# Open-domain QA

A fine-tuned version based on Gensyn/Qwen2.5-0.5B-Instruct, trained using the TRL framework and GRPO algorithm

This is a Mixture of Experts model obtained by merging multiple pre-trained language models using mergekit, aiming to create the most unconstrained text generation capability.

This model is a GGUF format model converted from Open-Orca/Mistral-7B-OpenOrca, suitable for text generation tasks.

A dialogue AI based on Causal Language Modeling (CLM) architecture, designed for natural language interaction, capable of generating coherent and contextually appropriate responses.

A 7-billion-parameter credibility-aware generation model fine-tuned on Mistral-7B, capable of understanding and utilizing contextual credibility for content generation.

StrangeMerges_17-7B-dare_ties is the result of merging two models, Gille/StrangeMerges_16-7B-slerp and Gille/StrangeMerges_12-7B-slerp, using the dare_ties merging method via LazyMergekit.

Blurdus-7b-v0.1 is a hybrid model obtained by merging three 7B-parameter models using LazyMergekit, demonstrating excellent performance across multiple benchmarks.

A dialogue model developed based on LLaMA 3B v2, supporting text generation tasks, with moderate performance on the open large model leaderboard.

This is a multimodal model published via arXiv paper 2312.17090, potentially capable of text and visual processing

A 7B-parameter large language model based on Llama 2 architecture, optimized through DPO training, supporting Chinese and English text generation

IDEFICS is an open-source reproduction of DeepMind's proprietary visual language model Flamingo. It is a multimodal model that can accept arbitrary sequences of images and text as input and generate text output.

orca_mini_13b is a text generation model trained on multiple high-quality datasets, focusing on instruction following and dialogue tasks.

CAMEL-33B is a large language model fine-tuned on LLaMA-33B, integrating CAMEL framework dialogue data, ShareGPT public dialogues, and Alpaca instruction data, excelling in multi-turn conversations and instruction comprehension.

InstructBLIP is a vision instruction-tuned version based on BLIP-2, using Vicuna-7B as the language model, focusing on vision-language tasks.

An open-domain question answering model trained on the complete Natural Questions (NQ) dataset, capable of answering various factual questions

A dense retriever based on BERT-base architecture, trained on Wikipedia articles to emulate BM25 behavior

A versatile large language model capable of handling various natural language processing tasks (inferred information)

This model is a high-performance open-domain chatbot capable of integrating multiple dialogue skills such as questioning, answering, knowledge demonstration, and empathy.

DPR (Dense Passage Retrieval) is a tool and model for open-domain question answering research. This model is a BERT-based question encoder trained on the Natural Questions (NQ) dataset.

BERT-based Dense Passage Retrieval (DPR) question encoder for open-domain QA research, trained on multiple QA datasets

KoGPT2 is a Korean GPT-2 model developed by SKT-AI, based on the Transformer architecture, suitable for various Korean text generation tasks.

SPARTA is an efficient open-domain QA model based on sparse Transformer matching retrieval, designed for information retrieval tasks.