

Aria Base 64K
Overview :
Aria-Base-64K is one of the foundational models in the Aria series, designed for research purposes and further training. It emerged after the long text pre-training phase, trained on 33 billion tokens (21 billion multimodal and 12 billion language tokens, with 69% being long texts). It is suitable for further pre-training or fine-tuning on long video question answering datasets or long document question answering datasets, even in resource-constrained environments, through post-training with short instruction tuning datasets tailored for long text scenarios. The model can comprehend up to 250 high-resolution images or up to 500 medium-resolution images, maintaining robust foundational performance in both language and multimodal contexts.
Target Users :
The target audience includes researchers and developers, particularly professionals who need to handle long texts and multimodal datasets. Aria-Base-64K offers a powerful pre-trained model suitable for scenarios such as video question answering and long document question answering, helping to enhance processing efficiency and accuracy.
Use Cases
- Develop a video question answering system using Aria-Base-64K to enhance understanding of video content.
- Apply Aria-Base-64K to long document question answering, improving efficiency in document retrieval and comprehension.
- Utilize Aria-Base-64K for joint inference of images and text to develop new multimodal applications.
Features
- Long text pre-training: Trained on 33 billion tokens, ideal for further pre-training or fine-tuning on long video question answering and long document question answering datasets.
- Multimodal understanding: Capable of understanding up to 250 high-resolution images or up to 500 medium-resolution images.
- Strong foundational performance: Maintains the same robust foundational performance as Aria-Base-8K across language and multimodal scenarios.
- Low ratio chat template training: Trained using only about 3% of the data in a chat template format, which may not be suitable for direct chat template usage.
- Quick start support: Provides quick installation and inference code examples for users to rapidly begin using the model.
- Advanced inference and fine-tuning: Offers a code repository that supports more advanced inference, examples, and fine-tuning on custom datasets.
How to Use
1. Install necessary libraries: Use pip to install transformers, accelerate, sentencepiece, and other libraries.
2. Load the model: Load the Aria-Base-64K model using AutoModelForCausalLM.from_pretrained.
3. Process input: Use AutoProcessor.from_pretrained to process text and image inputs.
4. Perform inference: Pass the processed inputs to the model to execute generation tasks.
5. Decode output: Use the processor to decode the tokens output by the model to obtain the final results.
6. Advanced usage: For more advanced inference and fine-tuning, access the code repository on GitHub.
Featured AI Tools

Gemini
Gemini is the latest generation of AI system developed by Google DeepMind. It excels in multimodal reasoning, enabling seamless interaction between text, images, videos, audio, and code. Gemini surpasses previous models in language understanding, reasoning, mathematics, programming, and other fields, becoming one of the most powerful AI systems to date. It comes in three different scales to meet various needs from edge computing to cloud computing. Gemini can be widely applied in creative design, writing assistance, question answering, code generation, and more.
AI Model
11.4M
Chinese Picks

Liblibai
LiblibAI is a leading Chinese AI creative platform offering powerful AI creative tools to help creators bring their imagination to life. The platform provides a vast library of free AI creative models, allowing users to search and utilize these models for image, text, and audio creations. Users can also train their own AI models on the platform. Focused on the diverse needs of creators, LiblibAI is committed to creating inclusive conditions and serving the creative industry, ensuring that everyone can enjoy the joy of creation.
AI Model
6.9M