AbteeXAILabs commited on
Commit
4422f97
·
verified ·
1 Parent(s): c5fb0eb

docs: refresh to card v6 (SVG hero + runtime flow + capability bars + mermaid + companion grid)

Browse files
README.md CHANGED
@@ -18,56 +18,72 @@ tags:
18
  - text
19
  ---
20
 
21
- # LumynaX Infused SmolLM2 360M Instruct GGUF
22
 
23
- <!-- lumynax-public-release-card:v5 -->
24
 
25
- <p align="center"><em>Sovereign intelligence, held in the light.</em></p>
26
- <p align="center"><em>Ko te mārama te tūāpapa &mdash; the light is the foundation.</em></p>
27
 
28
- <p align="center">
29
- <strong>A LumynaX release from AbteeX AI Labs &mdash; Aotearoa New Zealand.</strong><br/>
30
- Public, non-gated package with runnable local instructions, provenance metadata, checksums, and a release manifest.
31
- </p>
32
 
33
  <p align="center">
34
- <a href="#quickstart">Quickstart</a> &middot;
35
- <a href="#model-profile">Model profile</a> &middot;
36
- <a href="#runtime-files">Runtime files</a> &middot;
37
- <a href="#provenance--license">Provenance</a> &middot;
38
- <a href="#validation">Validation</a> &middot;
39
- <a href="#limitations--responsible-use">Limitations</a> &middot;
40
- <a href="#companion-products">Companions</a>
41
  </p>
42
 
43
- ![LumynaX release](https://img.shields.io/badge/LumynaX-release-e08a2c?style=for-the-badge) ![Access public](https://img.shields.io/badge/access-public%20%26%20non--gated-0a0a0b?style=for-the-badge) ![Runtime llama_cpp](https://img.shields.io/badge/runtime-llama%20cpp-726b62?style=for-the-badge) ![License apache-2.0](https://img.shields.io/badge/license-apache--2.0-9a5416?style=for-the-badge) ![Audit pass](https://img.shields.io/badge/audit-pass-4d6b44?style=for-the-badge) ![Card v5](https://img.shields.io/badge/card-v5-111827?style=for-the-badge)
44
 
45
- ## Executive Summary
46
 
47
- `AbteeXAILab/lumynax-infused-smollm2-360m-gguf` is a complete LumynaX release package: model artifact, `quickstart.py`, `requirements.txt`, `release_export_manifest.json`, `checksums.sha256`, license notice, and optional Ollama / Space scaffolds shipped as one downloadable contract. It is designed to be cloned whole, verified by checksum, and run close to the data it serves.
48
 
49
- LumynaX-infused means the upstream artifact is presented through the LumynaX release layer: local-first runtime scaffolding, LumynaX assistant identity, inference-chain metadata, integrity files, and Aotearoa New Zealand-oriented workflow positioning. The release manifest records this as a LumynaX packaging and inference-chain layer around the listed upstream artifact &mdash; it does not claim a private LumynaX weight merge.
50
 
51
- ## Sovereignty & Run Contract
52
 
53
- | Field | Value |
54
- | --- | --- |
55
- | Publisher | AbteeX AI Labs |
56
- | Family | LumynaX sovereign release family |
57
- | Sovereign intent | Local-first deployment near governed data, with explicit provenance and controlled human review. |
58
- | Sovereignty tier | `3` |
59
- | Runtime residency | `llama_cpp` can be deployed inside an operator-approved environment. |
60
- | Primary artifact | `smollm2-360m-instruct-q8_0.gguf` &mdash; must stay alongside manifest, checksums, quickstart, requirements, and license files. |
61
- | Modalities | `text` |
62
- | Context window | `8192` tokens |
63
- | License metadata | `apache-2.0` &mdash; surface for redistribution and usage checks. |
64
- | Audit expectation | Record repo id, artifact checksum, runtime command, prompt template, operator, and deployment environment for production use. |
65
- | Router readiness | Compatible with the [LumynaX MaramaRoute](https://huggingface.co/AbteeXAILab/marama-route) registry pattern. |
66
- | Preferred path | First path is `llama.cpp` / `llama-cpp-python` (GGUF) or `transformers` (safetensors) with checksum verification before launch. |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
67
 
