finbert_bert-base-uncased

This model is a fine-tuned version of ProsusAI/finbert on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8116
• Accuracy: 0.8752
• F1: 0.8758
• Precision: 0.8778
• Recall: 0.8752

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: 0.0001
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.8465	1.0	91	0.7610	0.6817	0.6643	0.6806	0.6817
0.5154	2.0	182	0.4672	0.8066	0.8082	0.8203	0.8066
0.331	3.0	273	0.4259	0.8393	0.8396	0.8407	0.8393
0.2461	4.0	364	0.5386	0.8315	0.8311	0.8405	0.8315
0.163	5.0	455	0.5392	0.8518	0.8496	0.8554	0.8518
0.1193	6.0	546	0.5441	0.8565	0.8559	0.8590	0.8565
0.0935	7.0	637	0.6496	0.8253	0.8218	0.8306	0.8253
0.0536	8.0	728	0.5461	0.8612	0.8609	0.8609	0.8612
0.0809	9.0	819	0.6680	0.8362	0.8350	0.8394	0.8362
0.0986	10.0	910	0.6303	0.8596	0.8597	0.8645	0.8596
0.0765	11.0	1001	0.7653	0.8300	0.8310	0.8511	0.8300
0.0507	12.0	1092	0.5176	0.8690	0.8691	0.8701	0.8690
0.0633	13.0	1183	0.9141	0.8268	0.8261	0.8370	0.8268
0.0529	14.0	1274	0.7537	0.8549	0.8552	0.8621	0.8549
0.0418	15.0	1365	0.9200	0.8346	0.8342	0.8441	0.8346
0.0151	16.0	1456	0.8578	0.8565	0.8549	0.8622	0.8565
0.0154	17.0	1547	0.8116	0.8752	0.8758	0.8778	0.8752
0.0054	18.0	1638	0.8926	0.8736	0.8733	0.8751	0.8736
0.0259	19.0	1729	0.9026	0.8705	0.8705	0.8709	0.8705
0.0036	20.0	1820	0.9616	0.8721	0.8713	0.8716	0.8721
0.0012	21.0	1911	0.9985	0.8658	0.8656	0.8655	0.8658
0.002	22.0	2002	0.9833	0.8690	0.8689	0.8688	0.8690

Framework versions

• Transformers 4.37.0
• Pytorch 2.1.2
• Datasets 2.1.0
• Tokenizers 0.15.1