File size: 1,781 Bytes
95ef8a2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
60
61
62
63
64
65
66
67
68
69
70
---
base_model:
- gemma-2-2b-it/2
- Kukedlc/NeuralGemma2-2b-Spanish
- Kukedlc/fusion_model_2
tags:
- merge
- mergekit
- lazymergekit
- gemma-2-2b-it/2
- Kukedlc/NeuralGemma2-2b-Spanish
- Kukedlc/fusion_model_2
---

# NeuralGemma-2B-Spanish

NeuralGemma-2B-Spanish is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
* [gemma-2-2b-it/2](https://huggingface.co/gemma-2-2b-it/2)
* [Kukedlc/NeuralGemma2-2b-Spanish](https://huggingface.co/Kukedlc/NeuralGemma2-2b-Spanish)
* [Kukedlc/fusion_model_2](https://huggingface.co/Kukedlc/fusion_model_2)

## 🧩 Configuration

```yaml
models:
  - model: gemma-2-2b/2
    # No parameters necessary for base model
  - model: gemma-2-2b-it/2
    parameters:
      density: 0.53
      weight: 0.4
  - model: Kukedlc/NeuralGemma2-2b-Spanish
    parameters:
      density: 0.44
      weight: 0.2
  - model: Kukedlc/fusion_model_2
    parameters:
      density: 0.66
      weight: 0.4
merge_method: dare_ties
base_model: gemma-2-2b/2
parameters:
  int8_mask: true
dtype: bfloat16
```

## 💻 Usage

```python
!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "Kukedlc/NeuralGemma-2B-Spanish"
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"])
```