# 🔬 Essential dev kit

## 🛠️ AI Tooling: Essential Stack

### ☁️ Cloud-Based Tools (Quick Start, Pay-as-you-go)

| Tool            | Use Case                      |
| --------------- | ----------------------------- |
| Google Colab    | Notebook prototyping w/ GPU   |
| AWS SageMaker   | Production-grade model ops    |
| Vertex AI (GCP) | Enterprise ML Ops & pipelines |
| Hugging Face    | Hosting, Transformers, Spaces |
| Kaggle Kernels  | Lightweight GPU compute       |

***

## 🔌 Hardware & GPU Drivers

### For Training & Inference:

* **NVIDIA CUDA Toolkit** *(Recommended)*\
  Industry standard for GPU-accelerated training.
* **AMD ROCm** *(Experimental but improving)*\
  Open alternative with growing support.
* **OpenCL / OpenCUDA**\
  Some success on edge devices and less common setups.

***

### Guidelines by Use Case:

* **LLMs:**\
  20–30GB VRAM or advanced offloading techniques required.
* **Vision Models:**
  * **Images:** 12–32GB VRAM
  * **Video:** 48–100GB VRAM for efficient processing pipelines.

***

### Edge/DIY Builds:

* **Raspberry Pi Cluster**\
  CPU-based inference and experimentation.
* **Jetson Orin / Xavier**\
  NVIDIA-powered edge inference solutions.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://doc.guidry-cloud.com/dev-kit-minimal.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.