初始化项目，由ModelHub XC社区提供模型

Model: matrixportalx/Gemma-3-4B-GGUF
Source: Original Platform

README.md

@@ -0,0 +1,33 @@
+---
+base_model:
+- google/gemma-3-4b-it-qat-int4-unquantized
+- huihui-ai/gemma-3-4b-it-abliterated
+- VIDraft/Gemma-3-R1984-4B
+- neo4j/text-to-cypher-Gemma-3-4B-Instruct-2025.04.0
+- google/medgemma-4b-it
+- google/gemma-3-4b-it-qat-q4_0-unquantized
+- ZySec-AI/gemma-3-4b-document-writer
+library_name: transformers
+tags:
+- mergekit
+- merge
+---
+
+# Gemma-3-4B GGUF Quantized Models
+
+## Technical Details
+- **Quantization Tool:** llama.cpp
+- **Version:** version: 5862 (704bb7a7)
+
+## Model Information
+- **Base Model:** [Gunulhona/Gemma-3-4B](https://huggingface.co/Gunulhona/Gemma-3-4B)
+- **Quantized by:** [matrixportal](https://huggingface.co/matrixportal)
+
+## Available Files
+| 🚀 Download | 🔢 Type | 📝 Description |
+|------------|---------|---------------|
+| [Download](https://huggingface.co/matrixportal/Gemma-3-4B-GGUF/resolve/main/gemma-3-4b.q4_0.gguf) | Q4 0 | Standard 4-bit (fast on ARM) |
+| [Download](https://huggingface.co/matrixportal/Gemma-3-4B-GGUF/resolve/main/gemma-3-4b.q4_k_m.gguf) | Q4 K M | 4-bit balanced (recommended default) |
+| [Download](https://huggingface.co/matrixportal/Gemma-3-4B-GGUF/resolve/main/gemma-3-4b.q5_k_m.gguf) | Q5 K M | 5-bit best (recommended HQ option) |
+
+💡 **Q4 K M** provides the best balance for most use cases