hfnlpmodels/eos_prediction_distilbert_1
This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:
- Train Loss: 0.0879
- Train Accuracy: 0.9714
- Validation Loss: 0.2180
- Validation Accuracy: 0.9305
- Epoch: 2
Model description
Predicts whether a given sentence is complete. Trained on sentences that were randomly truncate as '0' labels, hence some sentences which were grammatically correct may have been labelled as incomplete. Overall accuracy near 0.9 means that this was most likely a small factor.
Intended uses & limitations
Found better performance on shorter sentences.
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 4165, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
- training_precision: float32
Training results
Train Loss | Train Accuracy | Validation Loss | Validation Accuracy | Epoch |
---|---|---|---|---|
0.2709 | 0.8942 | 0.2015 | 0.9211 | 0 |
0.1421 | 0.9505 | 0.2055 | 0.9318 | 1 |
0.0879 | 0.9714 | 0.2180 | 0.9305 | 2 |
Framework versions
- Transformers 4.38.2
- TensorFlow 2.15.0
- Tokenizers 0.15.2
- Downloads last month
- 24
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for hfnlpmodels/eos_prediction_distilbert_1
Base model
distilbert/distilbert-base-uncased