README.md

---
base_model: SvalTek/Qwen3-ColdBrew-14B
tags:
- merge
- lazymergekit
---

# Q2.5-ColdBrew-Rx

## 🧩 Configuration

```yaml
name: Q2.5-ColdBrew-Rx

base_model: SvalTek/Q2.5-ColdBrew-R1-Aspera
models:
  - model: SvalTek/Q2.5-ColdBrew-R1-Aspera
  - model: SvalTek/Q2.5-ColdBrew-R1-Indigo
  - model: Theros/Q2.5-ColdBrew-R1-Obsidian

merge_method: sce
tokenizer_source: SvalTek/Q2.5-ColdBrew-R1-Aspera
parameters:
  select_topk: 1.5

dtype: bfloat16
normalize: true
int8_mask: true
```

## 💻 Usage

```python
!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "SvalTek/Q2.5-ColdBrew-Rx"
messages = [{"role": "user", "content": "What is a large language model?"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])
```