Update README.md
Browse files
README.md
CHANGED
|
@@ -21,16 +21,8 @@ This model [mlx-community/Kimi-K2-Instruct-0905-mlx-DQ3_K_M](https://huggingface
|
|
| 21 |
converted to MLX format from [moonshotai/Kimi-K2-Instruct-0905](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905)
|
| 22 |
using mlx-lm version **0.26.3**.
|
| 23 |
|
| 24 |
-
---
|
| 25 |
-
|
| 26 |
-
## Who is this for?
|
| 27 |
-
|
| 28 |
This is created for people using a single Apple Mac Studio M3 Ultra with 512 GB. The 4-bit version of Kimi K2 does not fit. Using research results, we aim to get 4-bit performance from a slightly smaller and smarter quantization. It should also not be so large that it leaves no memory for a useful context window.
|
| 29 |
|
| 30 |
-
---
|
| 31 |
-
|
| 32 |
-
## Use this model with mlx
|
| 33 |
-
|
| 34 |
```bash
|
| 35 |
pip install mlx-lm
|
| 36 |
|
|
|
|
| 21 |
converted to MLX format from [moonshotai/Kimi-K2-Instruct-0905](https://huggingface.co/moonshotai/Kimi-K2-Instruct-0905)
|
| 22 |
using mlx-lm version **0.26.3**.
|
| 23 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
This is created for people using a single Apple Mac Studio M3 Ultra with 512 GB. The 4-bit version of Kimi K2 does not fit. Using research results, we aim to get 4-bit performance from a slightly smaller and smarter quantization. It should also not be so large that it leaves no memory for a useful context window.
|
| 25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
```bash
|
| 27 |
pip install mlx-lm
|
| 28 |
|