metadata

license: apache-2.0
base_model: Twitter/twhin-bert-large
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
model-index:
  - name: financial-twhin-bert-large-3labels-pesudo-bb-only
    results: []
datasets:
  - kekunh/stock-related-tweets
  - zeroshot/twitter-financial-news-sentiment
language:
  - en
widget:
  - text: >-
      $KTOS: Kratos Defense and Security awarded a $39 million sole-source
      contract for Geolocation Global Support Service
    example_title: Example 1
  - text: >-
      $Google parent Alphabet Inc. reported revenue and earnings that fell short
      of analysts' expectations, showing the company's search advertising
      juggernaut was not immune to a slowdown in the digital ad market. The
      shares fell more than 6%.
    example_title: Example 2
  - text: $LJPC - La Jolla Pharma to reassess development of LJPC-401
    example_title: Example 3
  - text: >-
      Watch $MARK over 43c in after-hours for continuation targeting the 50c
      area initially
    example title: Example 4
  - text: >-
      $RCII: Rent-A-Center provides update - March revenues were off by about 5%
      versus last year
    example title: Example 5

financial-twhin-bert-large-3labels-pesudo-bb-only

This model is a fine-tuned version of Twitter/twhin-bert-large on finance-related tweets. It achieves the following results on the evaluation set:

Loss: 0.4379
Accuracy: 0.8847
F1: 0.8857

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: 3.812006227593217e-05
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.5859	1.0	2283	1.5711	0.2039	0.0800
0.1601	2.0	4566	0.4379	0.8847	0.8857
0.0875	3.0	6849	0.5111	0.8854	0.8868

Framework versions

Transformers 4.37.2
Pytorch 2.1.0+cu121
Datasets 2.17.1
Tokenizers 0.15.2