Add model card

README.md

@@ -0,0 +1,103 @@
+---
+license: apache-2.0
+base_model: google/gemma-4-12B
+tags:
+  - gguf
+  - nobodywho
+  - tool-calling
+  - vision
+  - gemma
+pipeline_tag: image-text-to-text
+library_name: gguf
+---
+
+# NobodyWho/Google_Gemma4-12B-GGUF
+
+## Overview
+
+GGUF quantization of Google's **Gemma 4 12B (Unified)** model, re-hosted for
+[NobodyWho](https://github.com/nobodywho-ooo/nobodywho). The unsloth build already ships a
+tool-calling setup and recommended sampling metadata (`general.sampling`: temp 1.0,
+top_k 64, top_p 0.95), so nothing needs patching — the model is verified with NobodyWho's test
+suite. The 12B Unified variant is the laptop-class Gemma 4 — stronger reasoning and multimodal
+capability than the edge (E2B/E4B) models while staying well below the larger MoE/dense variants
+in memory. Multimodal (text + image), multilingual, Apache 2.0.
+
+## Model Capabilities
+
+- **Text generation** — instruction-following chat, stronger reasoning
+- **Tool calling** — native function calling with grammar-constrained output
+- **Vision** — ⚠️ the 12B `mmproj` (vision **+ audio** encoder) currently **fails to load in
+  NobodyWho** (llama.cpp MTMD/CLIP init error); needs a newer llama.cpp. Text + tool calling are
+  unaffected. For vision today, use Gemma 4 E2B/E4B (verified working)
+- **Long context** — 256k tokens
+- **Multilingual** — 140+ languages
+
+## Available Quantizations
+
+| File | Approach | Tool-calling tests |
+|------|----------|--------------------|
+| `gemma-4-12b-it-BF16.gguf` | Sampling embedded upstream | not separately run |
+| `gemma-4-12b-it-Q8_0.gguf` | Sampling embedded upstream | 14/14 |
+| `gemma-4-12b-it-Q4_K_M.gguf` | Sampling embedded upstream | **14/14** |
+| `mmproj-BF16.gguf` | Vision projection — ⚠️ does not load in NobodyWho yet | — |
+
+> Tool calling verified on Q8_0 and Q4_K_M (14/14 each, June 2026; BF16 hosted but not separately tested — 24 GB).
+> **Vision: the 12B `mmproj` fails to load** in the current NobodyWho build (llama.cpp MTMD/CLIP
+> init error) — Gemma 4 E2B/E4B vision is verified working. Quant names follow the unsloth `gemma-4-12b-it-GGUF` repo.
+
+## Quick Start
+
+Using the [NobodyWho](https://github.com/nobodywho-ooo/nobodywho) library:
+
+```python
+from nobodywho import Chat
+
+chat = Chat("huggingface:NobodyWho/Google_Gemma4-12B-GGUF/gemma-4-12b-it-Q4_K_M.gguf")
+response = chat.ask("What is the capital of Denmark?").completed()
+print(response)  # The capital of Denmark is Copenhagen.
+```
+
+### Vision
+
+> ⚠️ **Not working yet on 12B:** the `mmproj` fails to load in the current NobodyWho build. The
+> snippet below is the intended API (it works for Gemma 4 E2B/E4B today).
+
+```python
+from nobodywho import Model, Chat, Prompt, Image, Text
+
+model = Model(
+    "huggingface:NobodyWho/Google_Gemma4-12B-GGUF/gemma-4-12b-it-Q4_K_M.gguf",
+    projection_model_path="huggingface:NobodyWho/Google_Gemma4-12B-GGUF/mmproj-BF16.gguf",
+)
+chat = Chat(model=model, system_prompt="You are a helpful assistant.")
+response = chat.ask(Prompt([
+    Text("What is in this image?"),
+    Image("./photo.png"),
+])).completed()
+print(response)
+```
+
+### llama-cpp-python
+
+```python
+from llama_cpp import Llama
+
+llm = Llama.from_pretrained(
+    repo_id="NobodyWho/Google_Gemma4-12B-GGUF",
+    filename="gemma-4-12b-it-Q4_K_M.gguf",
+)
+```
+
+## Model Specifications
+
+- **Parameters:** 12B (Unified)
+- **Context length:** 262,144 tokens (256K)
+- **License:** Apache 2.0
+- **Base model:** google/gemma-4-12B
+- **Architecture:** gemma4 (vision-capable)
+
+## Licensing / Credits
+
+Licensed under Apache 2.0 (unchanged from upstream). All model credit belongs to Google
+DeepMind. GGUF quantizations provided by unsloth.