finbert_roberta-base

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

• Loss: 0.7907
• Accuracy: 0.9033
• F1: 0.9034
• Precision: 0.9036
• Recall: 0.9033

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
0.8094	1.0	91	0.7239	0.6942	0.6824	0.6887	0.6942
0.439	2.0	182	0.4112	0.8471	0.8476	0.8527	0.8471
0.274	3.0	273	0.3978	0.8612	0.8596	0.8623	0.8612
0.2002	4.0	364	0.4319	0.8409	0.8399	0.8430	0.8409
0.123	5.0	455	0.4685	0.8674	0.8661	0.8685	0.8674
0.1251	6.0	546	0.4734	0.8690	0.8684	0.8689	0.8690
0.124	7.0	637	0.5604	0.8580	0.8574	0.8610	0.8580
0.0738	8.0	728	0.5583	0.8534	0.8546	0.8604	0.8534
0.1268	9.0	819	0.5665	0.8534	0.8524	0.8537	0.8534
0.0425	10.0	910	0.5959	0.8549	0.8561	0.8626	0.8549
0.1037	11.0	1001	0.4439	0.8752	0.8742	0.8760	0.8752
0.0762	12.0	1092	0.5998	0.8674	0.8668	0.8686	0.8674
0.0523	13.0	1183	0.5525	0.8783	0.8785	0.8792	0.8783
0.0291	14.0	1274	0.6588	0.8752	0.8747	0.8756	0.8752
0.0311	15.0	1365	0.6065	0.8830	0.8833	0.8839	0.8830
0.0146	16.0	1456	0.7469	0.8705	0.8701	0.8706	0.8705
0.0145	17.0	1547	0.6748	0.8861	0.8864	0.8872	0.8861
0.0013	18.0	1638	0.7708	0.8814	0.8815	0.8816	0.8814
0.0105	19.0	1729	0.8126	0.8908	0.8910	0.8918	0.8908
0.0025	20.0	1820	0.7727	0.8939	0.8938	0.8957	0.8939
0.0014	21.0	1911	0.8088	0.8939	0.8942	0.8958	0.8939
0.0015	22.0	2002	0.7766	0.9033	0.9033	0.9034	0.9033
0.0001	23.0	2093	0.7907	0.9033	0.9034	0.9036	0.9033
0.0002	24.0	2184	0.7945	0.9033	0.9034	0.9036	0.9033
0.0002	25.0	2275	0.7954	0.9033	0.9034	0.9036	0.9033

Framework versions

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