finetuning.py

import torch
from transformers import DistilBertTokenizerFast, DistilBertForSequenceClassification, Trainer, TrainingArguments
import pandas as pd
from datasets import load_dataset
from transformers import AutoTokenizer
from transformers import DataCollatorWithPadding
import evaluate
import numpy as np
from transformers import AutoModelForSequenceClassification, TrainingArguments, Trainer

#Prepare and load dataset
dataset = load_dataset('csv', data_files={'train': 'df_train.csv','test': 'df_test.csv'})
                                              
#Tonenizer and preprocessing function                                              
tokenizer = AutoTokenizer.from_pretrained('distilbert-base-uncased')

def preprocess_function(examples):
    return tokenizer(examples["text"], truncation=True,padding=True)

tokenized_data = dataset.map(preprocess_function, batched=True)
data_collator = DataCollatorWithPadding(tokenizer=tokenizer)

#Prepare accuracy functions
accuracy = evaluate.load("accuracy")

def compute_metrics(eval_pred):
    predictions, labels = eval_pred
    predictions = np.argmax(predictions, axis=1)
    print(predictions)
    return accuracy.compute(predictions=predictions, references=labels)
    
#Map the labels to the values in the column (here its 1 to 1, 0 to 0)
id2label = {0: 0, 1: 1}
label2id = {0: 0, 1: 1}

#Load model
model = AutoModelForSequenceClassification.from_pretrained(
    "distilbert-base-uncased", num_labels=2, id2label=id2label, label2id=label2id)

#Define the training arguments
training_args = TrainingArguments(
    output_dir="classifier",
    learning_rate=2e-5,
    per_device_train_batch_size=8,
    per_device_eval_batch_size=8,
    num_train_epochs=8,
    weight_decay=0.01,
    evaluation_strategy="epoch",
    save_strategy="epoch",
    load_best_model_at_end=True)

#Definer the trainer and train the model
trainer = Trainer(
    model=model,
    args=training_args,
    train_dataset=tokenized_data["train"],
    eval_dataset=tokenized_data["test"],
    tokenizer=tokenizer,
    data_collator=data_collator,
    compute_metrics=compute_metrics)

trainer.train()