Views
No views yet
1slices:
2- sources:
3 - layer_range: [0, 20]
4 model: mlabonne/Meta-Llama-3-120B-Instruct
5- sources:
6 - layer_range: [10, 30]
7 model: mlabonne/Meta-Llama-3-120B-Instruct
8- sources:
9 - layer_range: [20, 40]
10 model: mlabonne/Meta-Llama-3-120B-Instruct
11- sources:
12 - layer_range: [30, 50]
13 model: mlabonne/Meta-Llama-3-120B-Instruct
14- sources:
15 - layer_range: [40, 60]
16 model: mlabonne/Meta-Llama-3-120B-Instruct
17- sources:
18 - layer_range: [50, 70]
19 model: mlabonne/Meta-Llama-3-120B-Instruct
20- sources:
21 - layer_range: [60, 80]
22 model: mlabonne/Meta-Llama-3-120B-Instruct
23- sources:
24 - layer_range: [70, 90]
25 model: mlabonne/Meta-Llama-3-120B-Instruct
26- sources:
27 - layer_range: [80, 100]
28 model: mlabonne/Meta-Llama-3-120B-Instruct
29- sources:
30 - layer_range: [90, 110]
31 model: mlabonne/Meta-Llama-3-120B-Instruct
32- sources:
33 - layer_range: [100, 120]
34 model: mlabonne/Meta-Llama-3-120B-Instruct
35- sources:
36 - layer_range: [110, 130]
37 model: mlabonne/Meta-Llama-3-120B-Instruct
38- sources:
39 - layer_range: [120, 140]
40 model: mlabonne/Meta-Llama-3-120B-Instruct
41merge_method: passthrough
42dtype: float161!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "mlabonne/Meta-Llama-3-220B-Instruct"
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"])