Update Modelfile + README (MDM/Intune/Jamf/ABM/Autopilot/Graph/Cisco/.NET/Java)

Modelfile

@@ -1,6 +1,6 @@
-from hf://miladatefi/aaai-7b-dapt-gguf/aaai-7b-dapt-q8_0.gguf
+FROM /path/you/want/locally/aaai-7b-dapt-q8_0.gguf
 
-template """{{- if .System }}<|im_start|>system
+TEMPLATE """{{- if .System }}<|im_start|>system
 {{ .System }}<|im_end|>
 {{- end }}{{- range .Messages }}
 <|im_start|>{{ .Role }}
@@ -9,17 +9,16 @@ template """{{- if .System }}<|im_start|>system
 <|im_start|>assistant
 """
 
-system "You are **AAAI by Milad Atefi** (EN/DE). Focus: Intune, Jamf Pro, ABM/ASM, Windows Autopilot, Microsoft 365/Entra/Defender, PowerShell, bash/zsh, .NET backend. Be precise, practical, and prefer vendor-documented commands. If off-topic for enterprise device management / admin IT, reply exactly: **Out of scope.**"
+SYSTEM "You are AAAI by Milad Atefi (EN/DE). Focus: Intune, Jamf Pro, ABM/ASM, Windows Autopilot, Microsoft 365/Entra/Defender, PowerShell, MS Graph, Cisco networking, .NET, Java (admin tooling). If off-topic for enterprise device management / admin IT, reply exactly: Out of scope."
 
-parameter stop "<|im_end|>"
-parameter stop "<|im_start|>"
+PARAMETER stop "<|im_end|>"
+PARAMETER stop "<|im_start|>"
+PARAMETER num_ctx 8192
+PARAMETER temperature 0.1
+PARAMETER top_p 0.9
+PARAMETER repeat_penalty 1.15
 
-parameter num_ctx 8192
-parameter temperature 0.1
-parameter top_p 0.9
-parameter repeat_penalty 1.15
-
-message user "What's your name?"
-message assistant "AAAI by Milad Atefi."
-message user "Write a love poem about summer."
-message assistant "Out of scope."
+MESSAGE user "What's your name?"
+MESSAGE assistant "AAAI by Milad Atefi."
+MESSAGE user "Write a love poem about summer."
+MESSAGE assistant "Out of scope."

README.md

@@ -1,7 +1,5 @@
 ---
-language:
-- en
-- de
+language: [en, de]
 tags:
 - gguf
 - ollama
@@ -11,8 +9,14 @@ tags:
 - autopilot
 - apple-business-manager
 - microsoft-365
+- entra-id
+- defender
 - powershell
+- ms-graph
+- cisco
+- networking
 - dotnet
+- java
 base_model: Qwen/Qwen2.5-7B-Instruct
 license: other
 ---
@@ -23,20 +27,24 @@ license: other
 © atefi.ch — Validate in a lab before production.
 
 ## What it is
-Domain-adapted **enterprise device management** assistant (EN/DE). Focus areas:
-- **Microsoft Intune**: enrollment, compliance, configuration profiles, **Conditional Access**, App Protection Policies, **Windows Autopilot**
-- **Jamf Pro**: macOS/iOS/iPadOS configuration profiles, PPPC, restrictions, scripts, packaging
-- **Apple**: Apple Business/School Manager, Automated Device Enrollment, payload keys
+**AAAI** is a domain-adapted assistant (EN/DE) for **enterprise device management & admin IT**:
+
+- **Microsoft Intune**: enrollment, compliance, config profiles, **Conditional Access**, App Protection Policies, **Windows Autopilot**
+- **Jamf Pro**: macOS/iOS/iPadOS configuration profiles (PPPC, restrictions), scripts, packaging
+- **Apple**: ABM/ASM, **ADE**, payload keys
 - **Microsoft 365 / Entra ID / Defender**
-- **Scripting**: PowerShell, bash/zsh
-- **Backend patterns**: .NET web/API scaffolding
+- **Scripting**: **PowerShell**, bash/zsh, **Microsoft Graph**
+- **Networking**: **Cisco IOS/IOS-XE** core admin topics
+- **Backend**: **.NET** (Minimal APIs, admin portals), **Java** (admin tooling)
 
 **Base:** `Qwen/Qwen2.5-7B-Instruct`  
-**This repo:** **GGUF** build for **Ollama** (`Q8_0` ≈ 8 GB).  
-**Training:** Uses **my custom adapter** merged into base (data I collected on the above domains).
-
----
-
-## Quickstart (Ollama)
+**Training:** Uses **my custom adapter** (trained on my collected data in the above domains) **merged** into the base before quantization.  
+**This repo:** **GGUF** build for **Ollama** (`Q8_0` ≈ 8 GB).
 
-Install & create the model:
+## Quickstart (local GGUF)
+```bash
+# 1) place your file at /tmp/aaai.gguf (≈8.1 GB, starts with GGUF)
+# 2) Modelfile -> FROM /tmp/aaai.gguf  (see repo Modelfile for template/settings)
+ollama rm aaai 2>/dev/null || true
+ollama create aaai -f Modelfile
+echo "What's your name?" | ollama run aaai