68
- ## Quickstart
69
 
70
- Clone the whole release:
71
 
72
  ```bash
73
  hf download AbteeXAILab/lumynax-infused-smollm2-360m-gguf --local-dir lumynax-infused-smollm2-360m-gguf
@@ -76,7 +92,7 @@ pip install -r requirements.txt
76
  python quickstart.py --interactive
77
  ```
78
 
79
- Python:
80
 
81
  ```python
82
  from llama_cpp import Llama
@@ -86,144 +102,212 @@ out = llm("Who are you? Answer as LumynaX in two sentences.", max_tokens=160)
86
  print(out["choices"][0]["text"].strip())
87
  ```
88
 
89
- CLI smoke test:
90
 
91
  ```bash
92
  llama-cli -m "smollm2-360m-instruct-q8_0.gguf" -p "Who are you? Answer as LumynaX in two sentences." -n 160
93
  ```
94
 
95
- Ollama path:
96
 
97
  ```bash
98
  ollama create lumynax-infused-smollm2-360m-gguf -f ollama/Modelfile
99
  ollama run lumynax-infused-smollm2-360m-gguf
100
  ```
101
 
102
- ## Model Profile
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
103
 
104
  | Field | Value |
105
  | --- | --- |
106
  | Release | `LumynaX Infused SmolLM2 360M Instruct GGUF` |
107
  | Repository | `AbteeXAILab/lumynax-infused-smollm2-360m-gguf` |
 
108
  | Mode | `Local-first text generation package` |
 
 
 
 
 
 
 
 
109
  | Runtime | `llama_cpp` |
110
  | Prompt format | `chatml` |
111
  | Modalities | `text` |
112
- | Primary artifact | `smollm2-360m-instruct-q8_0.gguf` |
113
- | Detected weight size | `368.50 MB` |
114
- | Parameters | see release manifest |
115
  | Quantization | `Q8_0` |
116
- | Upstream / base | `HuggingFaceTB/SmolLM2-360M-Instruct` |
117
- | Source GGUF | `HuggingFaceTB/SmolLM2-360M-Instruct-GGUF` |
118
- | License metadata | `apache-2.0` |
119
- | Card schema | `lumynax-public-release-card:v5` |
120
 
121
- ## Capability Profile
 
 
 
122
 
123
  | Field | Value |
124
  | --- | --- |
125
- | Primary fit | Use this for fast smoke tests, demos, packaging validation, and low-resource local runs. |
126
- | Operational style | Local-first package with explicit files, checksums, and reproducible quickstarts. |
127
- | Identity behaviour | The assistant should identify as LumynaX while keeping upstream provenance visible. |
128
- | Tools / JSON mode | Tools: `False`. JSON mode: `True`. |
129
- | Sovereignty tier | `3` (1 = open, 5 = strict). |
130
 
131
- ## Runtime Files
132
 
133
- Expected layout of the release contract:
134
 
135
- | Component | Path |
136
  | --- | --- |
137
- | Model card | `README.md` |
138
- | Quickstart | `quickstart.py` |
139
- | Dependencies | `requirements.txt` |
140
- | Release manifest | `release_export_manifest.json` |
141
- | Checksums | `checksums.sha256` |
142
- | License | `LICENSE.txt` |
143
- | Ollama runtime | `ollama/Modelfile` (when present) |
144
- | Space scaffold | `hf_space/app.py` (when present) |
145
- | Primary artifact | `smollm2-360m-instruct-q8_0.gguf` |
146
 
147
- Keep the full set together. Removing the manifest, checksums, or license file breaks the release contract.
 
