Brief-details: Advanced text embedding model with 335M parameters, achieving SOTA performance on MTEB benchmark. Specialized for semantic similarity and retrieval tasks.
Brief-details: Microsoft's efficient 3.8B parameter LLM optimized for instruction following & reasoning. Supports 4K context, trained on 4.9T tokens with strong math/logic capabilities.
Brief-details: A production-ready text embedding model with 137M parameters, featuring Matryoshka architecture for flexible dimensionality (64-768) and long context support up to 8192 tokens.
Brief-details: Multi-language BERT model fine-tuned for sentiment analysis on product reviews, supporting 6 languages with 1.5M+ downloads and high accuracy rates
BRIEF DETAILS: A distilled RoBERTa model fine-tuned for financial sentiment analysis, achieving 98.23% accuracy on financial news classification with 82.1M parameters.
BRIEF DETAILS: Microsoft's LayoutLMv3 is a 125M-parameter multimodal Transformer for Document AI, combining text and image processing for document understanding tasks.
Brief-details: E5-large-v2 is a 335M parameter transformer model for text embeddings, featuring 24 layers and 1024d embeddings, optimized for semantic search and similarity tasks.
Brief-details: BERT large cased model (335M params) trained on BookCorpus and Wikipedia. Optimized for masked language modeling with 24 layers and 16 attention heads.
RMBG-1.4: State-of-the-art background removal model with 44.1M parameters. Trained on 12,000+ professional images, supports multiple categories including e-commerce and advertising.
BRIEF DETAILS: CodeBERT is Microsoft's pre-trained model for programming and natural languages, with 1.5M+ downloads. Built on RoBERTa for code understanding and generation.
Brief-details: Microsoft's DeBERTaV3 base model with 86M parameters, featuring ELECTRA-style pre-training and gradient-disentangled embedding sharing, achieving SOTA results on NLU tasks.
Brief Details: An 8B parameter Llama-3 variant optimized for instruction-following, featuring GGUF quantization with multiple precision options (2-8 bit) for efficient deployment
Brief-details: Google's advanced T5 v1.1 XXL model with GEGLU activation and improved architecture, pre-trained on C4 corpus for text-to-text tasks. Apache 2.0 licensed.
Brief-details: Cross-encoder model optimized for MS Marco passage ranking, achieving 73.04 NDCG@10 on TREC DL 19 with 2500 docs/sec processing speed on V100 GPU.
Brief Details: Qwen2.5-7B-Instruct-GGUF is a quantized instruction-tuned language model with 7.62B parameters, optimized for GGUF format and various bit precisions.
Brief-details: RoBERTa-based emotion classification model trained on Reddit data, supporting 28 emotion labels with strong performance on gratitude and amusement detection.
BRIEF-DETAILS: A 400M parameter multilingual sentence embedding model optimized for retrieval tasks, featuring flexible dimension options (512-8192) and simplified prompting
Brief Details: BLIP image captioning model by Salesforce - State-of-the-art vision-language model for generating image descriptions with 517 likes and 1.6M+ downloads
Brief-details: TinyLlama-1.1B-Chat is a compact 1.1B parameter LLM trained on 3T tokens, optimized for chat applications with Apache 2.0 license. Combines efficiency with Llama 2 architecture compatibility.
Brief Details: ESM-2 protein language model with 3B parameters and 36 layers. Trained on masked language modeling for protein sequence analysis. MIT licensed.
BRIEF-DETAILS: Multi-view diffusion model for 3D generation, based on MVDream architecture. Popular with 1.6M+ downloads. Supports diverse viewing angles for 3D content creation.
