|
import torch |
|
from typing import Dict, List, Any |
|
from tokenizers import Tokenizer |
|
import sys |
|
import os |
|
import warnings |
|
|
|
|
|
sys.path.append(os.path.dirname(__file__)) |
|
|
|
from model import build_transformer |
|
|
|
warnings.simplefilter("ignore", category=FutureWarning) |
|
|
|
|
|
class EndpointHandler: |
|
def __init__(self, path: str = ""): |
|
""" |
|
Initialize the handler. Load the model and tokenizer. |
|
""" |
|
device = torch.device("cuda" if torch.cuda.is_available() else "cpu") |
|
self.device = device |
|
|
|
|
|
self.model_weights_path = os.path.join(path, "EN-IT.pt") |
|
self.tokenizer_src_path = os.path.join(path, "tokenizer_en.json") |
|
self.tokenizer_tgt_path = os.path.join(path, "tokenizer_it.json") |
|
|
|
|
|
self.tokenizer_src = Tokenizer.from_file(self.tokenizer_src_path) |
|
self.tokenizer_tgt = Tokenizer.from_file(self.tokenizer_tgt_path) |
|
|
|
|
|
self.model = build_transformer( |
|
src_vocab_size=self.tokenizer_src.get_vocab_size(), |
|
tgt_vocab_size=self.tokenizer_tgt.get_vocab_size(), |
|
src_seq_len=350, |
|
tgt_seq_len=350, |
|
d_model=512, |
|
num_layers=6, |
|
num_heads=8, |
|
dropout=0.1, |
|
d_ff=2048 |
|
).to(self.device) |
|
|
|
|
|
print(f"Loading weights from: {self.model_weights_path}") |
|
checkpoint = torch.load(self.model_weights_path, map_location=self.device) |
|
self.model.load_state_dict(checkpoint["model_state_dict"]) |
|
self.model.eval() |
|
|
|
def __call__(self, data: Dict[str, Any]) -> List[Dict[str, Any]]: |
|
""" |
|
Process the incoming request and return the translation. |
|
""" |
|
try: |
|
inputs = data.get("inputs", "") |
|
if not inputs: |
|
return [{"error": "No 'inputs' provided in request"}] |
|
|
|
source = self.tokenizer_src.encode(inputs) |
|
source = torch.cat([ |
|
torch.tensor([self.tokenizer_src.token_to_id("[SOS]")], dtype=torch.int64), |
|
torch.tensor(source.ids, dtype=torch.int64), |
|
torch.tensor([self.tokenizer_src.token_to_id("[EOS]")], dtype=torch.int64), |
|
torch.tensor([self.tokenizer_src.token_to_id("[PAD]")] * (350 - len(source.ids) - 2), dtype=torch.int64) |
|
], dim=0).to(self.device) |
|
|
|
source_mask = (source != self.tokenizer_src.token_to_id("[PAD]")).unsqueeze(0).unsqueeze(1).int().to(self.device) |
|
encoder_output = self.model.encode(source, source_mask) |
|
|
|
decoder_input = torch.empty(1, 1).fill_(self.tokenizer_tgt.token_to_id("[SOS]")).type_as(source).to(self.device) |
|
predicted_words = [] |
|
|
|
while decoder_input.size(1) < 350: |
|
decoder_mask = torch.triu( |
|
torch.ones((1, decoder_input.size(1), decoder_input.size(1))), |
|
diagonal=1 |
|
).type(torch.int).type_as(source_mask).to(self.device) |
|
|
|
out = self.model.decode(encoder_output, source_mask, decoder_input, decoder_mask) |
|
prob = self.model.project(out[:, -1]) |
|
_, next_word = torch.max(prob, dim=1) |
|
|
|
decoder_input = torch.cat( |
|
[decoder_input, torch.empty(1, 1).type_as(source).fill_(next_word.item()).to(self.device)], dim=1) |
|
|
|
decoded_word = self.tokenizer_tgt.decode([next_word.item()]) |
|
if next_word == self.tokenizer_tgt.token_to_id("[EOS]"): |
|
break |
|
|
|
predicted_words.append(decoded_word) |
|
|
|
predicted_translation = " ".join(predicted_words).replace("[EOS]", "").strip() |
|
|
|
return [{"translation": predicted_translation}] |
|
except Exception as e: |
|
return [{"error": str(e)}] |
|
|