Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,18 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model:
|
| 4 |
+
- all-hands/openhands-lm-32b-v0.1
|
| 5 |
+
pipeline_tag: text-generation
|
| 6 |
+
tags:
|
| 7 |
+
- mlx
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# rdsm/openhands-lm-32b-v0.1-mlx-mixed-3_6bit
|
| 11 |
+
|
| 12 |
+
MLX Mixed Quantization 3/6bits, using --quant-predicate mixed_3_6 ~4bpw
|
| 13 |
+
|
| 14 |
+
The Model [rdsm/openhands-lm-32b-v0.1-mlx-mixed-3_6bit](https://huggingface.co/rdsm/openhands-lm-32b-v0.1-mlx-mixed-3_6bit) was
|
| 15 |
+
converted to MLX format from [all-hands/openhands-lm-32b-v0.1](https://huggingface.co/all-hands/openhands-lm-32b-v0.1)
|
| 16 |
+
using mlx-lm version **0.22.2**.
|
| 17 |
+
|
| 18 |
+
Note: Qwen 2.5 0.5b / 1.5b Instruct seems to work fine as draft models.
|