documentation on the model format

README.md

@@ -9,10 +9,28 @@
 
 [![Discord](https://dcbadge.vercel.app/api/server/ollama?style=flat&compact=true)](https://discord.gg/ollama)
 
-Create, run, and share large language models (LLMs). Ollama bundles a model’s weights, configuration, prompts, and more into self-contained packages that can run on any machine.
-
 > Note: Ollama is in early preview. Please report any issues you find.
 
+Create, run, and share portable large language models (LLMs). Ollama bundles a model’s weights, configuration, prompts, and more into self-contained packages that can run on any machine.
+
+### Portable Large Language Models (LLMs)
+
+Package models as a series of layers in a portable, easy to manage format.
+
+#### The idea behind Ollama
+
+- Universal model format that can run anywhere: desktop, cloud servers & other devices.
+- Encapsulate everything a model needs to operate – weights, configuration, and data – into a single package.
+- Build custom models from base models like Meta's [Llama 2](https://ai.meta.com/llama/)
+- Share large models without having to transmit large amounts of data.
+
+<picture>
+  <source media="(prefers-color-scheme: dark)" height="480" srcset="https://github.com/jmorganca/ollama/assets/251292/2e05cf23-e3c6-403e-9910-3d622801f4b8">
+  <img alt="logo" height="480" src="https://github.com/jmorganca/ollama/assets/251292/2e05cf23-e3c6-403e-9910-3d622801f4b8">
+</picture>
+
+This format is inspired by the [image spec](https://github.com/opencontainers/image-spec) originally introduced by Docker for Linux containers. Ollama extends this format to package large language models.
+
 ## Download
 
 - [Download](https://ollama.ai/download) for macOS on Apple Silicon (Intel coming soon)