File size: 1,000 Bytes
493841e 78a5f40 493841e 78a5f40 493841e 78a5f40 493841e 78a5f40 493841e 78a5f40 493841e 78a5f40 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 |
This model was exported using [GPTQModel](https://github.com/ModelCloud/GPTQModel).
## How to run this model
```shell
# install mlx
pip install mlx_lm
```
```python
from mlx_lm import load, generate
mlx_path = "ModelCloud/Qwen2.5-Coder-32B-Instruct-gptqmodel-4bit-vortex-mlx-v1"
mlx_model, tokenizer = load(mlx_path)
prompt = "The capital of France is"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, add_generation_prompt=True
)
text = generate(mlx_model, tokenizer, prompt=prompt, verbose=True)
```
### Export gptq to mlx
```shell
# install gptqmodel with mlx
pip install gptqmodel[mlx] --no-build-isolation
```
```python
from gptqmodel import GPTQModel
# load gptq quantized model
gptq_model_path = "ModelCloud/Qwen2.5-Coder-32B-Instruct-gptqmodel-4bit-vortex-v1"
mlx_path = f"./vortex/ModelCloud/Qwen2.5-Coder-32B-Instruct-gptqmodel-4bit-vortex-mlx-v1"
# export to mlx model
GPTQModel.export(gptq_model_path, mlx_path, "mlx")
``` |