Merge branch 'vllm_doc_update_121' of github.com:terrastackai/terratorch into vllm_doc_update_121

Author: romeokienzler

docs/guide/vllm/intro.md

@@ -7,24 +7,28 @@ can be served with vLLM.
 
 TerraTorch models can be served with vLLM in _tensor-to-tensor_ or
 _image-to-image_ mode. The tensor-to-tensor mode is the default mode and is
-natively enabled by vLLM. For the image-to-image mode instead, TerraTorch uses a
-feature in vLLM called
+natively enabled by vLLM. For the image-to-image mode, TerraTorch uses a feature
+in vLLM called
 [IOProcessor plugins](https://docs.vllm.ai/en/v0.13.0/design/io_processor_plugins/#writing-an-io-processor-plugin),
 enabling processing and generation of data in any modality (e.g., geoTiff). In
-TerraTorch we provide pre-defined IOProcessor plugins, check the list
-[here](./vllm_io_plugins.md#available-terratorch-ioprocessor-plugins).
+TerraTorch, we provide pre-defined IOProcessor plugins. Check the list of
+[available plugins](./vllm_io_plugins.md#available-terratorch-ioprocessor-plugins).
 
-To enable your model to be served via vLLM, follow the below steps:
+To enable your model to be served via vLLM, follow the steps below:
 
-1. Verify the model you want to serve is either already a core model, or learn
-   how to [add your model to TerraTorch](../models.md#adding-a-new-model).
-2. [Prepare your model for serving with vLLM](./prepare_your_model.md).
-3. If serving in image-to-image mode
-   [Learn about IOProcessor plugins](./vllm_io_plugins.md), identify an existing
-   one suiting your model or
+1. **Ensure TerraTorch Integration**: Verify the model you want to serve is
+   either already a core model, or learn how to
+   [add your model to TerraTorch](../models.md#adding-a-new-model).
+2. **Create a Model _config.json_**: Create a
+   [vLLM compatible _config.json_](./prepare_your_model.md).
+3. **Determine IOProcessor Plugin Needs**: If serving in image-to-image mode,
+   identify an [IOProcessor plugin](./vllm_io_plugins.md) that suits your model
+   or
    [build one yourself](https://docs.vllm.ai/en/latest/design/io_processor_plugins/).
-4. Ensure your model weights and config.json are either hosted on Hugging Face,
-   or stored in a local directory and accessible by your vLLM instance.
-5. Start a vLLM serving instance that loads your model and perform an inference
-   in [tensor-to-tensor mode](./serving_a_model_tensor.md) or in
-   [image-to-image mode](./serving_a_model_image.md).
+4. **Make your Model Accessible to vLLM**: Host your model weights and
+   config.json on Hugging Face, or store them in a local directory accessible by
+   the vLLM instance.
+
+To validate the steps above, start a vLLM serving instance that loads your model
+and perform an inference in [tensor-to-tensor mode](./serving_a_model_tensor.md)
+or in [image-to-image mode](./serving_a_model_image.md).