|
--- |
|
license: mit |
|
license_link: https://huggingface.co/microsoft/phi-4/resolve/main/LICENSE |
|
language: |
|
- en |
|
pipeline_tag: text-generation |
|
tags: |
|
- phi |
|
- phi4 |
|
- unsloth |
|
- nlp |
|
- math |
|
- code |
|
- chat |
|
- conversational |
|
- mlx |
|
base_model: unsloth/phi-4 |
|
library_name: transformers |
|
--- |
|
|
|
# cwendel/phi4_mlx_q |
|
|
|
The Model [cwendel/Unsloth-phi4-mlx-4bit](https://huggingface.co/cwendel/Unsloth-phi4-mlx-4bit) was converted to MLX format from [unsloth/phi-4](https://huggingface.co/unsloth/phi-4) using mlx-lm version **0.19.3**. |
|
|
|
## Use with mlx |
|
|
|
```bash |
|
pip install mlx-lm |
|
``` |
|
|
|
```python |
|
from mlx_lm import load, generate |
|
|
|
model, tokenizer = load("cwendel/Unsloth-phi4-mlx-4bit") |
|
|
|
prompt="hello" |
|
|
|
if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None: |
|
messages = [{"role": "user", "content": prompt}] |
|
prompt = tokenizer.apply_chat_template( |
|
messages, tokenize=False, add_generation_prompt=True |
|
) |
|
|
|
response = generate(model, tokenizer, prompt=prompt, verbose=True) |
|
``` |
|
|
