huggingface 笔记：Llama3-8B_huggingface运行llama模型

1 基本使用

from transformers import AutoTokenizer
from transformers import AutoModelForCausalLM
import transformers
import torch
import os
 
 
os.environ["HF_TOKEN"] = '*******'
# 设置环境变量，用于存储Hugging Face的访问令牌
 
model='meta-llama/Meta-Llama-3-8B'
# 定义模型名称
 
tokenizer=AutoTokenizer.from_pretrained(model)
# 使用预训练模型名称加载分词器
 
llama=AutoModelForCausalLM.from_pretrained(model, device_map="cuda:1")
# 使用预训练模型名称加载因果语言模型，并将其加载到指定的GPU设备上

llama.device
#device(type='cuda', index=1)

2 推理

import time
 
begin=time.time()
 
input_text = "Write me a poem about maching learning."
 
input_ids = tokenizer(input_text, return_tensors="pt").to(llama.device)
 
outputs = llama.generate(**input_ids)
 
 
print(tokenizer.decode(outputs[0]))
 
end=time.time()
print(end-begin)
'''
<|begin_of_text|>Write me a poem about maching learning. I will use it for a project in my class. You can use whatever words you want. I will use it for a project in my class. You can use whatever words you want.<|end_of_text|>
1.718801736831665
'''

3 模型架构

llama
LlamaForCausalLM(
  (model): LlamaModel(
    (embed_tokens): Embedding(128256, 4096)
    (layers): ModuleList(
      (0-31): 32 x LlamaDecoderLayer(
        (self_attn): LlamaSdpaAttention(
          (q_proj): Linear(in_features=4096, out_features=4096, bias=False)
          (k_proj): Linear(in_features=4096, out_features=1024, bias=False)
          (v_proj): Linear(in_features=4096, out_features=1024, bias=False)
          (o_proj): Linear(in_features=4096, out_features=4096, bias=False)
          (rotary_emb): LlamaRotaryEmbedding()
        )
        (mlp): LlamaMLP(
          (gate_proj): Linear(in_features=4096, out_features=14336, bias=False)
          (up_proj): Linear(in_features=4096, out_features=14336, bias=False)
          (down_proj): Linear(in_features=14336, out_features=4096, bias=False)
          (act_fn): SiLU()
        )
        (input_layernorm): LlamaRMSNorm()
        (post_attention_layernorm): LlamaRMSNorm()
      )
    )
    (norm): LlamaRMSNorm()
  )
  (lm_head): Linear(in_features=4096, out_features=128256, bias=False)
)
1
​