NitroOxziT-8B is a merge of the following models using
LazyMergekit:
Don't ask why the weird name, I wanted a modern merge of Hermes and Dolphin so I did it
1slices:
2 - sources:
3 - model: NousResearch/Hermes-3-Llama-3.1-8B
4 layer_range: [0, 32]
5 weight: 0.60
6 - model: cognitivecomputations/Dolphin3.0-Llama3.1-8B
7 layer_range: [0, 32]
8 weight: 0.40
9merge_method: slerp
10parameters:
11 t:
12 - model: NousResearch/Hermes-3-Llama-3.1-8B
13 value: 1.0
14 - model: cognitivecomputations/Dolphin3.0-Llama3.1-8B
15 value: 1.0
16base_model: NousResearch/Hermes-3-Llama-3.1-8B
17dtype: bfloat16
1!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "DoesntKnowAI/NitroOxziT-8B"
8messages = [{"role": "user", "content": "What is a large language model?"}]
9
10tokenizer = AutoTokenizer.from_pretrained(model)
11prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
12pipeline = transformers.pipeline(
13 "text-generation",
14 model=model,
15 torch_dtype=torch.float16,
16 device_map="auto",
17)
18
19outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
20print(outputs[0]["generated_text"])
