bert-base-uncased model was restricted to 1 hidden layer and
fine-tuned for sequence classification on the
imdb dataset loaded using the
from transformers import AutoTokenizer, AutoModelForSequenceClassification pretrained = "lannelin/bert-imdb-1hidden" tokenizer = AutoTokenizer.from_pretrained(pretrained) model = AutoModelForSequenceClassification.from_pretrained(pretrained) LABELS = ["negative", "positive"] def get_sentiment(text: str): inputs = tokenizer.encode_plus(text, return_tensors='pt') output = model(**inputs).squeeze() return LABELS[(output.argmax())] print(get_sentiment("What a terrible film!"))
No special consideration given to limitations and bias.
Any bias held by the imdb dataset may be reflected in the model's output.
Initialised with bert-base-uncased
Fine tuned on imdb
The model was fine-tuned for 1 epoch with a batch size of 64, a learning rate of 5e-5, and a maximum sequence length of 512.
Accuracy on imdb test set: 0.87132
- Downloads last month