inference.py

import torch        # Import PyTorch
import torch.optim as optim
import torch.optim.lr_scheduler as lr_scheduler
from torch.utils.data import DataLoader
from torch import nn
from transformers import AutoModel, AutoTokenizer

class DebertaEvaluator(nn.Module):
    
    def __init__(self):
        super().__init__()
        
        self.deberta = AutoModel.from_pretrained('microsoft/deberta-v3-base')
        self.dropout = nn.Dropout(0.5)
        self.linear = nn.Linear(768, 6)
        
    def forward(self, input_id, mask):
        output = self.deberta(input_ids=input_id, attention_mask=mask)
        output_pooled = torch.mean(output.last_hidden_state, 1)
        dropout_output = self.dropout(output_pooled)
        linear_output = self.linear(dropout_output)
        
        return linear_output

def inference(input_text):
    saved_model_path = './'
    model = torch.load(saved_model_path + 'fine-tuned-model.pt', map_location=torch.device('cpu'))
    tokenizer = torch.load(saved_model_path + 'fine-tuned-tokenizer.pt', map_location=torch.device('cpu'))
    model.eval()
    input = tokenizer(input_text)
    input_ids = torch.Tensor(input['input_ids']).to(torch.device('cpu')).long()
    input_ids.resize_(1,len(input_ids))
    print(input_ids)
    mask = torch.Tensor(input['attention_mask']).to(torch.device('cpu'))
    mask.resize_(1, len(mask))
    output = model(input_ids, mask)
    
    return output.tolist()

if __name__ == "__main__":
    inference()