# Unsupervised learning

This is a transformers model for zero-shot classification tasks, which can classify text without a large amount of labeled data.

HVI-CIDNet is a deep learning model designed for low-light image enhancement, based on a novel HVI color space.

An 8B-parameter unaligned beta model based on the LLaMA-3 architecture, offering multiple quantization versions to suit different hardware needs

Grayscale underwater fish detection model trained on YOLO11n architecture using semi-supervised learning techniques

Depth Anything is a depth estimation model that leverages large-scale unlabeled data to enhance performance, suitable for monocular depth estimation tasks.

DPT model using DINOv2 as backbone for depth estimation tasks.

This is an unsupervised SimCSE-based Japanese sentence embedding model, specifically designed for generating high-quality Japanese sentence embeddings.

A summarization model based on the Bart-Large architecture, specifically designed to accurately summarize research paper content into a format suitable for slide presentations.

A zero-shot text classification model trained with Self-Supervised Tuning (SSTuning), based on ALBERT-xxlarge-v2 architecture, which can be directly applied to tasks like sentiment analysis and topic classification without fine-tuning.

A zero-shot text classification model based on sentence-transformers, achieving text classification without labeled data through implicit training

Portuguese zero-shot classification model based on DeBERTa-v3 architecture, suitable for performing text classification tasks without fine-tuning

This is a multilingual sentence embedding model based on the ConGen framework, which maps sentences to a 768-dimensional vector space, suitable for tasks such as semantic search.

This is a Thai sentence similarity model based on the RoBERTa architecture, capable of mapping sentences into a 768-dimensional vector space, suitable for tasks like semantic search.

Experimental classification model based on fastText word vectors, supporting zero-shot classification tasks

A text summarization model fine-tuned on the CiteSum dataset based on facebook/bart-large, specifically designed for generating title-style summaries of scientific literature.

A PyTorch-based image classification model for detecting anomalies in images.

A GPT-2 model trained from scratch on the OSCAR Spanish corpus, supporting Spanish text generation tasks

TILDE is a model based on the BERT architecture, mainly used for text retrieval and language modeling tasks.

FlauBERT is a French BERT model trained on a large-scale French corpus, developed by the French National Center for Scientific Research.

A Wav2Vec2 base model pre-trained on 24.1k hours of unlabeled English data from the VoxPopuli corpus, suitable for speech recognition tasks.

A chemical domain BERT model trained based on TSDAE (Transformer-based Sequential Denoising Auto-Encoder), focusing on sentence similarity tasks

This is a model based on zero-shot classification technology, capable of classifying text without specific training.

SBERT-base-nli-v2 is a transformer-based sentence embedding model specifically designed for sentence similarity calculation and semantic search tasks.