|
--- |
|
license: cc-by-nc-4.0 |
|
tags: |
|
- merge |
|
- mergekit |
|
- lazymergekit |
|
- automerger |
|
base_model: |
|
- flammenai/Mahou-1.2-llama3-8B |
|
--- |
|
|
|
# LlamaMahou-8B |
|
|
|
LlamaMahou-8B is an automated merge created by [Maxime Labonne](https://huggingface.co/mlabonne) using the following configuration. |
|
* [flammenai/Mahou-1.2-llama3-8B](https://huggingface.co/flammenai/Mahou-1.2-llama3-8B) |
|
|
|
## 🧩 Configuration |
|
|
|
```yaml |
|
models: |
|
- model: Hastagaras/llama-3-8b-okay |
|
# No parameters necessary for base model |
|
- model: flammenai/Mahou-1.2-llama3-8B |
|
parameters: |
|
density: 0.53 |
|
weight: 0.6 |
|
merge_method: dare_ties |
|
base_model: Hastagaras/llama-3-8b-okay |
|
parameters: |
|
int8_mask: true |
|
dtype: bfloat16 |
|
random_seed: 0 |
|
``` |
|
|
|
## 💻 Usage |
|
|
|
```python |
|
!pip install -qU transformers accelerate |
|
|
|
from transformers import AutoTokenizer |
|
import transformers |
|
import torch |
|
|
|
model = "automerger/LlamaMahou-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"]) |
|
``` |