# Open-source Model

FairyR1-32B is a large language model with 32B parameters, developed by PKU-DS-LAB, focusing on text generation tasks.

DMind-1 is a text generation foundation model dedicated to the free dissemination of knowledge.

Cosmos-Reason1-7B is a 7B-parameter foundational model released by NVIDIA, specializing in image-to-text tasks.

This is a transformers model published on the Hugging Face Hub. The specific functions and uses are to be supplemented.

OLMo-2-0425-1B-Instruct is a 1-billion-parameter instruction-finetuned language model developed by AllenAI, focused on text generation tasks.

GLM-4-32B-0414-abliterated is a large-scale language model based on the GLM architecture, with a parameter scale of 32B, suitable for text generation tasks.

A lightweight version of the Qwen2.5 large language model released by Alibaba PAI, focusing on efficient text generation tasks

DeepSeek-V3-0324 is a powerful foundational model focused on text generation tasks, designed to deliver high-quality text generation capabilities.

GGUF quantized version of Stable Diffusion 2.1, supporting multiple quantization precisions, suitable for text-to-image generation tasks.

Llama-3 is a large language model based on the Transformer architecture, supporting English language processing.

Llama 3 is the next-generation open-source large language model introduced by Meta, featuring enhanced performance and broader application scenarios.

Lawyer LLaMA is a large language model based on the LLaMA architecture, specializing in Chinese legal tasks.

LLaVA-Next-Video is an open-source multimodal dialogue model, fine-tuned with multimodal instruction-following data on large language models, supporting video and text multimodal interactions.

SPLADE-v3-Doc is the document version of the SPLADE model, focusing on document-side inference and suitable for scenarios such as information retrieval.

Vikhr is a bilingual (Russian and English) open-source instruction-following large language model, specifically designed for Russian-speaking users, capable of conversation and task assistance.

LLaVA is an open-source multimodal chatbot, fine-tuned on large language models with multimodal instruction-following data.

Openchat 3.5 0106 is an open-source dialogue model based on the Mistral architecture, developed by the OpenChat team. This model focuses on delivering high-quality conversational experiences and supports various dialogue tasks.

Large-scale vision-language model based on Vision Transformer architecture, supporting zero-shot image classification tasks

Dolphin 2.7 Mixtral 8X7B is a large language model based on the Mixtral architecture, focusing on code generation and instruction-following tasks.

LLaVA is an open-source multimodal chatbot, based on the LLaMA/Vicuna architecture, fine-tuned with multimodal instruction-following data.

LLaVA is an open-source multimodal chatbot, fine-tuned on GPT-generated multimodal instruction data based on the LLaMA/Vicuna model.

ToRA-7B is a 7B-parameter tool-integrated reasoning agent, specifically designed for solving mathematical problems by combining natural language reasoning with external tool usage.

This model is obtained by adapting bloom-1b7 to Italian, using the MAD-X language adaptation strategy to adjust the original BLOOM model.

A 125M-parameter language model fine-tuned on the Super-Natural Instructions dataset, suitable for various natural language processing tasks.

Vicuna is a chat assistant fine-tuned based on Llama 2, with training data collected from user-shared conversations on ShareGPT.

FLAN-T5 is a language model optimized through instruction fine-tuning based on the T5 model, supporting multilingual task processing and outperforming the original T5 model with the same parameter count.

A Russian-to-English machine translation model developed by the Language Technology Research Group at the University of Helsinki, based on the Transformer-align architecture.

A Chinese-to-English machine translation model developed by the University of Helsinki, based on the OPUS corpus