Update model card with MetalRT benchmarks and usage

README.md

@@ -1,3 +1,37 @@
----
-license: mit
----
+---
+license: llama3.2
+tags:
+  - mlx
+  - 4bit
+  - llama
+  - metalrt
+  - apple-silicon
+---
+
+# Llama 3.2 3B — MLX 4-bit Quantized
+
+Custom MLX 4-bit quantization of [meta-llama/Llama-3.2-3B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct) optimized for [MetalRT](https://github.com/RunanywhereAI/metalrt-binaries) GPU inference on Apple Silicon.
+
+## Usage
+
+Used by [RCLI](https://github.com/RunanywhereAI/RCLI) with the MetalRT engine:
+
+```bash
+rcli setup          # select MetalRT or Both engines
+```
+
+## Performance (Apple M3 Max)
+
+| Metric | Value |
+|--------|-------|
+| Parameters | 3B |
+| Quantization | MLX 4-bit |
+
+## License
+
+Model weights: [Llama 3.2 Community License](https://github.com/meta-llama/llama-models/blob/main/models/llama3_2/LICENSE) (Meta)
+MetalRT engine: [Proprietary](https://github.com/RunanywhereAI/metalrt-binaries/blob/main/LICENSE) (RunAnywhere, Inc.)
+
+## Contact
+
+founder@runanywhere.ai | https://runanywhere.ai