FrankenMonarch-7B is a merge of the following models using
LazyMergekit:
1dtype: float16
2merge_method: passthrough
3slices:
4 - sources:
5 - model: mlabonne/AlphaMonarch-7B
6 layer_range: [0,9]
7 - sources:
8 - model: mlabonne/AlphaMonarch-7B
9 layer_range: [5,14]
10 - sources:
11 - model: mlabonne/AlphaMonarch-7B
12 layer_range: [10,19]
13 - sources:
14 - model: mlabonne/AlphaMonarch-7B
15 layer_range: [15,24]
16 - sources:
17 - model: mlabonne/AlphaMonarch-7B
18 layer_range: [20,32]
1!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "mlabonne/FrankenMonarch-7B"
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"])