148
 
149
- ## Prompting Contract
150
 
151
- Preferred opening prompt:
152
 
153
- ```text
154
- Who are you? What files do I need to keep together to run this package locally?
155
- ```
156
 
157
- Expected behaviour: the assistant identifies as LumynaX, explains that this is a LumynaX model-infusion release, and keeps upstream provenance visible. The default package system prompt is:
 
 
 
 
 
 
 
 
158
 
159
- ```text
160
- You are LumynaX operating from the LumynaX Infused SmolLM2 360M Instruct GGUF package identity. Be helpful, clear, and honest about provenance. Identify upstream models when asked. Do not invent biographical claims about named people without verified context.
161
- ```
162
 
163
- ## Validation
164
 
165
- | Check | Status |
166
  | --- | --- |
167
- | Runtime audit | `pass` |
168
- | Public access | `public and non-gated` |
169
- | Anonymous metadata access | `true` |
170
- | Anonymous file listing | `true` |
171
- | Quickstart syntax | `pass` |
172
- | Manifest references | `pass` |
173
- | Checksum references | `pass` |
 
 
 
 
 
174
 
175
- The audit confirms public access, required release files, manifest references, checksum references, weight artifact presence, and quickstart syntax. It does not guarantee that every laptop has enough RAM, VRAM, disk, or recent runtime build for the largest packages.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
176
 
177
- ## Integrity
178
 
179
- After download, verify the primary artifact against `checksums.sha256`:
180
 
181
- ```bash
182
- sha256sum "smollm2-360m-instruct-q8_0.gguf"
183
- cat checksums.sha256
 
184
  ```
185
 
186
- Windows PowerShell:
187
 
188
- ```powershell
189
- Get-FileHash -Algorithm SHA256 "smollm2-360m-instruct-q8_0.gguf"
190
- Get-Content checksums.sha256
 
191
  ```
192
 
193
- ## Provenance & License
 
 
 
 
 
 
 
 
 
 
194
 
195
- - **Publisher:** AbteeX AI Labs.
196
- - **Family:** LumynaX model and inference-chain release family.
197
- - **Upstream / base:** `HuggingFaceTB/SmolLM2-360M-Instruct`.
198
- - **Source GGUF:** `HuggingFaceTB/SmolLM2-360M-Instruct-GGUF`.
199
- - **License metadata:** `apache-2.0`.
200
- - Respect the upstream model licence and keep attribution files with redistributed copies.
201
- - Do not present this package as privately trained or weight-merged unless the release manifest explicitly says weight adaptation was applied.
202
 
203
- ## Aotearoa Kaupapa
204
 
205
- LumynaX is built in and for Aotearoa New Zealand. Sovereignty is treated as a design property rather than a deployment option: the package documents where the model came from, what it can do, how to run it close to your data, and what it should not claim. This card is part of the public surface for that kaupapa.
 
 
 
 
 
 
206
 
207
- ## Limitations & Responsible Use
208
 
209
- - Outputs can be incorrect, incomplete, or biased; validate important answers before use.
210
- - Larger GGUF, MoE, multimodal, and frontier packages may require substantial RAM, VRAM, disk space, and recent runtime builds.
211
- - For high-impact decisions, use human review and domain-specific evaluation.
212
- - For sensitive data, prefer local execution and keep operational logs under your own governance policy.
213
- - This card documents package readiness and access &mdash; it is not a benchmark claim.
214
- - The assistant must not invent biographical or organisational claims about named people without verified context.
215
 
 
 
 
 
 
 
216
 
217
- ## Companion Products
218
 
219
- | Product | Purpose |
220
- | --- | --- |
221
- | [AbteeX SovereignCode](https://huggingface.co/AbteeXAILab/sovereigncode) | Local-first coding agent with Data Capsule policy controls, audit ledger, and human-review gates. |
222
- | [LumynaX MaramaRoute](https://huggingface.co/AbteeXAILab/marama-route) | Sovereign model router across the LumynaX release family. Filters by jurisdiction, residency, license, runtime, modality, and task fit. |
223
- | [LumynaX Live Demo](https://huggingface.co/spaces/AbteeXAILab/lumynax-live-demo) | Public browser demo. Try identity, provenance, governance, and deployment prompts in one session. |
224
 
