File size: 1,730 Bytes
969cd78 |
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 |
import json
import torch
import torch.nn.functional as F
from transformers import BertTokenizer
from .sentiment_classifier import SentimentClassifier
with open("config.json") as json_file:
config = json.load(json_file)
class Model:
def __init__(self):
self.device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
self.tokenizer = BertTokenizer.from_pretrained(config["BERT_MODEL"])
classifier = SentimentClassifier(len(config["CLASS_NAMES"]))
classifier.load_state_dict(
torch.load(config["PRE_TRAINED_MODEL"], map_location=self.device)
)
classifier = classifier.eval()
self.classifier = classifier.to(self.device)
def predict(self, text):
encoded_text = self.tokenizer.encode_plus(
text,
max_length=config["MAX_SEQUENCE_LEN"],
add_special_tokens=True,
return_token_type_ids=False,
pad_to_max_length=True,
return_attention_mask=True,
return_tensors="pt",
)
input_ids = encoded_text["input_ids"].to(self.device)
attention_mask = encoded_text["attention_mask"].to(self.device)
with torch.no_grad():
probabilities = F.softmax(self.classifier(input_ids, attention_mask), dim=1)
confidence, predicted_class = torch.max(probabilities, dim=1)
predicted_class = predicted_class.cpu().item()
probabilities = probabilities.flatten().cpu().numpy().tolist()
return (
config["CLASS_NAMES"][predicted_class],
confidence,
dict(zip(config["CLASS_NAMES"], probabilities)),
)
model = Model()
def get_model():
return model
|