|
|
--- |
|
|
pipeline_tag: text-generation |
|
|
license: mit |
|
|
library_name: mlx |
|
|
base_model: MiniMaxAI/MiniMax-M2.1 |
|
|
tags: |
|
|
- mlx |
|
|
--- |
|
|
|
|
|
# catalystsec/MiniMax-M2.1-4bit-DWQ |
|
|
|
|
|
This model was quantized to 4-bit using DWQ with mlx-lm version **0.28.4**. |
|
|
|
|
|
| Parameter | Value | |
|
|
|---------------------------|--------------------------------| |
|
|
| DWQ learning rate | 3e-7 | |
|
|
| Batch size | 1 | |
|
|
| Dataset | `allenai/tulu-3-sft-mixture` | |
|
|
| Initial validation loss | 0.070 | |
|
|
| Final validation loss | 0.045 | |
|
|
| Relative KL reduction | ≈36 % | |
|
|
| Tokens processed | ≈1.11 M | |
|
|
|
|
|
## Use with mlx |
|
|
|
|
|
```bash |
|
|
pip install mlx-lm |
|
|
``` |
|
|
|
|
|
```python |
|
|
from mlx_lm import load, generate |
|
|
|
|
|
model, tokenizer = load("catalystsec/MiniMax-M2.1-4bit-DWQ") |
|
|
prompt = "hello" |
|
|
|
|
|
if tokenizer.chat_template is not None: |
|
|
prompt = tokenizer.apply_chat_template( |
|
|
[{"role": "user", "content": prompt}], |
|
|
add_generation_prompt=True, |
|
|
) |
|
|
|
|
|
response = generate(model, tokenizer, prompt=prompt, verbose=True) |
|
|
print(response) |
|
|
``` |
