BRIEF-DETAILS: Large-scale BERT model (32k vocab) trained on 256GB legal corpus, optimized for legal text analysis with 1.7M training steps and custom legal vocabulary
Brief Details: LaTeX OCR model by yhshin - Specialized in converting mathematical equation images to LaTeX code, hosted on HuggingSpace
Brief Details: T5-small model fine-tuned on Inspec dataset for scientific paper keyphrase generation. Optimized for academic abstract analysis with 0.32 F1@M score.
Brief Details: Chinese BART-base model (139M params) fine-tuned for text summarization on LCSTS dataset, developed by IDEA-CCNL for efficient Chinese text compression.
Brief Details: SBERT model trained on 20K UNO sustainable development goals dataset, optimized for sentence similarity with 768-dimensional vectors.
Brief Details: BERT-based model for fine-grained scientific document similarity in computer science, using contextual sentence vectors and co-citation training
Brief-details: A lightweight 68M parameter LLaMA-based model optimized for VLLM inference, featuring random initialization for experimental purposes
Brief-details: Optimized 8B parameter Llama 3.1 model with 4-bit quantization and unsloth acceleration, achieving 2.4x faster inference with 58% less memory usage
BRIEF-DETAILS: An 8B parameter LLaMA model optimized for code tasks, fine-tuned from mattshumer/Llama-3-8B-16K using Unsloth and TRL for 2x faster training speed.
BRIEF-DETAILS: A 7B parameter medical LLM optimized with AWQ 4-bit quantization, specialized for healthcare tasks with strong performance on medical reasoning benchmarks
Brief-details: BERT-based classification model for chatbot interactions, developed by smusa. Specialized for text classification tasks in conversational AI contexts.
BRIEF-DETAILS: Fine-tuned GEMMA 2B model specialized for IT customer support interactions, trained on Bitext Customer Support Dataset for enhanced response accuracy
Brief-details: MoviePortrait is a specialized image generation model focused on creating portrait-style images, triggered by the specific prompt prefix "In the style of OHWX"
Brief Details: Blazarot_blazaroshi is a specialized AI model designed for image generation, available on Civitai and hosted on Hugging Face, created by digiplay.
Brief Details: An XL-scale prediction model by Laxhar, hosted on HuggingFace. Designed for advanced prediction tasks with expanded capabilities and robust architecture.
Brief Details: Sentence embedding model that converts text to 384-dimensional vectors, optimized for ONNX runtime with both hidden state and pooler outputs
Brief Details: Anime-style image generation model optimized for cute faces and danbooru prompts. Features good NSFW capability with recommended Anything-V3.0 VAE.
Brief Details: DALLE2-PyTorch is an open-source PyTorch implementation of OpenAI's DALLE-2 text-to-image generation model, developed by LAION for research and experimentation.
Brief-details: Cross-encoder model optimized for MS Marco passage ranking, achieving 74.31 NDCG@10 on TREC DL 19, processes 960 docs/sec on V100 GPU
BRIEF DETAILS: 34B parameter reward model trained on Yi-34B-Chat, optimized for helpfulness and safety using Berkeley's Nectar dataset. Achieves 76.7% accuracy across preference metrics.
Brief-details: BERT-based multi-vector model for fine-grained computer science paper similarity using contextual sentence embeddings and Wasserstein Distance matching
