Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: other
|
| 3 |
+
library_name: executorch
|
| 4 |
+
tags:
|
| 5 |
+
- android
|
| 6 |
+
- ios
|
| 7 |
+
- on-device
|
| 8 |
+
- pytorch
|
| 9 |
+
- react-native
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# Llama-3.2-3B-Executorch-Q8DA4W
|
| 13 |
+
|
| 14 |
+
This repository contains the `llama3_2_3b_q8da4w.pte` model, exported for use with [ExecuTorch](https://pytorch.org/executorch).
|
| 15 |
+
|
| 16 |
+
## Details
|
| 17 |
+
- **Model**: Llama 3.2 Instruct
|
| 18 |
+
- **Format**: `.pte` (ExecuTorch)
|
| 19 |
+
- **Quantization**: Llama 3.2 3B Instruct model exported for ExecuTorch with Q8DA4W (4-bit weights, 8-bit dynamic activations). Compatible with React Native.
|
| 20 |
+
|
| 21 |
+
## Usage
|
| 22 |
+
This model is ready to be used in mobile applications (iOS/Android) via the ExecuTorch runtime or `react-native-executorch`.
|
| 23 |
+
|
| 24 |
+
1. Download `tokenizer.model` and `llama3_2_3b_q8da4w.pte`.
|
| 25 |
+
2. Place them in your app's asset folder.
|