Update README.md

README.md

@@ -22,23 +22,30 @@ library_name: transformers
 
 # **Qwen2.5-VL-7B-Abliterated-Caption-it**
 
-> **Qwen2.5-VL-7B-Abliterated-Caption-it** is a fine-tuned version of **Qwen2.5-VL-7B-Instruct**, optimized for **Abliterated Captioning** / **Uncensored Captioning**. This model excels at generating detailed, context-rich, and high-fidelity captions across **diverse image categories** and **variational aspect ratios**, offering robust visual understanding without filtering or censorship.
+> The **Qwen2.5-VL-7B-Abliterated-Caption-it** model is a fine-tuned version of **Qwen2.5-VL-7B-Instruct**, tailored for **Abliterated Captioning** / **Uncensored Image Captioning**. This variant is designed to generate highly detailed and descriptive captions across a broad range of visual categories, including images with complex, sensitive, or nuanced content—across varying aspect ratios and resolutions.
 
-# Key Enhancements
+# Key Highlights
 
-* **Uncensored & Detailed Captioning**: Capable of producing in-depth captions for a wide range of image types, including complex or non-standard visual content.
+* **Abliterated / Uncensored Captioning**: Fine-tuned to bypass common content filters while preserving factual and descriptive richness across diverse visual categories.
 
-* **Aspect-Ratio-Aware Visual Description**: Robust performance across images of varying sizes, orientations, and layouts.
+* **High-Fidelity Descriptions**: Generates comprehensive captions for general, artistic, technical, abstract, and low-context images.
 
-* **Context-Aware Multimodal Reasoning**: Understands visual scenes in conjunction with textual prompts, enabling accurate and comprehensive interpretations.
+* **Robust Across Aspect Ratios**: Capable of accurately captioning images with wide, tall, square, and irregular dimensions.
 
-* **Support for OCR, Layout, and Visual QA Tasks**: Maintains strong performance on document-type images, retaining capability for text extraction and visual question answering.
+* **Variational Detail Control**: Produces outputs with both high-level summaries and fine-grained descriptions as needed.
 
-* **Instruction-Tuned for Precision**: Fine-tuned to follow user prompts and provide captions tailored to user intent, even with minimal or ambiguous input.
+* **Foundation on Qwen2.5-VL Architecture**: Leverages the strengths of the Qwen2.5-VL-7B multimodal model for visual reasoning, comprehension, and instruction-following.
 
-* **Multilingual and Multi-Domain Compatibility**: Provides accurate captioning for content across languages and specialized domains.
+* **Multilingual Output Capability**: Can support multilingual descriptions (English as default), adaptable via prompt engineering.
 
-* **Stable Across Benchmark Tests**: Competent performance across visual-language benchmarks such as COCO, DocVQA, TextVQA, and others, including ablative scenarios.
+# Training Details
+
+This model was fine-tuned using the following datasets:
+
+* **[prithivMLmods/blip3o-caption-mini-arrow](https://huggingface.co/datasets/prithivMLmods/blip3o-caption-mini-arrow)**
+* **Private/unlisted datasets** curated for uncensored and domain-specific image captioning tasks.
+
+The training objective focused on enhancing performance in unconstrained, descriptive image captioning—especially for edge cases commonly filtered out in standard captioning benchmarks.
 
 # Quick Start with Transformers
 
@@ -60,7 +67,7 @@ messages = [
                 "type": "image",
                 "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
             },
-            {"type": "text", "text": "Describe this image."},
+            {"type": "text", "text": "Describe this image in detail."},
         ],
     }
 ]
@@ -90,20 +97,17 @@ print(output_text)
 
 # Intended Use
 
-This model is intended for:
+This model is suited for:
 
-* Generating rich, uncensored captions from diverse image categories, including stylized, medical, artistic, and real-world scenarios.
-* Supporting flexible aspect ratios and complex visual scenes.
-* Captioning tasks that demand unfiltered, detailed outputs without masking sensitive or subtle content.
-* Multimodal reasoning between visual and textual content for creative, research, and analytical use cases.
-* Caption generation for datasets that require fine-grained annotation or description, such as image-based storytelling or surveillance contexts.
-* Multilingual caption generation for global image datasets.
-* Vision-based interaction systems that need unrestricted and accurate scene understanding.
+* Generating detailed and unfiltered image captions for general-purpose or artistic datasets.
+* Content moderation research, red-teaming, and generative safety evaluations.
+* Enabling descriptive captioning for visual datasets typically excluded from mainstream models.
+* Use in creative applications (e.g., storytelling, art generation) that benefit from rich descriptive captions.
+* Captioning for non-standard aspect ratios and stylized visual content.
 
 # Limitations
 
-* May produce uncensored or sensitive outputs not suitable for all applications.
-* Performance may vary on abstract or adversarial images outside of the training domain.
-* High computational demands; not optimized for edge or low-resource devices.
-* Accuracy on handwritten or heavily distorted content may be reduced.
-* As with most generative models, hallucination or context drift is possible in some scenarios. 
+* May produce explicit, sensitive, or offensive descriptions depending on image content and prompts.
+* Not suitable for deployment in production systems requiring content filtering or moderation.
+* Can exhibit variability in caption tone or style depending on input prompt phrasing.
+* Accuracy for unfamiliar or synthetic visual styles may vary.