import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM, Trainer, TrainingArguments
from datasets import load_dataset

# Model ve tokenizer
model_name = "distilgpt2"  # veya kendi modelin "1c1/7cpc"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    device_map="auto",
    load_in_8bit=True
)

# Dataset örnek
dataset = load_dataset("wikitext", "wikitext-2-raw-v1", split="train")

def tokenize_function(examples):
    return tokenizer(examples["text"], truncation=True, padding="max_length", max_length=128)

tokenized_dataset = dataset.map(tokenize_function, batched=True)

# Training ayarları
training_args = TrainingArguments(
    output_dir="./results",
    num_train_epochs=1,
    per_device_train_batch_size=1,
    save_steps=500,
    save_total_limit=2,
    logging_dir="./logs",
    logging_steps=50,
)

trainer = Trainer(
    model=model,
    args=training_args,
    train_dataset=tokenized_dataset,
)

# Eğitim fonksiyonu
def train_model():
    trainer.train()
    return "Model eğitildi!"

# Gradio arayüzü
demo = gr.Interface(fn=train_model, inputs=[], outputs="text")
demo.launch()