tasks/text.py

from transformers import AutoTokenizer, AutoModelForSequenceClassification
from fastapi import APIRouter
from datetime import datetime
from datasets import load_dataset
from sklearn.metrics import accuracy_score
import torch
from torch.utils.data import Dataset, DataLoader

from .utils.evaluation import TextEvaluationRequest
from .utils.emissions import tracker, clean_emissions_data, get_space_info

router = APIRouter()

DESCRIPTION = "Climate Disinformation Detection"
ROUTE = "/text"

@router.post(ROUTE, tags=["Text Task"], description=DESCRIPTION)
async def evaluate_text(request: TextEvaluationRequest):
    """
    Evaluate text classification for climate disinformation detection.
    """
    # Get space info
    username, space_url = get_space_info()

    # Define the label mapping
    LABEL_MAPPING = {
        "0_not_relevant": 0,
        "1_not_happening": 1,
        "2_not_human": 2,
        "3_not_bad": 3,
        "4_solutions_harmful_unnecessary": 4,
        "5_science_unreliable": 5,
        "6_proponents_biased": 6,
        "7_fossil_fuels_needed": 7
    }

    # Load and prepare the dataset
    dataset = load_dataset(request.dataset_name)

    # Convert string labels to integers
    dataset = dataset.map(lambda x: {"label": LABEL_MAPPING[x["label"]]})

    # Split dataset
    train_test = dataset["train"].train_test_split(test_size=request.test_size, seed=request.test_seed)
    test_dataset = train_test["test"]
    
    # Start tracking emissions
    tracker.start()
    tracker.start_task("inference")

    try:
        # Get texts and labels
        texts = test_dataset["quote"]
        labels = test_dataset["label"]
        
        # Load model and tokenizer from local directory
        model_dir = "./"
        tokenizer = AutoTokenizer.from_pretrained(model_dir)
        model = AutoModelForSequenceClassification.from_pretrained(model_dir)
        
        # Define dataset class
        class TextDataset(Dataset):
            def __init__(self, texts, labels, tokenizer, max_len=128):
                self.texts = texts
                self.labels = labels
                self.tokenizer = tokenizer
                self.max_len = max_len

            def __len__(self):
                return len(self.texts)

            def __getitem__(self, idx):
                text = self.texts[idx]
                label = self.labels[idx]
                encodings = self.tokenizer(
                    text,
                    max_length=self.max_len,
                    padding='max_length',
                    truncation=True,
                    return_tensors="pt"
                )
                return {
                    'input_ids': encodings['input_ids'].squeeze(0),
                    'attention_mask': encodings['attention_mask'].squeeze(0),
                    'labels': torch.tensor(label, dtype=torch.long)
                }

        # Create dataset and dataloader
        test_dataset = TextDataset(texts, labels, tokenizer)
        test_loader = DataLoader(test_dataset, batch_size=16)

        # Model inference
        model.eval()
        predictions = []
        ground_truth = []
        device = 'cpu'
        
        with torch.no_grad():
            for batch in test_loader:
                input_ids = batch['input_ids'].to(device)
                attention_mask = batch['attention_mask'].to(device)
                labels = batch['labels'].to(device)

                outputs = model(input_ids=input_ids, attention_mask=attention_mask)
                _, predicted = torch.max(outputs.logits, 1)

                predictions.extend(predicted.cpu().numpy())
                ground_truth.extend(labels.cpu().numpy())

        # Stop tracking emissions
        emissions_data = tracker.stop_task()
        
        # Calculate accuracy
        accuracy = accuracy_score(test_dataset["label"], predictions)
        
        # Prepare results
        results = {
            "username": username,
            "space_url": space_url,
            "submission_timestamp": datetime.now().isoformat(),
            "model_description": DESCRIPTION,
            "accuracy": float(accuracy),
            "energy_consumed_wh": emissions_data.energy_consumed * 1000,
            "emissions_gco2eq": emissions_data.emissions * 1000,
            "emissions_data": clean_emissions_data(emissions_data),
            "api_route": ROUTE,
            "dataset_config": {
                "dataset_name": request.dataset_name,
                "test_size": request.test_size,
                "test_seed": request.test_seed
            }
        }
        
        return results
        
    except Exception as e:
        # Stop tracking in case of error
        tracker.stop_task()
        raise e