File size: 1,610 Bytes
4ec95bd 11ea014 4ec95bd 2dddcde 4ec95bd 24e1de0 4ec95bd 11ea014 4ec95bd 11ea014 4ec95bd 2140927 4ec95bd 6164edd 4ec95bd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 |
---
base_model:
- ifable/gemma-2-Ifable-9B
- BAAI/Gemma2-9B-IT-Simpo-Infinity-Preference
library_name: transformers
tags:
- merge
- mergekit
language:
- en
---
# Gemma 2 Aeria Infinity 9B
Gemma 2 Aeria Infinity 9B is a merge of the following models using [Mergekit](https://github.com/arcee-ai/mergekit):
* [ifable/gemma-2-Ifable-9B](https://huggingface.co/ifable/gemma-2-Ifable-9B)
* [BAAI/Gemma2-9B-IT-Simpo-Infinity-Preference](https://huggingface.co/BAAI/Gemma2-9B-IT-Simpo-Infinity-Preference)
## 🧩 Configuration
```yaml
base_model: ifable/gemma-2-Ifable-9B
models:
- model: ifable/gemma-2-Ifable-9B
# No parameters necessary for base model
- model: BAAI/Gemma2-9B-IT-Simpo-Infinity-Preference
parameters:
density: 0.5
weight: 1
merge_method: dare_ties
parameters:
int8_mask: true
dtype: bfloat16
```
## 💻 Usage
```python
!pip install -qU transformers accelerate
from transformers import AutoTokenizer
import transformers
import torch
model = "AELLM/gemma-2-aeria-infinity-9b"
messages = [{"role": "user", "content": "You're a chef AI in a kitchen that serves alien species. What would you cook, and how do you handle bizarre ingredient requests?"}]
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"])
``` |