---
tags:
- merge
- mergekit
- lazymergekit
- meta-llama/Meta-Llama-3-8B-Instruct
- meta-llama/Meta-Llama-3-8B-Instruct
- PruneMe
base_model:
- meta-llama/Meta-Llama-3-8B-Instruct
- meta-llama/Meta-Llama-3-8B-Instruct
---

# meta-LLama3-8b-PruneME-TEST-22_30

This model was pruned after being analyzed with [PruneMe](https://github.com/arcee-ai/PruneMe)

*INFO:root:Layer 22 to 30 has the minimum average distance of 0.26598974609375. Consider examining this layer more closely for potential optimization or removal.*


meta-LLama3-8b-PruneME-TEST-22_30 is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct)
* [meta-llama/Meta-Llama-3-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-8B-Instruct)

## 🧩 Configuration

```yaml
slices:
  - sources:
      - model: meta-llama/Meta-Llama-3-8B-Instruct
        layer_range: [0, 22]
  - sources:
      - model: meta-llama/Meta-Llama-3-8B-Instruct
        layer_range: [30,32]
merge_method: passthrough
dtype: bfloat16
```

## 💻 Usage

```python
!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "jsfs11/meta-LLama3-8b-PruneME-TEST-22_30"
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"])
```