You can deploy this model and start earning money today!

lannelin/
bert-imdb-1hidden

Name: bert-imdb-1hidden
Availability: InStock

Views

No views yet

bert-imdb-1hidden

Model description

A bert-base-uncased model was restricted to 1 hidden layer and fine-tuned for sequence classification on the imdb dataset loaded using the datasets library.

Intended uses & limitations

How to use

python

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)[0].squeeze()

    return LABELS[(output.argmax())]

print(get_sentiment("What a terrible film!"))

Limitations and bias

No special consideration given to limitations and bias.

Any bias held by the imdb dataset may be reflected in the model's output.

Training data

Initialised with bert-base-uncased

Fine tuned on imdb

Training procedure

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.

Eval results

Accuracy on imdb test set: 0.87132