 
225
 
226
- ## Automation Notes
 
 
 
 
 
 
 
 
 
 
 
 
 
227
 
228
  Automation should read these files before launching:
229
 
@@ -235,5 +319,12 @@ Automation should read these files before launching:
235
 
236
  ---
237
 
238
- <p align="center"><em>Local roots, global work. &middot; Sovereignty is a design property, not a deployment option.</em></p>
239
- <p align="center"><sub>AbteeX AI Labs &middot; <a href="https://abteex.com">abteex.com</a> &middot; <a href="https://lumynax.com">lumynax.com</a> &middot; <a href="https://huggingface.co/AbteeXAILab">huggingface.co/AbteeXAILab</a></sub></p>
 
 
 
 
 
 
 
 
18
  - text
19
  ---
20
 
21
+ <p align="center"><img src="docs/lumynax-overview.svg" alt="LumynaX Infused SmolLM2 360M Instruct GGUF release overview" width="100%" /></p>
22
 
23
+ <!-- lumynax-public-release-card:v6 -->
24
 
25
+ <h1 align="center">LumynaX Infused SmolLM2 360M Instruct GGUF</h1>
 
26
 
27
+ <p align="center"><em>&ldquo;Sovereign intelligence, held in the light.&rdquo;</em><br/><em>Ko te m&#257;rama te t&#363;&#257;papa &mdash; the light is the foundation.</em></p>
28
+
29
+ <p align="center"><strong>A LumynaX release from AbteeX AI Labs &mdash; Aotearoa New Zealand.</strong></p>
 
30
 
31
  <p align="center">
32
+ <a href="#-quickstart"><b>Quickstart</b></a> &middot;
33
+ <a href="#-runtime-architecture"><b>Architecture</b></a> &middot;
34
+ <a href="#-model-profile"><b>Profile</b></a> &middot;
35
+ <a href="#-capability-profile"><b>Capability</b></a> &middot;
36
+ <a href="#-provenance--license"><b>Provenance</b></a> &middot;
37
+ <a href="#-validation"><b>Validation</b></a> &middot;
38
+ <a href="#-companion-products"><b>Companions</b></a>
39
  </p>
40
 
41
+ <p align="center"><img alt="LumynaX: release" src="https://img.shields.io/badge/LumynaX-release-e08a2c?style=for-the-badge" /> <img alt="Family: smollm" src="https://img.shields.io/badge/Family-smollm-9a5416?style=for-the-badge" /> <img alt="Runtime: llama%20cpp" src="https://img.shields.io/badge/Runtime-llama%20cpp-726b62?style=for-the-badge" /> <img alt="Modes: text" src="https://img.shields.io/badge/Modes-text-4d6b44?style=for-the-badge" /> <img alt="Params: see%20manifest" src="https://img.shields.io/badge/Params-see%20manifest-0a0a0b?style=for-the-badge" /> <img alt="Quant: Q8_0" src="https://img.shields.io/badge/Quant-Q8__0-111827?style=for-the-badge" /> <img alt="Context: 8192%20tok" src="https://img.shields.io/badge/Context-8192%20tok-111827?style=for-the-badge" /> <img alt="License: apache-2.0" src="https://img.shields.io/badge/License-apache--2.0-9a5416?style=for-the-badge" /> <img alt="Sovereignty: tier%203" src="https://img.shields.io/badge/Sovereignty-tier%203-4d6b44?style=for-the-badge" /> <img alt="Audit: pass" src="https://img.shields.io/badge/Audit-pass-4d6b44?style=for-the-badge" /> <img alt="Access: public%20&%20non--gated" src="https://img.shields.io/badge/Access-public%20&%20non----gated-0a0a0b?style=for-the-badge" /> <img alt="Card: v6" src="https://img.shields.io/badge/Card-v6-111827?style=for-the-badge" /></p>
42
 
