InternLM

InternLM is a family of downloadable language models for developers and researchers building AI assistants or adapting models to their own applications. It includes base models for further training and chat models tuned to follow instructions.

• Reasoning modes: InternLM3-8B-Instruct supports ordinary conversational responses and a deep thinking mode for more complex reasoning tasks.
• Chat and function calling: InternLM2.5-Chat combines instruction following with conversation and function calling for applications that connect a model to other tools.
• Long documents: InternLM2.5-7B-Chat-1M accepts up to one million tokens of context, allowing an application to provide lengthy documents or reference material in a single request.
• Model adaptation: Base models provide a starting point for customization, while InternLM2-Reward models assess response preferences for reinforcement learning workflows. InternLM2 also supports coding tasks.

The models work with Hugging Face Transformers and use Python and PyTorch. The family includes smaller and larger variants, so choosing it involves selecting a model as well as the software that will run it. The repository uses the Apache 2.0 license.