File size: 5,096 Bytes
6278e08
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
from fastapi import APIRouter
from datetime import datetime
from datasets import load_dataset
from sklearn.metrics import accuracy_score
import random

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

#packages needed for inference
from sentence_transformers import SentenceTransformer
from xgboost import XGBClassifier
import pickle
import torch
import os

router = APIRouter()

DESCRIPTION = "Embedding + Neural Network"
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

    # Set the device to MPS (if available)
    device = torch.device("mps" if torch.backends.mps.is_available() else "cpu")
    print(f"Using device: {device}")

    model_name = "sentence-transformers/paraphrase-multilingual-mpnet-base-v2"  # You can use other Sentence Transformers models as needed
    sentence_model = SentenceTransformer(model_name)

    # Convert each sentence into a vector representation (embedding)
    embeddings = sentence_model.encode(test_dataset['quote'], convert_to_tensor=True)

    # Make random predictions (placeholder for actual model inference)
    true_labels = test_dataset["label"]

    """
    from torch import nn, optim

    class SimpleNN2(nn.Module):
        def __init__(self, input_dim, output_dim):
            super(SimpleNN2, self).__init__()
            self.fc1 = nn.Linear(input_dim, 128)  # Reduce hidden units
            self.fc2 = nn.Linear(128, 64)  # Further reduce units
            self.fc3 = nn.Linear(64, output_dim)
            self.relu = nn.ReLU()
            self.dropout = nn.Dropout(0.3)  # Add dropout
            self.batch_norm1 = nn.BatchNorm1d(128)
            self.batch_norm2 = nn.BatchNorm1d(64)

        def forward(self, x):
            x = self.relu(self.batch_norm1(self.fc1(x)))
            x = self.dropout(x)  # Apply dropout
            x = self.relu(self.batch_norm2(self.fc2(x)))
            x = self.dropout(x)  # Apply dropout
            x = self.fc3(x)  # Output raw logits
            return x
    """

    current_file_path = os.path.abspath(__file__)
    current_dir = os.path.dirname(current_file_path)

    # model_nn = torch.load(os.path.join(current_dir,"model_nn.pth"), map_location=device)
    model_nn = torch.jit.load(os.path.join(current_dir,"model_nn_scripted.pth"), map_location=device)  
        

    # Set the model to evaluation mode
    model_nn.eval()

    # Make predictions
    with torch.no_grad():
        outputs = model_nn(embeddings)
        _, predicted = torch.max(outputs, 1)  # Get the class with the highest score

    # Decode the predictions back to original labels using label_encoder
    predictions = predicted.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