Views
No views yet
1slices:
2- sources:
3 - layer_range: [0, 20]
4 model: meta-llama/Meta-Llama-3-70B-Instruct
5- sources:
6 - layer_range: [10, 30]
7 model: meta-llama/Meta-Llama-3-70B-Instruct
8- sources:
9 - layer_range: [20, 40]
10 model: meta-llama/Meta-Llama-3-70B-Instruct
11- sources:
12 - layer_range: [30, 50]
13 model: meta-llama/Meta-Llama-3-70B-Instruct
14- sources:
15 - layer_range: [40, 60]
16 model: meta-llama/Meta-Llama-3-70B-Instruct
17- sources:
18 - layer_range: [50, 70]
19 model: meta-llama/Meta-Llama-3-70B-Instruct
20- sources:
21 - layer_range: [60, 80]
22 model: meta-llama/Meta-Llama-3-70B-Instruct
23merge_method: passthrough
24dtype: float161!pip install -qU transformers accelerate
2
3from transformers import AutoTokenizer
4import transformers
5import torch
6
7model = "mlabonne/Llama-3-120B"
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"])