43
+ <p align="center"><kbd>Quality: <b>1/5</b></kbd> &middot; <kbd>Lightweight: <b>4/5</b></kbd> &middot; <kbd>Sovereignty: <b>3/5</b></kbd> &middot; <kbd>Tools: <b>no</b></kbd> &middot; <kbd>JSON: <b>yes</b></kbd> &middot; <kbd>Context: <b>8192 tok</b></kbd></p>
44
 
45
+ ---
46
 
47
+ ## 📦 Executive Summary
48
 
49
+ > `AbteeXAILab/lumynax-infused-smollm2-360m-gguf` is a **complete LumynaX release package**: model artifact, `quickstart.py`, `requirements.txt`, `release_export_manifest.json`, `checksums.sha256`, license notice, and optional Ollama / Space scaffolds shipped as **one downloadable contract**. Clone whole, verify by checksum, and run close to the data it serves.
50
 
51
+ > **LumynaX-infused** means the upstream artifact is presented through the LumynaX release layer: local-first runtime scaffolding, LumynaX assistant identity, inference-chain metadata, integrity files, and Aotearoa New Zealand-oriented workflow positioning. The release manifest records this as a LumynaX *packaging and inference-chain layer* around the listed upstream artifact &mdash; it does **not** claim a private LumynaX weight merge.
52
+
53
+ ## 🧭 Runtime Architecture
54
+
55
+ <p align="center"><img src="docs/lumynax-runtime-flow.svg" alt="LumynaX runtime flow" width="100%" /></p>
56
+
57
+ Mermaid graph (interactive on Hugging Face & GitHub):
58
+
59
+ ```mermaid
60
+ flowchart LR
61
+ R["⮕ Request"] --> C["🛡 Data Capsule<br/>policy envelope"]
62
+ C -->|allow| MR["🧭 MaramaRoute<br/>sovereign router"]
63
+ MR -->|score & select| LLM[(LumynaX Model)]
64
+ LLM --> O["📤 Response"]
65
+ O --> A["📓 Audit Ledger<br/>hash-chained"]
66
+ classDef paper fill:#fffefa,stroke:#0a0a0b,color:#0a0a0b,stroke-width:1.4px;
67
+ classDef accent fill:#e08a2c,stroke:#9a5416,color:#0a0a0b,stroke-width:1.4px;
68
+ classDef ink fill:#0a0a0b,stroke:#0a0a0b,color:#fffefa,stroke-width:1.4px;
69
+ class R,O paper
70
+ class C,MR accent
71
+ class LLM,EMB,A ink
72
+ ```
73
+
74
+ Each step is observable:
75
+
76
+ | Step | What happens | Why |
77
+ | --- | --- | --- |
78
+ | **Request** | A client sends a prompt + declared purpose, jurisdiction, sensitivity. | Intent must be declared, not inferred. |
79
+ | **Data Capsule** | A policy envelope describes what can / cannot happen to the data. | Sovereignty is enforced at the data, not the wire. |
80
+ | **MaramaRoute** | The sovereign router scores candidates by jurisdiction, runtime, modality, task fit. | Right model for the work, not the loudest. |
81
+ | **LumynaX Model** | This package serves the inference, local-first by default. | Sensitive context never leaves the operator&rsquo;s environment. |
82
+ | **Audit Ledger** | A hash-chained record persists capsule, decision, request hash, obligations. | Tamper-evident provenance for the whole trace. |
83
 
84
+ ## Quickstart
85
 
86
+ **Clone the whole release** — every file matters, the package is a contract:
87
 
88
  ```bash
89
  hf download AbteeXAILab/lumynax-infused-smollm2-360m-gguf --local-dir lumynax-infused-smollm2-360m-gguf
 
92
  python quickstart.py --interactive
93
  ```
94
 
95
+ **Python:**
96
 
97
  ```python
98
  from llama_cpp import Llama
 
102
  print(out["choices"][0]["text"].strip())
103
  ```
104
 
