ddiddu's picture
Update README.md
967accb
metadata
license: mit
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: bert_financial_phrasebank
    results: []
datasets:
  - financial_phrasebank
library_name: transformers

tps_sentimental_analysis

This model is a fine-tuned version of bert-base-uncased on the financial_phrasebank dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2586
  • Accuracy: 0.9604

Model description

A fine-tuned version of bert-base-uncased

Intended uses & limitations

Sentimental Analysis

Lines Emotions
Hi, Harper. I’m really happy you came. Positive
Happy Father’s Day. Positive
It was Christmas. Neutral
HARPER sits at a table alone in a room. Neutral
I am mad at you badly. Negative

Training and evaluation data

financial_phrasebank

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 1.0 114 0.5293 0.8230
No log 2.0 228 0.0804 0.9779
No log 3.0 342 0.0367 0.9867
No log 4.0 456 0.1544 0.9646
0.3241 5.0 570 0.0497 0.9912
0.3241 6.0 684 0.0520 0.9912
0.3241 7.0 798 0.0318 0.9912
0.3241 8.0 912 0.0628 0.9912
0.0218 9.0 1026 0.0777 0.9867
0.0218 10.0 1140 0.0866 0.9867

Framework versions

  • Transformers 4.30.1
  • Pytorch 2.1.0+cu118
  • Tokenizers 0.13.3