File size: 5,556 Bytes
4d6e8c2
 
 
 
ef8f2ae
 
 
4d6e8c2
 
 
 
 
 
eba6cfa
1c33274
70f5f26
1c33274
70f5f26
4d6e8c2
 
70f5f26
 
 
 
 
4d6e8c2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
70f5f26
 
 
 
 
4d6e8c2
 
 
ef8f2ae
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
eba6cfa
 
 
 
 
 
 
 
ef8f2ae
 
eba6cfa
 
ef8f2ae
eba6cfa
 
 
 
 
 
 
 
 
 
 
ef8f2ae
eba6cfa
 
70f5f26
 
 
 
 
4d6e8c2
 
 
 
 
 
 
 
 
 
 
 
70f5f26
4d6e8c2
 
 
 
1c33274
4d6e8c2
 
 
 
 
 
 
eba6cfa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
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 transformers import AutoTokenizer, AutoModelForSequenceClassification

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

router = APIRouter()

DESCRIPTION = "Evaluate text classification for 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.
    
    Current Model: Random Baseline
    - Makes random predictions from the label space (0-7)
    - Used as a baseline for comparison
    """
    # 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")

    #--------------------------------------------------------------------------------------------
    # YOUR MODEL INFERENCE CODE HERE
    # Update the code below to replace the random baseline by your model inference within the inference pass where the energy consumption and emissions are tracked.
    #--------------------------------------------------------------------------------------------   
    
    # Make random predictions (placeholder for actual model inference)
    true_labels = test_dataset["label"]
    texts=test_dataset["quote"]
    labels=test_dataset["label"]

    model_dir = "./"

    tokenizer = AutoTokenizer.from_pretrained(model_dir)
    model = AutoModelForSequenceClassification.from_pretrained(model_dir)

    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)
            }

    test_dataset = TextDataset(texts, labels, tokenizer)
    test_loader = DataLoader(test_dataset, batch_size=16)

    # model.eval()
    # predictions = []
    # with torch.no_grad():
    #     for inputs, labels in test_loader:
    #         inputs, labels = inputs.to('cpu'), labels.to('cpu')
    #         outputs = model(inputs)
    #         _, predicted = torch.max(outputs, 1)
    #         predictions.extend(predicted.cpu().numpy())
    model.eval()
    predictions = []
    ground_truth = []
    DEVICE='cpu'
    with torch.no_grad():
        for batch in test_loader:
            # Access each component of the batch dictionary
            input_ids = batch['input_ids'].to(DEVICE)
            attention_mask = batch['attention_mask'].to(DEVICE)
            labels = batch['labels'].to(DEVICE)
    
            # Forward pass
            outputs = model(input_ids=input_ids, attention_mask=attention_mask)
            _, predicted = torch.max(outputs.logits, 1)
    
            # Store predictions and ground truth
            predictions.extend(predicted.cpu().numpy())
            ground_truth.extend(labels.cpu().numpy())


    #--------------------------------------------------------------------------------------------
    # YOUR MODEL INFERENCE STOPS HERE
    #--------------------------------------------------------------------------------------------   

    
    # Stop tracking emissions
    emissions_data = tracker.stop_task()
    
    # Calculate accuracy
    accuracy = accuracy_score(true_labels, predictions)
    
    # Prepare results dictionary
    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