CoolFace
Modelpublic

hw2942/Erlangshen-Longformer-110M-finetuning-wallstreetcn-morning-news-vix-sz50-v2

sourceHugging Faceapache-2.0updated 3y agoView on Hugging Face
0likes7downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

Erlangshen-Longformer-110M-finetuning-wallstreetcn-morning-news-vix-sz50-v2

This model is a fine-tuned version of IDEA-CCNL/Erlangshen-Longformer-110M on the dataset of Wallstreetcn Morning News(Sample link: https://wallstreetcn.com/articles/3691904?keyword=%E8%A7%81%E9%97%BB%E6%97%A9%E9%A4%90) to predict VIX of SZ50 up or down. It achieves the following results on the evaluation set:

  • —Loss: 3.5413
  • —Accuracy: 0.5577

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: 5e-05
  • —trainbatchsize: 8
  • —evalbatchsize: 8
  • —seed: 42
  • —optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • —lrschedulertype: linear
  • —num_epochs: 20

Training results

Training LossEpochStepValidation LossAccuracy
No log1.0190.68470.5962
No log2.0380.99580.5385
No log3.0571.07580.5962
No log4.0761.67910.5769
No log5.0952.34080.5385
No log6.01142.46840.5385
No log7.01332.84360.5385
No log8.01522.58920.5385
No log9.01712.56520.5
No log10.01903.03160.5385
No log11.02092.98710.5577
No log12.02283.44730.5385
No log13.02473.39940.5192
No log14.02663.41790.5
No log15.02853.45920.5192
No log16.03043.49510.5192
No log17.03233.51340.5577
No log18.03423.52890.5577
No log19.03613.53800.5577
No log20.03803.54130.5577

Framework versions

  • —Transformers 4.30.2
  • —Pytorch 2.0.1+cu118
  • —Datasets 2.13.1
  • —Tokenizers 0.13.3