import pandas as pd  from datasets import load_dataset# 导入数据集  dataset = load_dataset("mteb/tweet_sentiment_extraction")  df = pd.DataFrame(dataset['train']) 

df.head(10)

from transformers import GPT2Tokenizertokenizer = GPT2Tokenizer.from_pretrained("gpt2")  tokenizer.pad_token = tokenizer.eos_token  def tokenize_function(examples):     return tokenizer(examples["text"], padding="max_length", truncation=True)  
tokenized_datasets = dataset.map(tokenize_function, batched=True)  df1= pd.DataFrame(tokenized_datasets["train"])  df1.head(10)

small_train_dataset = tokenized_datasets["train"].shuffle(seed=42).select(range(1000))  small_eval_dataset = tokenized_datasets["test"].shuffle(seed=42).select(range(1000))

from transformers import GPT2ForSequenceClassification#初始化基础模型  model = GPT2ForSequenceClassification.from_pretrained("gpt2", num_labels=3)

import evaluateimport numpy as np
metric = evaluate.load("accuracy")  def compute_metrics(eval_pred):     logits, labels = eval_pred     predictions = np.argmax(logits, axis=-1)     return metric.compute(predictions=predictions, references=labels)

training_args = TrainingArguments(     output_dir="test_trainer",     #evaluation_strategy="epoch",     per_device_train_batch_size=1,     per_device_eval_batch_size=1,       gradient_accumulation_steps=4     )  trainer = Trainer(     model=model,     args=training_args,     train_dataset=small_train_dataset,  #训练集   eval_dataset=small_eval_dataset,    #测试集   compute_metrics=compute_metrics,    #评估函数  )  trainer.train()

trainer.evaluate()