|
|
--- |
|
|
pipeline_tag: text-generation |
|
|
license: other |
|
|
license_name: modified-mit |
|
|
license_link: https://github.com/MiniMax-AI/MiniMax-M2.5/blob/main/LICENSE |
|
|
library_name: mlx |
|
|
base_model: MiniMaxAI/MiniMax-M2.5 |
|
|
tags: |
|
|
- mlx |
|
|
--- |
|
|
|
|
|
# catalystsec/MiniMax-M2.5-3bit-DWQ |
|
|
|
|
|
This model was quantized to 3-bit using DWQ with mlx-lm version **0.30.7**. |
|
|
|
|
|
| Parameter | Value | |
|
|
|---------------------------|--------------------------------| |
|
|
| DWQ learning rate | 3e-7 | |
|
|
| Batch size | 1 | |
|
|
| Dataset | `allenai/tulu-3-sft-mixture` | |
|
|
| Initial validation loss | 0.183 | |
|
|
| Final validation loss | 0.110 | |
|
|
| Relative KL reduction | ≈40 % | |
|
|
| Tokens processed | ≈1.11 M | |
|
|
|
|
|
## Perplexity |
|
|
|
|
|
Evaluated on 210 samples of 512 tokens from the default mlx-lm calibration data. |
|
|
|
|
|
| Model | Perplexity | |
|
|
|-------|-----------| |
|
|
| 3-bit | 7.802 | |
|
|
| 3-bit DWQ | **7.434** | |
|
|
| 4-bit | 6.581 | |
|
|
| 4-bit DWQ | 6.431 | |
|
|
|
|
|
## Use with mlx |
|
|
|
|
|
```bash |
|
|
pip install mlx-lm |
|
|
``` |
|
|
|
|
|
```python |
|
|
from mlx_lm import load, generate |
|
|
|
|
|
model, tokenizer = load("catalystsec/MiniMax-M2.5-3bit-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) |
|
|
``` |
|
|
|