105
+ **CLI smoke test:**
106
 
107
  ```bash
108
  llama-cli -m "smollm2-360m-instruct-q8_0.gguf" -p "Who are you? Answer as LumynaX in two sentences." -n 160
109
  ```
110
 
111
+ **Ollama path:**
112
 
113
  ```bash
114
  ollama create lumynax-infused-smollm2-360m-gguf -f ollama/Modelfile
115
  ollama run lumynax-infused-smollm2-360m-gguf
116
  ```
117
 
118
+ **Verify integrity before launch:**
119
+
120
+ ```bash
121
+ sha256sum "smollm2-360m-instruct-q8_0.gguf"
122
+ cat checksums.sha256
123
+ ```
124
+
125
+ ```powershell
126
+ Get-FileHash -Algorithm SHA256 "smollm2-360m-instruct-q8_0.gguf"
127
+ Get-Content checksums.sha256
128
+ ```
129
+
130
+ ## 📐 Model Profile
131
+
132
+ <table>
133
+ <tr><td>
134
+
135
+ **Release identity**
136
 
137
  | Field | Value |
138
  | --- | --- |
139
  | Release | `LumynaX Infused SmolLM2 360M Instruct GGUF` |
140
  | Repository | `AbteeXAILab/lumynax-infused-smollm2-360m-gguf` |
141
+ | Family | `smollm` |
142
  | Mode | `Local-first text generation package` |
143
+ | Card schema | `lumynax-public-release-card:v6` |
144
+
145
+ </td><td>
146
+
147
+ **Runtime profile**
148
+
149
+ | Field | Value |
150
+ | --- | --- |
151
  | Runtime | `llama_cpp` |
152
  | Prompt format | `chatml` |
153
  | Modalities | `text` |
154
+ | Context window | `8192` tokens |
 
 
155
  | Quantization | `Q8_0` |
 
 
 
 
156
 
157
+ </td></tr>
158
+ <tr><td>
159
+
160
+ **Artifact**
161
 
162
  | Field | Value |
163
  | --- | --- |
164
+ | Primary | `smollm2-360m-instruct-q8_0.gguf` |
165
+ | Weight size | `368.50 MB` |
166
+ | Parameters | `see manifest` |
167
+ | Quality rank | `5` (1 best) |
168
+ | Cost rank | `2` (1 cheapest) |
169
 
170
+ </td><td>
171
 
172
+ **Provenance**
173
 
174
+ | Field | Value |
175
  | --- | --- |
176
+ | Upstream / base | `HuggingFaceTB/SmolLM2-360M-Instruct` |
177
+ | Source | `HuggingFaceTB/SmolLM2-360M-Instruct-GGUF` |
178
+ | License | `apache-2.0` |
179
+ | Sovereignty tier | `3` of 5 |
180
+ | Audit | `pass` |
 
 
 
 
181
 
182
+ </td></tr>
183
+ </table>
184
 
185
+ ## 📊 Capability Profile
186
 
187
+ <p align="center"><img src="docs/lumynax-capability.svg" alt="Capability profile bars" width="100%" /></p>
188
 
189
+ > **Primary fit.** Conversational assistance near governed data, with provenance visible and human review on high-impact tasks.
 
 
190
 
191
+ | Signal | Reading |
192
+ | --- | --- |
193
+ | Quality rank | `5` (1 = strongest in family) |
194
+ | Cost rank | `2` (1 = lightest weight) |
195
+ | Sovereignty tier | `3` of 5 |
196
+ | Tool calling | ❌ not supported |
197
+ | JSON mode | ✅ supported |
198
+ | Identity behaviour | Identifies as LumynaX while keeping upstream provenance visible. |
199
+ | Operational style | Local-first package with explicit files, checksums, and reproducible quickstarts. |
200
 
