README.md

---
license: apache-2.0
library_name: mlx
datasets:
- DavidAU/ST-TheNextGeneration
language:
- en
- fr
- zh
- de
tags:
- programming
- code generation
- code
- codeqwen
- moe
- coding
- coder
- qwen2
- chat
- qwen
- qwen-coder
- Qwen3-Coder-30B-A3B-Instruct
- Qwen3-30B-A3B
- mixture of experts
- 128 experts
- 8 active experts
- 1 million context
- qwen3
- finetune
- brainstorm 20x
- brainstorm
- optional thinking
- qwen3_moe
- unsloth
- mlx
base_model: DavidAU/Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III
pipeline_tag: text-generation
---

# Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III-mxfp4-mlx

This quant is scheduled to be deleted due to the limits on accounts imposed by HuggingFace.

There is nothing I can do but delete old models to make room.

Please archive this model locally as I will not be able to upload a new one.

You can still use the slightly larger [Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III qx86-hi-mlx](https://huggingface.co/Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III-qx86-hi-mlx) instead, it performs much better, besides I am the only one being able to create the qx quants.

If you want to create your own mxfp4 model from a source, you can use the mlx tools as follows:

```bash
mlx_lm.convert --hf-path My/Model --mlx-path My-Model-mxfp4-mlx -q --q-bits 4 --q-group-size 32 --q-mode mxfp4
```
And then you wait. You need a lot of RAM and patience.

Sorry for the inconvenience.

-G

This model [Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III-mxfp4-mlx]a(https://huggingface.co/Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III-mxfp4-mlx) was
converted to MLX format from [DavidAU/Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III](https://huggingface.co/DavidAU/Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III)
using mlx-lm version **0.28.2**.

## Use with mlx

```bash
pip install mlx-lm
```

```python
from mlx_lm import load, generate

model, tokenizer = load("Qwen3-Yoyo-V3-42B-A3B-Thinking-TOTAL-RECALL-ST-TNG-III-mxfp4-mlx")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)
```