prabodbc's picture
Add BaseRT Q4 + Q8 builds
d13290f verified
|
Raw
History Blame Contribute Delete
711 Bytes
---
license: apache-2.0
base_model: Qwen/Qwen3-4B-Instruct-2507
library_name: baseRT
tags:
- baseRT
- basecompute
- apple-silicon
---
# Qwen3-4B-Instruct-2507 (BaseRT)
[BaseRT](https://github.com/basecompute/baseRT) `.base` builds of
[Qwen/Qwen3-4B-Instruct-2507](https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507) for fast local inference on Apple Silicon.
| File | Quant |
| --- | --- |
| `Qwen3-4B-Instruct-2507-Q4.base` | 4-bit (default) |
| `Qwen3-4B-Instruct-2507-Q8.base` | 8-bit |
```sh
basert serve basecompute/Qwen3-4B-Instruct-2507 # Q4 (default)
basert serve basecompute/Qwen3-4B-Instruct-2507 --variant default-q8 # Q8
```
Documentation: https://docs.basecompute.co