reallexi/lexi-rm-agent
The reallexi/lexi-rm-agent is a 0.5 billion parameter instruction-tuned language model developed by Reallexi LLC, based on the Qwen2.5-0.5B-Instruct architecture. This model is provided as a standalone export with its adapter merged into the base weights, eliminating the need for PEFT at runtime. It is designed for general language understanding and generation tasks, offering efficient performance due to its compact size.
Loading preview...
Overview
The reallexi/lexi-rm-agent is a compact, instruction-tuned language model developed by Reallexi LLC. It is built upon the Qwen/Qwen2.5-0.5B-Instruct base architecture and features 0.5 billion parameters with a 32768 token context length. This model is distributed as a standalone export, meaning its fine-tuning adapter has been pre-merged into the base weights, simplifying deployment as no separate PEFT (Parameter-Efficient Fine-Tuning) adapter is required during runtime.
Key Characteristics
- Developer: Reallexi LLC
- Base Model: Qwen2.5-0.5B-Instruct
- Parameter Count: 0.5 billion parameters
- Context Length: 32768 tokens
- Deployment: Standalone export with merged adapter, requiring no PEFT at runtime.
Usage Considerations
This model is suitable for applications where a smaller, efficient language model is preferred, especially when computational resources are a concern. Its pre-merged adapter streamlines integration, making it ready for immediate use in various instruction-following tasks. Users should ensure that reallexi-model.json, NOTICE, and all applicable upstream license files are kept with the model for compliance.