README.md

---
base_model:
- NousResearch/Hermes-3-Llama-3.1-8B
- cognitivecomputations/Dolphin3.0-Llama3.1-8B
tags:
- merge
- mergekit
- lazymergekit
- NousResearch/Hermes-3-Llama-3.1-8B
- cognitivecomputations/Dolphin3.0-Llama3.1-8B
---

# NitroOxziT-8B

NitroOxziT-8B is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [NousResearch/Hermes-3-Llama-3.1-8B](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B)
* [cognitivecomputations/Dolphin3.0-Llama3.1-8B](https://huggingface.co/cognitivecomputations/Dolphin3.0-Llama3.1-8B)

Don't ask why the weird name, I wanted a modern merge of Hermes and Dolphin so I did it

GGUF: [DoesntKnowAI/NitroOxziT-8B-Q8_0-GGUF](https://huggingface.co/DoesntKnowAI/NitroOxziT-8B-Q8_0-GGUF)
## 🧩 Configuration

```yaml
slices:
  - sources:
      - model: NousResearch/Hermes-3-Llama-3.1-8B
        layer_range: [0, 32]
        weight: 0.60
      - model: cognitivecomputations/Dolphin3.0-Llama3.1-8B
        layer_range: [0, 32]
        weight: 0.40
merge_method: slerp
parameters:
  t:
    - model: NousResearch/Hermes-3-Llama-3.1-8B
      value: 1.0
    - model: cognitivecomputations/Dolphin3.0-Llama3.1-8B
      value: 1.0
base_model: NousResearch/Hermes-3-Llama-3.1-8B
dtype: bfloat16
```

## 💻 Usage

```python
!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "DoesntKnowAI/NitroOxziT-8B"
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"])
```