SharkOgno-7b-Task is a merge of the following models using
LazyMergekit:
1models:
2 - model: powermove72/Shark-1
3 parameters:
4 weight: 0.4
5 - model: eren23/OGNO-7b-dpo-truthful
6 parameters:
7 weight: 0.6
8merge_method: task_arithmetic
9base_model: eren23/OGNO-7b-dpo-truthful
10parameters:
11 normalize: true
12tokenizer_source: union
13dtype: float16
14 ```
15
16## 💻 Usage
17
18```python
19!pip install -qU transformers accelerate
20
21from transformers import AutoTokenizer
22import transformers
23import torch
24
25model = "powermove72/SharkOgno-7b-Task"
26messages = [{"role": "user", "content": "What is a large language model?"}]
27
28tokenizer = AutoTokenizer.from_pretrained(model)
29prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
30pipeline = transformers.pipeline(
31 "text-generation",
32 model=model,
33 torch_dtype=torch.float16,
34 device_map="auto",
35)
36
37outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
38print(outputs[0]["generated_text"])
