# Multimodal generation

OmniGen2 is a powerful and efficient unified multimodal model composed of a 3B vision-language model and a 4B diffusion model, supporting visual understanding, text-to-image generation, instruction-guided image editing, and context generation.

BLIP is a Transformer-based image-to-text generation model that can generate natural language descriptions for input images.

SkyReels V2 is the first open-source video generation model adopting an autoregressive diffusion forced architecture, supporting unlimited-length movie generation, achieving state-of-the-art performance among public models.

A speech-language model based on Qwen2.5-7B extension, supporting speech-text interleaved training and cross-modal generation

4-bit quantized version based on Qwen2-VL-7B model, trained using Unsloth and Huggingface TRL library, achieving 2x inference speedup

Brazilian Portuguese image captioning model based on Swin Transformer and DistilBERTimbau

Show-o is a PyTorch-based any-to-any conversion model focused on multimodal task processing.

Show-o is an any-to-any conversion model based on PyTorch, supporting input and output conversion across multiple modalities.

MGM-7B is an open-source multimodal chatbot trained on Vicuna-7B-v1.5, supporting high-definition image understanding, reasoning, and generation.

A text-to-video generation model based on the AllenNLP library, capable of generating corresponding video content according to the input text description.

CheXagent is a foundational model for chest X-ray interpretation, which can assist the medical field in professionally interpreting chest X-ray images.

A Persian image captioning model based on ViT+RoBERTa architecture, specifically designed to generate Persian text descriptions from images

BLIP model fine-tuned on HL Narratives dataset for generating high-level narrative image descriptions

This model is an image captioning model based on the BLIP architecture, specifically fine-tuned for high-level scene descriptions.

This model can convert text descriptions into video content and is suitable for various creative and automated scenarios.

This is a text-to-video model that can convert the input text description into corresponding video content.

Text2Video-Zero is a zero-shot text-to-video tool supporting edge guidance and mystical style

This is an image captioning model trained on a translated version (English-Russian) of the COCO2014 dataset, capable of generating Russian descriptions for input images.

A French incremental pre-trained language model based on GPT-fr with text-to-image generation capabilities

MolT5 is a large language model for translation between molecules and natural language, built on the T5 architecture.