Brief-details: PUMA is a unified multimodal LLM that enables multi-granular visual generation and understanding, supporting diverse text-to-image tasks and precise image editing with balanced control and creativity.
Brief-details: Quantized version of Qwen2-VL-OCR-2B-Instruct with multiple compression variants (0.5GB-1.4GB), optimized for efficient deployment with GGUF format.
BRIEF-DETAILS: GGUF quantized version of Claude 3.5 Sonnet model, offering multiple compression options from 3.3GB to 16.2GB with varying quality-size tradeoffs
BRIEF DETAILS: 12B parameter GGUF quantized model with multiple compression variants, optimized for efficiency. Features Q2 to Q8 quantization options with file sizes from 4.9GB to 13.1GB.
Brief-details: Sa2VA-26B is a multimodal LLM that combines SAM2 and LLaVA capabilities for advanced image/video understanding, segmentation, and QA tasks at 26B parameters.
BRIEF-DETAILS: A specialized BERT-based model designed for ethnicity classification and analysis, created by pparasurama and hosted on HuggingFace.
Brief Details: A specialized AI model focused on creating realistic analog photography effects, developed by digiplay. Available on HuggingFace for artistic image generation.
Brief Details: A creative AI image generation model focused on blending fantasy and reality, specializing in detailed scenes with unique character interactions and atmospheric elements.
Brief Details: ControlNet model for depth-based image generation, trained on 3M depth-image pairs using Stable Diffusion 1.5. Enables precise control over image generation using depth maps.
BRIEF DETAILS: Norwegian Wav2vec2 speech recognition model with 300M parameters, optimized for Bokmål dialect, suitable for ASR tasks and speech processing
BRIEF-DETAILS: HunyuanDiT is a powerful multi-resolution Diffusion Transformer supporting both Chinese and English text-to-image generation, featuring 25-step generation and advanced Chinese understanding capabilities.
BRIEF-DETAILS: 4-bit quantized version of Mistral-Nemo-Instruct optimized for MLX framework, offering efficient deployment on Apple Silicon
BRIEF-DETAILS: Indonesian BERT model fine-tuned for emotion classification tasks, offering text-based emotion analysis for Indonesian language content using transformer architecture
Brief Details: MajicMIX Realistic v5 is a specialized AI image generation model focused on photorealistic outputs, particularly excelling in human portrait generation with enhanced detail and natural features.
Brief-details: NextPhoto_v1 is a specialized AI image model focused on photorealistic outputs, developed by digiplay and available on Hugging Face, designed for high-quality photo generation.
BRIEF DETAILS: Qwen2.5-14B-Instruct optimized by Unsloth for 4-bit quantization. Features improved accuracy, 60% less memory usage, and 2x faster training speeds.
Brief-details: Core ML optimized version of Llama 2 (7B) chat model, converted to float16 with 64 token sequence length, designed for Apple devices
Brief Details: Stable Diffusion 3 Medium optimized with TensorRT - A specialized version of SD3 designed for efficient inference and deployment
Brief-details: LLaMA-13B is a powerful language model with 13 billion parameters, developed by HuggingLlama. Available under non-commercial license, designed for research and academic use.
Brief-details: Llama3-TAIDE-LX-8B-Chat-Alpha1 is an 8B parameter chat model developed by TAIDE, featuring specialized Chinese language capabilities and community licensing.
Brief Details: LLaMA-based 8B parameter reasoning model by Shaleen123. Specialized architecture focusing on logical reasoning capabilities. Available on Hugging Face.
