Chronos 13B V2
About
The Chronos 13B V2 is an advanced large language model adept at generating human-like text, excelling in chat, roleplay, and storywriting tasks. Based on the LLaMA v2 Base model architecture, it is particularly effective in producing coherent and lengthy outputs. Optimally performing with Alpaca formatting, it supports context lengths of up to 4096 tokens. Quantized versions of the model, in formats such as GGML and GGUF, offer flexibility between processing speed and accuracy. A notable variant, Chronos Hermes 13B V2, merges Chronos with Nous-Hermes-Llama2-13b models, enhancing coherency and creativity. However, its reliance on specific formatting may limit prompt flexibility, and higher-bit quantization can demand more computational resources. The model's language proficiency leans toward English, and it may carry biases from its training data.