# Low VRAM optimization

Qwen2.5-Omni is an end-to-end multimodal model capable of perceiving multiple modalities including text, images, audio, and video, while generating text and natural speech responses in a streaming manner.

The first real-time high-quality video generation model based on DiT architecture, capable of generating 1216×704 resolution videos at 30fps

A standard PEFT LoRA model trained on fal/AuraFlow, specializing in text-to-image and image-to-image generation tasks for Domo-kun characters.

FLUX is a text-to-image generation model that integrates 5 high-quality fine-tuned adapters, capable of producing images in various styles

Wan2.1-Fun-1.3B is a text-to-video generation model that supports multi-resolution training and first/last frame prediction.

This is a LoRA adapter based on the Wan2.1-T2V-14B model, designed for generating origami-style videos.

A fine-tuned version based on microsoft/phi-3-mini-4k-instruct, specifically designed for role-playing dialogues with various characters

Wan2.1-Fun-1.3B is a text-to-video generation model developed by Alibaba PAI team, supporting multi-resolution training and first/last frame prediction.

A LoRA model trained on Steamboat Willie animation clips for generating text-to-video content in golden age animation style

CogView4-6B is a text-to-image model based on the GLM-4-9B foundation model, supporting both Chinese and English, capable of generating high-quality images.

AWQ quantized version of DeepSeek R1 model, optimized for float16 overflow issues and supports efficient inference deployment

The 8-bit quantized version of OmniGen-v1, suitable for text-to-image and image-to-image tasks, supporting multimodal input.

Stable Diffusion 3.5 Large Model is a multimodal diffusion transformer (MMDiT) text-to-image model, with significant improvements in image quality, text layout, complex prompt understanding, and resource efficiency.

A LyCORIS adapter based on FLUX.1-dev, specializing in text-to-image generation tasks, particularly suitable for work environments.

A text-to-image model formed by merging Schnell, fine-tuned Dev, and Hyper, recommended steps 4-8, with significant quality improvement at 4 steps

This is a quantized version of the NeuralDaredevil-8B-abliterated model, providing model files of various quantization types, suitable for users with different hardware conditions and requirements.

ALMA-7B-Ja-V2 is a machine translation model supporting Japanese-English bidirectional translation, with improved performance through additional training on the previous version.

Ether Real Mix is a text-to-image generation model based on stable diffusion technology, focusing on generating high-quality, artistic-style images.

RWKV is a large language model combining the advantages of RNN and Transformer, supporting efficient training and fast inference with unlimited context processing capability.

An anime-style text-to-image model fine-tuned on Stable Diffusion, specializing in 'Cool Japan' cultural content

GPT-J 6B fine-tuned model for Python code generation, specialized in Python programming assistance

This is a 6B-parameter Vietnamese causal language model based on GPT-J architecture, specifically trained for Vietnamese news content.