# Low VRAM Requirement

A quantized version of the DeepSeek R1 0528 model from DeepSeek AI, an autoregressive language model based on an optimized Transformer architecture, which can be used for commercial and non-commercial purposes.

The DeepSeek-R1-0528 model after quantization processing significantly reduces the requirements for GPU memory and disk space by quantizing the weights to the INT4 data type.

Wan2.1 is an open and advanced foundational model for video generation, supporting various video generation and editing tasks.

A losslessly compressed version of stabilityai/stable-diffusion-3.5-large using DFloat11 format, reducing size by 30% while maintaining 100% accuracy

Qwen3-14B-FP8-dynamic is an optimized large language model. By quantizing activation values and weights to the FP8 data type, it effectively reduces GPU memory requirements and improves computational throughput.

Wan2.1 is an open and advanced large-scale video generation model with top-tier performance, capable of running on consumer-grade GPUs and excelling in multitask processing.

DeepCoder-14B-Preview is a code generation model developed based on DeepSeek-R1-Distill-Qwen-14B, focusing on solving verifiable programming problems.

The GGUF quantized version of Lumina is a model specifically designed for generating high-quality images, supporting text-prompt-based generation of highly matched images.

Quantized version of DeepSeek-R1-Distill-Qwen-32B, reducing memory requirements and improving computational efficiency through INT8 weight quantization and activation quantization

The FP8 quantized version of DeepSeek-R1-Distill-Llama-70B, which optimizes inference performance by reducing the number of bits of weights and activations.

Quantized version based on PixArt-alpha/PixArt-XL-2-1024-MS, supporting efficient text-to-image tasks

The GGUF quantized version of Stable Diffusion 3.5 Medium, suitable for text-to-image tasks and capable of running on legacy devices.

The GGUF quantized version of Stable Diffusion 3.5 Large Turbo, suitable for image generation tasks, providing more efficient runtime performance.

This is a text-to-image generation model combining Hands XL, SD 1.5, and FLUX.1-dev technologies, focusing on high-quality image generation.

FP8 quantized version of Meta Llama 3.1 8B Instruct model, featuring an optimized transformer architecture autoregressive language model with 128K context length support.

FLUX.1-dev is a text-to-image generation model quantized to INT4 format using Optimum Quanto, suitable for non-commercial use.

A quantized version of Meta-Llama-3.1-8B-Instruct, optimized to reduce disk space and GPU memory requirements, suitable for chat assistant scenarios in English business and research.

This is the INT4 quantized version of the Meta-Llama-3.1-8B-Instruct model, quantized using the GPTQ algorithm, suitable for multilingual dialogue scenarios.

FP8 quantized version of DeepSeek-Coder-V2-Lite-Instruct, suitable for commercial and research use in English, optimized for inference efficiency.

MaPO is a reference-free, energy-efficient, and memory-friendly alignment method for text-to-image diffusion models

KOALA-Lightning-700M is an efficient text-to-image model trained through knowledge distillation based on SDXL-Lightning, significantly improving inference speed while maintaining generation quality

A French Q&A model fine-tuned based on Llama-2-13b-chat, supporting tasks like Baroque style text generation