054 | 0.057 | +5.6% |

## Output Format

QwenStoryteller2 produces enhanced outputs with improved consistency:

1. **Chain-of-Thought Analysis (`<think></think>`):** More accurate structured analysis with:
   - Improved character tables with consistent identity references
   - Better object tracking with enhanced spatial coordination
   - More reliable setting categorization
   - Stronger narrative structure modeling

2. **Grounded Story:** Enhanced narrative with specialized XML tags:
   - `<gdi>`: Image tags for specific frames
   - `<gdo>`: Entity reference tags with improved accuracy
   - `<gda>`: Action tags with better character-action alignment
   - `<gdl>`: Location/landmark tags with enhanced spatial grounding

## Key Features

- **Enhanced Cross-Frame Consistency:** Superior character and object identity maintenance through contrastive learning
- **Improved Pronoun Grounding:** Better alignment of pronouns with visual entities (up to 99.1% for "he", 98.6% for "she")
- **Reduced Hallucinations:** Fewer incorrect entity connections and spurious references
- **Robust Entity Discrimination:** Learned ability to distinguish when cross-frame connections are appropriate
- **Better Structural Quality:** Near-perfect adherence to expected output format (97.5%)

## Limitations

- Precision scores show some reduction compared to the original model due to increased recall
- Training data derived from movies may introduce cinematic biases
- Entity re-identification still relies primarily on visual similarity within bounding boxes
- Performance validated only on 7B parameter scale
- Optimal real-to-synthetic story ratio (2:1) may not generalize to all scenarios

## Citation

```bibtex
TODO

@misc{oliveira2025storyreasoningdatasetusingchainofthought,
      title={StoryReasoning Dataset: Using Chain-of-Thought for Scene Understanding and Grounded Story Generation}, 
      author={Daniel A. P. Oliveira and David Martins de Matos},
      year={2025},
      eprint={2505.10292},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2505.10292}
}
```

## Contact

For questions or feedback regarding this model, please contact:
- Daniel A. P. Oliveira ([email protected])