LLaMA2

wikipedia:LLaMA2 pretrained models are trained on 2 trillion tokens.

• https://ai.meta.com/llama/

• ollama run llama2

Related

• q4
• Ollama uses 4-bit quantization

See also

• Ollama, [ show | server | run | rm ]
• LLaMA, LLaMA2, Llama 3-V 405B
• Meta AI, LLaMA3
• LLM, MLLM, LoRA, LLaMA, LLaMA3, QLoRA, Falcon, PaLM 2, Gemini, Mixtral 8x7B, BitNet, Measuring Massive Multitask Language Understanding (MMLU), NVLM, LangChain, LlamaIndex, MCP, vLLM, DeepEval, NotebookLM, Costory.io