# Low resource consumption

Qwen3 30B A3B is a large language model that has undergone 8-bit quantization using the GPTQ method, suitable for efficient inference scenarios.

A lightweight Russian-English bilingual language model fine-tuned based on SmolLM2-135M, with enhanced Russian processing capabilities but notable limitations

An INT4 weight quantization version based on openai/whisper-large-v3-turbo, supporting efficient audio-to-text tasks

This model is converted from google/gemma-3-12b-it to GGUF format, suitable for the llama.cpp framework.

An efficient text-to-image AI model that generates high-quality images in just 4 steps, supporting multiple hardware-optimized formats.

An image watermark processing model based on SDXL-Turbo, focusing on image-to-image transformation tasks.

Phi-3-mini-4k-instruct-graph is a fine-tuned version of Microsoft's Phi-3-mini-4k-instruct, specifically designed for entity relationship extraction from general text data, aiming to achieve comparable quality and accuracy to GPT-4 in generating entity relationship graphs.

This is a CTranslate2 converted version of the OpenAI Whisper large-v3 model, specifically fine-tuned for Japanese with an additional 4,000 training steps, supporting multilingual speech recognition.

Gerbil-A-32m is an A-grade model with 32 million parameters, trained on 640 million tokens, suitable for various natural language processing tasks.

An AI illustration generation model trained using Dreambooth technology, suitable for text-to-image tasks

Vi-(Arcane)-v1 is a Dreambooth model trained by asukii using TheLastBen's fast DreamBooth notebook, suitable for text-to-image generation tasks.

This is a binary classification model trained via AutoNLP for IMDB review sentiment analysis