tinyllama constant

ModelInfo const tinyllama

TinyLlama 1.1B Chat — lightweight model suitable for most phones.

~638 MB download, 2 GB RAM recommended, 2048 token context.

Implementation

static const tinyllama = ModelInfo(
  id: 'tinyllama',
  name: 'TinyLlama 1.1B Chat',
  fileName: 'tinyllama-1.1b-chat-v1.0.Q4_K_M.gguf',
  ggufUrl:
      'https://huggingface.co/TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF/resolve/main/tinyllama-1.1b-chat-v1.0.Q4_K_M.gguf',
  sizeMB: 638,
  minRamGB: 2,
  context: 2048,
  chatTemplate: _zephyrTemplate,
);