hi, i am prithiv! i am a graduate engineer [UG 2024] in information technology from gcee, focused on working with LLM enhancements, computer vision models, and improving multimodal AI capabilities.
| Camel-Doc-OCR-080125 | Lumian2-VLR-7B-Thinking |
|---|---|
 |
|
 |
|
| Advanced Qwen2.5-VL fine-tuned specialist for document retrieval, content extraction, and analysis recognition, delivering superior document comprehension with Opendoc2-Analysis-Recognition training. | High-fidelity vision-language reasoning system with explicit grounded reasoning, producing structured reasoning traces aligned with visual coordinates for explainable multimodal understanding and step-by-step analysis. |
| Qwen2.5-VL-7B-Abliterated-Caption-it | DeepCaption-VLA-7B |
|---|---|
|
|
 |
 |
| Uncensored captioning specialist generating highly detailed descriptions across complex visual categories and sensitive content, optimized for comprehensive image analysis across varying aspect ratios. | Precision image captioning model focused on defining visual properties, object attributes, and scene details with exceptional descriptive accuracy across diverse image spectrums and dimensional variations. |
| Megalodon-OCR-Sync-0713 | DREX-062225-exp |
|---|---|
|
|
 |
 |
| Qwen2.5-VL-3B specialist trained on 200K image pairs including 70K Corvus-OCR-Caption-Mix samples, optimized for document OCR captioning, image reasoning, and visual analysis across variational dimensions. | Document Retrieval and Extraction eXpert built on docscopeOCR-7B architecture, specialized for superior document analysis and information extraction with Opendoc2-Analysis-Recognition training optimization. |
For more, visit:
(Open LLM Leaderboard)
| Galactic-Qwen-14B-Exp2 | Sombrero-Opus-14B-Elite5 |
|---|---|
|
|
 |
 |
| Advanced reasoning powerhouse optimized for general-purpose intelligence, excelling in contextual understanding, logical deduction, and complex multi-step problem-solving with superior performance metrics. | Elite conversational AI fine-tuned using advanced chain-of-thought reasoning methodologies and specialized datasets, delivering enhanced comprehension, structured responses, and intelligent dialogue capabilities. |
| Viper-Coder-v1.1 | Dinobot-Opus-14B-Exp |
|---|---|
|
|
 |
 |
| Specialized coding AI fine-tuned on synthetic datasets leveraging cutting-edge coding logits and CoT methodologies, optimizing chain-of-thought reasoning and advanced logical problem-solving for programming tasks. | High-performance abliterated model based on Qwen 2.5 architecture, engineered for superior reasoning, detailed explanations, and conversational excellence with enhanced contextual understanding and multi-step analytical capabilities. |
For more, visit:
| Stranger Zone | Stranger Guard |
|---|---|
|
|
 |
|
| Building illustration adapters for diffusion models, The Stranger Zone specializes in intelligence development, focusing on fine-tuning models for computer vision ; text-to-image specialized adapters (LoRA). | Stranger Guard specializes in building strict content moderation models, with a core focus on advanced computer vision tasks. Our team develops precision-driven AI systems capable of detecting, classifying, and moderating visual content at scale. |