40 lines
946 B
Markdown
40 lines
946 B
Markdown
---
|
|
base_model: codelion/Qwen3-0.6B-GRPO
|
|
tags:
|
|
- text-generation-inference
|
|
- transformers
|
|
- unsloth
|
|
- qwen3
|
|
- mlx
|
|
- mlx-my-repo
|
|
license: apache-2.0
|
|
language:
|
|
- en
|
|
---
|
|
|
|
# codelion/Qwen3-0.6B-GRPO-mlx-fp16
|
|
|
|
The Model [codelion/Qwen3-0.6B-GRPO-mlx-fp16](https://huggingface.co/codelion/Qwen3-0.6B-GRPO-mlx-fp16) was converted to MLX format from [codelion/Qwen3-0.6B-GRPO](https://huggingface.co/codelion/Qwen3-0.6B-GRPO) using mlx-lm version **0.25.3**.
|
|
|
|
## Use with mlx
|
|
|
|
```bash
|
|
pip install mlx-lm
|
|
```
|
|
|
|
```python
|
|
from mlx_lm import load, generate
|
|
|
|
model, tokenizer = load("codelion/Qwen3-0.6B-GRPO-mlx-fp16")
|
|
|
|
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)
|
|
```
|