|
import gradio as gr |
|
from transformers import pipeline, Trainer, TrainingArguments, DistilBertForSequenceClassification, DistilBertTokenizer |
|
from datasets import load_dataset |
|
import torch |
|
import os |
|
|
|
|
|
dataset = load_dataset("tanquangduong/spam-detection-dataset-splits") |
|
|
|
|
|
tokenizer = DistilBertTokenizer.from_pretrained("distilbert-base-uncased") |
|
model = DistilBertForSequenceClassification.from_pretrained("distilbert-base-uncased") |
|
|
|
|
|
def tokenize_function(examples): |
|
return tokenizer(examples['text'], truncation=True, padding="max_length", max_length=128) |
|
|
|
tokenized_datasets = dataset.map(tokenize_function, batched=True) |
|
|
|
|
|
train_dataset = tokenized_datasets["train"].shuffle(seed=42).select(range(2000)) |
|
test_dataset = tokenized_datasets["test"].shuffle(seed=42).select(range(500)) |
|
|
|
|
|
training_args = TrainingArguments( |
|
output_dir="./results", |
|
evaluation_strategy="epoch", |
|
save_strategy="epoch", |
|
learning_rate=2e-5, |
|
per_device_train_batch_size=16, |
|
per_device_eval_batch_size=16, |
|
num_train_epochs=1, |
|
weight_decay=0.01, |
|
save_total_limit=2, |
|
load_best_model_at_end=True, |
|
) |
|
|
|
|
|
trainer = Trainer( |
|
model=model, |
|
args=training_args, |
|
train_dataset=train_dataset, |
|
eval_dataset=test_dataset, |
|
) |
|
|
|
|
|
if os.path.exists("./results/checkpoint-1"): |
|
print("Riprendi l'addestramento dal checkpoint...") |
|
trainer.train(resume_from_checkpoint="./results/checkpoint-1") |
|
else: |
|
print("Inizia l'addestramento da zero...") |
|
trainer.train() |
|
|
|
|
|
def classify_email(text): |
|
classifier = pipeline("text-classification", model=model, tokenizer=tokenizer, framework="pt") |
|
result = classifier(text) |
|
label = result[0]['label'] |
|
score = result[0]['score'] |
|
return {label: score} |
|
|
|
|
|
iface = gr.Interface(fn=classify_email, |
|
inputs="text", |
|
outputs="label", |
|
title="ZeroSpam Email Classifier", |
|
description="Inserisci l'email da analizzare per determinare se è spam o phishing.") |
|
|
|
|
|
iface.launch(share=True) |
|
|