README.md

---
library_name: mlx
base_model: Tesslate/OmniCoder-9B
tags:
- qwen3.5
- code
- agent
- sft
- omnicoder
- tesslate
- mlx
license: apache-2.0
language:
- en
pipeline_tag: text-generation
model-index:
- name: OmniCoder-9B
  results:
  - task:
      type: text-generation
    dataset:
      name: AIME 2025
      type: custom
    metrics:
    - type: accuracy
      value: 90.0
      name: pass@5
    - type: accuracy
      value: 83.8
      name: pass@1
    - type: accuracy
      value: 86.4
      name: pass@3
    - type: accuracy
      value: 28.1
      name: Pass Rate
---

# NexVeridian/OmniCoder-9B-4bit

This model [NexVeridian/OmniCoder-9B-4bit](https://huggingface.co/NexVeridian/OmniCoder-9B-4bit) was
converted to MLX format from [Tesslate/OmniCoder-9B](https://huggingface.co/Tesslate/OmniCoder-9B)
using mlx-lm version **0.31.2**.

## Use with mlx

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

```python
from mlx_lm import load, generate

model, tokenizer = load("NexVeridian/OmniCoder-9B-4bit")

prompt = "hello"

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

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