Modelpublic

hw2942/Erlangshen-Longformer-330M-finetuning-wallstreetcn-morning-news-vix-sz50-1

sourceHugging Faceapache-2.0updated 3y agoView on Hugging Face

0likes12downloads

Model Card

Erlangshen-Longformer-330M-finetuning-wallstreetcn-morning-news-vix-sz50-1

This model is a fine-tuned version of IDEA-CCNL/Erlangshen-Longformer-330M on an unknown dataset. It achieves the following results on the evaluation set:

—Loss: 0.6971
—Accuracy: 0.5

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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	150	0.7366	0.5
No log	2.0	300	0.7063	0.5
No log	3.0	450	0.6971	0.5
0.7857	4.0	600	0.6971	0.5
0.7857	5.0	750	0.6971	0.5
0.7857	6.0	900	0.6971	0.5
0.696	7.0	1050	0.6971	0.5
0.696	8.0	1200	0.6971	0.5
0.696	9.0	1350	0.6971	0.5
0.6916	10.0	1500	0.6971	0.5

Framework versions

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