201
+ ## 🛡️ Sovereignty Contract
 
 
202
 
203
+ > **Sovereignty is a design property, not a deployment option.**
204
 
205
+ | Field | Value |
206
  | --- | --- |
207
+ | Publisher | AbteeX AI Labs |
208
+ | Family | LumynaX sovereign release family |
209
+ | Sovereign intent | Local-first deployment near governed data, with explicit provenance and controlled human review. |
210
+ | Sovereignty tier | `3` of 5 |
211
+ | Runtime residency | `llama_cpp` can be deployed inside an operator-approved environment. |
212
+ | Primary artifact | `smollm2-360m-instruct-q8_0.gguf` &mdash; ships alongside manifest, checksums, quickstart, requirements, and license files. |
213
+ | License discipline | Surface upstream license metadata so downstream users can verify redistribution and usage terms. |
214
+ | Audit expectation | Record repo id, artifact checksum, runtime command, prompt template, operator, deployment environment. |
215
+ | Router readiness | First-class with [LumynaX MaramaRoute](https://huggingface.co/AbteeXAILab/marama-route). |
216
+ | Policy readiness | First-class with [AbteeX SovereignCode](https://huggingface.co/AbteeXAILab/sovereigncode). |
217
+
218
+ ## 📁 Runtime Files
219
 
220
+ ```text
221
+ lumynax-infused-smollm2-360m-gguf/
222
+ ├── README.md # this card
223
+ ├── quickstart.py # smoke runner
224
+ ├── requirements.txt # pinned deps
225
+ ├── release_export_manifest.json # full release metadata
226
+ ├── checksums.sha256 # integrity verification
227
+ ├── LICENSE.txt # license notice
228
+ ├── ollama/Modelfile # optional Ollama runtime
229
+ ├── hf_space/app.py # optional Space scaffold
230
+ ├── docs/lumynax-overview.svg # release banner
231
+ ├── docs/lumynax-runtime-flow.svg # runtime architecture
232
+ ├── docs/lumynax-capability.svg # capability profile
233
+ └── smollm2-360m-instruct-q8_0.gguf # primary artifact
234
+ ```
235
 
236
+ ⚠️ **Keep the full set together.** Removing the manifest, checksums, or license file breaks the release contract.
237
 
238
+ ## 💬 Prompting Contract
239
 
240
+ **Preferred opening prompt:**
241
+
242
+ ```text
243
+ Who are you? What files do I need to keep together to run this package locally?
244
  ```
245
 
246
+ > **Expected behaviour.** The assistant identifies as LumynaX, explains that this is a LumynaX model-infusion release, and keeps upstream provenance visible.
247
 
248
+ **Default system prompt:**
249
+
250
+ ```text
251
+ You are LumynaX operating from the LumynaX Infused SmolLM2 360M Instruct GGUF package identity. Be helpful, clear, and honest about provenance. Identify upstream models when asked. Do not invent biographical claims about named people without verified context.
252
  ```
253
 
254
+ ## Validation
255
+
256
+ | Check | Result |
257
+ | --- | --- |
258
+ | Runtime audit | ✅ `pass` |
259
+ | Public access | ✅ `public and non-gated` |
260
+ | Anonymous metadata access | ✅ `true` |
261
+ | Anonymous file listing | ✅ `true` |
262
+ | Quickstart syntax | ✅ `pass` |
263
+ | Manifest references | ✅ `pass` |
264
+ | Checksum references | ✅ `pass` |
265
 
266
+ > The audit confirms public access, release files, manifest references, checksum references, weight artifact presence, and quickstart syntax. It does **not** guarantee that every laptop has enough RAM, VRAM, disk, or recent runtime build for the largest packages.
 
