eeshclusive's picture
Training complete
b52010d
metadata
base_model: allenai/scibert_scivocab_cased
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: scibert-finetuned-ner
    results: []

scibert-finetuned-ner

This model is a fine-tuned version of allenai/scibert_scivocab_cased on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.3459
  • Precision: 0.5666
  • Recall: 0.5191
  • F1: 0.5418
  • Accuracy: 0.9363

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 2e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
No log 1.0 121 0.3648 0.3157 0.3390 0.3269 0.8945
No log 2.0 242 0.3177 0.5280 0.3348 0.4097 0.9253
No log 3.0 363 0.2599 0.5143 0.4326 0.4700 0.9315
No log 4.0 484 0.2825 0.5360 0.4227 0.4726 0.9336
0.2574 5.0 605 0.2968 0.5473 0.4922 0.5183 0.9350
0.2574 6.0 726 0.3193 0.5857 0.4894 0.5332 0.9377
0.2574 7.0 847 0.3327 0.5513 0.4879 0.5177 0.9356
0.2574 8.0 968 0.3315 0.5658 0.5121 0.5376 0.9363
0.0678 9.0 1089 0.3413 0.5465 0.5163 0.5310 0.9361
0.0678 10.0 1210 0.3459 0.5666 0.5191 0.5418 0.9363

Framework versions

  • Transformers 4.34.1
  • Pytorch 2.1.0+cu118
  • Datasets 2.14.6
  • Tokenizers 0.14.1