Brief Details: Specialized anime-style image generation model with strong capabilities in character and landscape creation, featuring high-quality detail rendering and versatile prompt handling.
BRIEF-DETAILS: Versatile instant voice cloning model enabling multi-language speech generation with style control and zero-shot capabilities. MIT licensed, supports English & Chinese.
Brief-details: A powerful 25.5B parameter multimodal LLM combining InternViT-6B and InternLM2-20B, featuring dynamic high-resolution processing and strong bilingual capabilities.
Brief-details: Stable Diffusion 2.0-based model specialized in futuristic Sci-Fi imagery, featuring high-quality 3D visuals triggered by "future style" token. 402 likes, 767 downloads.
Brief-details: Mini-omni is a multimodal LLM capable of real-time speech-to-speech conversation with streaming audio output, built on Qwen2-0.5B base model for English language processing.
Brief-details: MiniGPT-4 combines BLIP-2's visual encoder with Vicuna LLM for advanced vision-language understanding, trained in two stages for enhanced image comprehension and natural conversation.
Brief Details: ChatGLM-6B-INT4 is a quantized bilingual LLM with 6B parameters, optimized for Chinese-English dialogue, requiring only 6GB VRAM for inference.
BRIEF-DETAILS: Vicuna-13b-delta-v1.1 is a fine-tuned LLaMA variant trained on 70K ShareGPT conversations, offering strong chat capabilities for researchers and hobbyists
BRIEF DETAILS: A powerful 46.7B parameter Mixtral-based model fine-tuned with DPO, achieving state-of-the-art performance. Features ChatML format and extensive benchmark improvements.
Brief Details: AsiaFacemix is a specialized AI model focused on improving Asian facial features in image generation, based on basil mix, dreamlike, and ProtoGen models. Licensed under OpenRail.
Brief-details: Dolphin 2.9 is an 8B parameter LLaMA3-based model optimized for conversational AI, coding, and instruction-following with uncensored capabilities and 4k context length.
Brief-details: A Stable Diffusion model fine-tuned for classic animation-style image generation, specializing in Disney-like character rendering with 412 likes and 338 downloads.
BRIEF DETAILS: A specialized Stable Diffusion model for generating pixel art sprite sheets from 4 angles (front, back, left, right). Apache-2.0 licensed with 1.3K+ downloads.
Brief-details: A powerful 46.7B parameter MoE model quantized for efficient deployment, supporting 5 languages with various GGUF formats for different performance/quality trade-offs
Brief-details: Compact 968M-parameter multimodal model optimized for edge devices. Features 9x token reduction and DPO training for reliable visual-text processing.
Brief Details: MiniCPM-V-2 is a 3.43B parameter bilingual multimodal LLM achieving GPT-4V-level performance, supporting high-res images and efficient deployment on mobile devices.
Brief Details: MistralLite - A fine-tuned Mistral-7B model optimized for long context (32K tokens) with enhanced retrieval capabilities. Built by Amazon.
Brief-details: Solar Pro Preview is a 22B parameter LLM optimized for single GPU deployment, offering performance comparable to 70B models with enhanced instruction-following capabilities and MMLU benchmark excellence.
BRIEF DETAILS: First open-source Chinese Stable Diffusion model trained on 20M filtered Chinese image-text pairs. Uses CLIP-based filtering and specialized text encoder for Chinese concept alignment.
Brief-details: WizardLM-7B-Uncensored is an unfiltered variant of WizardLM, trained without alignment constraints for customizable fine-tuning. Built on PyTorch.
BRIEF-DETAILS: A versatile text-to-image model created by hassanblend, featuring specialized diffusion techniques with 436 likes and 1,993 downloads. Licensed under CreativeML OpenRAIL-M.