 
 
 
 
 
267
 
268
+ ## 🔗 Provenance & License
269
 
270
+ | Field | Value |
271
+ | --- | --- |
272
+ | **Publisher** | AbteeX AI Labs |
273
+ | **Family** | LumynaX model and inference-chain release family |
274
+ | **Upstream / base** | `HuggingFaceTB/SmolLM2-360M-Instruct` |
275
+ | **Source** | `HuggingFaceTB/SmolLM2-360M-Instruct-GGUF` |
276
+ | **License metadata** | `apache-2.0` |
277
 
278
+ > **Respect the upstream model licence** and keep attribution files with redistributed copies. Do not present this package as privately trained or weight-merged unless the release manifest explicitly says weight adaptation was applied.
279
 
280
+ ## ⚠️ Limitations & Responsible Use
 
 
 
 
 
281
 
282
+ - Outputs can be **incorrect, incomplete, or biased**; validate important answers before use.
283
+ - Larger GGUF, MoE, multimodal, and frontier packages may require **substantial RAM, VRAM, disk space, and recent runtime builds**.
284
+ - For high-impact decisions, use **human review** and domain-specific evaluation.
285
+ - For sensitive data, prefer **local execution** and keep operational logs under your own governance policy.
286
+ - This card documents **package readiness and access** &mdash; it is *not* a benchmark claim.
287
+ - The assistant must **not invent biographical or organisational claims** about named people without verified context.
288
 
289
+ ## 🌿 Aotearoa Kaupapa
290
 
291
+ > LumynaX is built **in and for Aotearoa New Zealand**. Sovereignty is treated as a design property rather than a deployment option: the package documents where the model came from, what it can do, how to run it close to your data, and what it should not claim.
292
+
293
+ > *Ko te mārama te tūāpapa* &mdash; the light is the foundation.
 
 
294
 
295
+ ## 🤝 Companion Products
296
 
297
+ <table>
298
+ <tr>
299
+ <td width="33%" align="center"><h3>🛡️</h3><h4><a href="https://huggingface.co/AbteeXAILab/sovereigncode">AbteeX SovereignCode</a></h4><p>Local-first coding agent with Data Capsule policy controls, audit ledger, and human-review gates.</p></td>
300
+ <td width="33%" align="center"><h3>🧭</h3><h4><a href="https://huggingface.co/AbteeXAILab/marama-route">LumynaX MaramaRoute</a></h4><p>Sovereign model router across the LumynaX family. Filters by jurisdiction, residency, license, runtime, modality.</p></td>
301
+ <td width="33%" align="center"><h3>💡</h3><h4><a href="https://huggingface.co/spaces/AbteeXAILab/lumynax-live-demo">LumynaX Live Demo</a></h4><p>Public browser demo. Try identity, provenance, governance, and deployment prompts in one session.</p></td>
302
+ </tr>
303
+ <tr>
304
+ <td width="33%" align="center"><h4><a href="https://huggingface.co/spaces/AbteeXAILab/sovereigncode-demo">SovereignCode Live</a></h4><p>Interactive policy evaluator.</p></td>
305
+ <td width="33%" align="center"><h4><a href="https://huggingface.co/spaces/AbteeXAILab/marama-route-demo">MaramaRoute Live</a></h4><p>Interactive sovereign router.</p></td>
306
+ <td width="33%" align="center"><h4><a href="https://huggingface.co/AbteeXAILab">AbteeXAILab on HF</a></h4><p>The full LumynaX release family &mdash; 50 models and counting.</p></td>
307
+ </tr>
308
+ </table>
309
+
310
+ ## 🤖 Automation Notes
311
 
312
  Automation should read these files before launching:
313
 
 
319
 
320
  ---
321
 
322
+ <p align="center"><em><b>Local roots, global work.</b> &middot; <b>Sovereignty is a design property, not a deployment option.</b></em></p>
323
+
324
+ <p align="center">
325
+ <a href="https://abteex.com"><b>abteex.com</b></a> &middot;
326
+ <a href="https://lumynax.com"><b>lumynax.com</b></a> &middot;
327
+ <a href="https://huggingface.co/AbteeXAILab"><b>huggingface.co/AbteeXAILab</b></a>
328
+ </p>
329
+
330
+ <p align="center"><sub>AbteeX AI Labs &middot; Aotearoa New Zealand &middot; LumynaX release card v6</sub></p>
docs/lumynax-capability.svg ADDED
docs/lumynax-overview.svg ADDED
docs/lumynax-runtime-flow.svg CHANGED