# Instruction Tuning

Mistral Small 24B is a large language model developed by the Mistral AI team, featuring 24 billion parameters and supporting multilingual conversation and instruction handling. Through instruction tuning, it generates high-quality text content applicable in various scenarios like chat, writing, and programming assistance. Its key advantages include powerful language generation capabilities, multilingual support, and efficient inference. This model caters to individuals and businesses requiring high-performance language processing, offers an open-source license, supports local deployment and quantization optimizations, making it suitable for scenarios with data privacy requirements.

EXAONE-3.5-32B-Instruct is a series of instruction-tuned bilingual (English and Korean) generation models developed by LG AI Research, consisting of variants ranging from 2.4B to 32B parameters. These models support long context processing of up to 32K tokens and exhibit state-of-the-art performance in real-world use cases and long context understanding, while remaining competitive in general domains compared to similarly sized recently released models.

EXAONE-3.5-2.4B-Instruct-GGUF is a series of bilingual (English and Korean) generative models fine-tuned for instructions, developed by LG AI Research, with parameter sizes ranging from 2.4B to 32B. These models support long context processing of up to 32K tokens, demonstrating state-of-the-art performance in real-world use cases and long context comprehension, while remaining competitive in the general domain compared to similarly sized recently released models. The significance of this model lies in its optimization for deployment on small or resource-constrained devices, while delivering powerful performance.

MAmmoTH-VL is a large-scale multimodal reasoning platform that significantly enhances the performance of multimodal large language models (MLLMs) on various multimodal tasks through instruction tuning techniques. The platform has created a dataset consisting of 12 million instruction-response pairs using open models, covering a wide range of reasoning-intensive tasks and providing detailed and accurate reasoning steps. MAmmoTH-VL has achieved state-of-the-art performance on benchmarks such as MathVerse, MMMU-Pro, and MuirBench, showcasing its importance in education and research.

LLaVA-Video is a large multimodal model (LMM) focused on video instruction tuning, addressing the challenge of acquiring high-quality raw data from the internet by creating a high-quality synthetic dataset, LLaVA-Video-178K. This dataset includes detailed video descriptions, open-ended questions, and multiple-choice questions, aimed at enhancing the understanding and reasoning capabilities of video language models. The LLaVA-Video model has demonstrated outstanding performance across various video benchmarks, validating the effectiveness of its dataset.

Data-Juicer is a comprehensive multimodal data processing system aimed at delivering higher quality, richer, and more digestible data for large language models (LLMs). It offers a systematic and reusable data processing library, supports collaborative development between data and models, allows rapid iteration through a sandbox lab, and provides features like data and model feedback loops, visualization, and multidimensional automated evaluation, helping users better understand and improve their data and models. Data-Juicer is actively maintained and regularly enhanced with more features, data recipes, and datasets.

Llama3.1-8B-Chinese-Chat is an instruction-tuned language model based on the Meta-Llama-3.1-8B-Instruct architecture, designed for both Chinese and English users. It offers multiple capabilities such as role-playing and tool usage. The model is fine-tuned using the ORPO algorithm, significantly reducing the incidence of Chinese questions answered in English and minimizing language mixing, particularly excelling in role-playing, functionality invocation, and mathematical reasoning.

Llama3.1-70B-Chinese-Chat is an instruction-tuned language model based on the Meta-Llama-3.1-70B-Instruct model, specifically designed for Chinese and English bilingual users, with diverse capabilities such as role-playing and tool utilization. This model is fine-tuned using the ORPO algorithm, significantly reducing instances where Chinese questions receive English answers and cases of mixed language responses, particularly showing notable improvements in role-playing, function invocation, and mathematical abilities.

Gemma-2-27B-Chinese-Chat is the first instruction-tuned language model based on google/gemma-2-27b-it, designed specifically for Chinese and English users. It possesses various capabilities, including role-playing and tool usage. Fine-tuned using the ORPO algorithm, the model has significantly improved performance in Chinese and English conversations, role-playing, and mathematical calculations.

MG-LLaVA is a machine learning language model (MLLM) designed to enhance the visual processing capabilities of models. It achieves this by incorporating a multi-granularity visual pipeline, encompassing low-resolution, high-resolution, and object-centric features. An additional high-resolution visual encoder is introduced to capture finer details, and a Conv-Gate fusion network is used to integrate these high-resolution features with the base visual features. Furthermore, object-level features derived from offline detector bounding boxes are integrated to further refine the model's object recognition abilities. Trained via instruction tuning on publicly available multimodal data, MG-LLaVA exhibits exceptional perceptual skills.

Mistral-7B-Instruct-v0.2 is a large language model fine-tuned with instructions based on the Mistral-7B-v0.2 model. It features a 32k context window and a 1e6 Rope Theta value. The model can generate corresponding text outputs according to given instructions, supporting various tasks such as Q&A, writing, and translation. Through instruction-tuning, the model can better understand and execute instructions. Although the model currently lacks a targeted review mechanism, it will continue to be optimized to support deployment in more scenarios.

Gemma-7B-IT is an instruction tuning model developed by Google with a parameter count of 7B, utilizing the Gemini architecture and designed to enhance mathematical abilities, logical reasoning, and code generation. This model can run on a standard laptop without the need for substantial AI computing power, making it suitable for various application scenarios.