| --- |
| license: apache-2.0 |
| --- |
| |
| model base: https://huggingface.co/microsoft/deberta-v3-base |
|
|
| dataset: https://github.com/ramybaly/Article-Bias-Prediction |
|
|
|
|
| training parameters: |
| - devices: 2xH100 |
| - batch_size: 100 |
| - epochs: 5 |
| - dropout: 0.05 |
| - max_length: 512 |
| - learning_rate: 3e-5 |
| - warmup_steps: 100 |
| - random_state: 239 |
| |
| |
| training methodology: |
| - sanitize dataset following specific rule-set, utilize random split as provided in the dataset |
| - train on train split and evaluate on validation split in each epoch |
| - evaluate test split only on the model that performed best on validation loss |
| |
| result summary: |
| - throughout the five training epochs, model of fourth epoch achieved the lowest validation loss of 0.1909 |
| - on test split fourth epoch model achieved f1 score of 0.9427 and a test loss of 0.2168 |
| |
| usage: |
| |
| ``` |
| model = AutoModelForSequenceClassification.from_pretrained("premsa/political-bias-prediction-allsides-DeBERTa") |
| tokenizer = AutoTokenizer.from_pretrained("premsa/political-bias-prediction-allsides-DeBERTa") |
| nlp = pipeline("text-classification", model=model, tokenizer=tokenizer) |
| print(nlp("the masses are controlled by media.")) |
| ``` |
| |
