sentiment-seq_bn-1

This model is a fine-tuned version of indolem/indobert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3078
• Accuracy: 0.8797
• Precision: 0.8572
• Recall: 0.8499
• F1: 0.8534

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
• train_batch_size: 30
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.5593	1.0	122	0.5129	0.7318	0.6697	0.6453	0.6532
0.481	2.0	244	0.4831	0.7343	0.6993	0.7295	0.7054
0.4234	3.0	366	0.3974	0.8221	0.7926	0.7616	0.7740
0.3701	4.0	488	0.3780	0.8396	0.8128	0.7890	0.7992
0.3499	5.0	610	0.3612	0.8471	0.8135	0.8268	0.8195
0.3165	6.0	732	0.3760	0.8271	0.7953	0.8377	0.8072
0.2968	7.0	854	0.3342	0.8697	0.8438	0.8403	0.8420
0.2812	8.0	976	0.3311	0.8672	0.8463	0.8260	0.8351
0.2682	9.0	1098	0.3269	0.8722	0.8463	0.8446	0.8454
0.2596	10.0	1220	0.3145	0.8797	0.8560	0.8524	0.8541
0.2464	11.0	1342	0.3138	0.8697	0.8503	0.8278	0.8377
0.2415	12.0	1464	0.3126	0.8847	0.8697	0.8459	0.8565
0.2354	13.0	1586	0.3136	0.8822	0.8694	0.8392	0.8521
0.2303	14.0	1708	0.3172	0.8747	0.8463	0.8563	0.8510
0.2172	15.0	1830	0.3120	0.8822	0.8656	0.8442	0.8537
0.2159	16.0	1952	0.3116	0.8622	0.8319	0.8400	0.8357
0.2192	17.0	2074	0.3123	0.8847	0.8717	0.8434	0.8557
0.2124	18.0	2196	0.3150	0.8647	0.8340	0.8467	0.8399
0.2077	19.0	2318	0.3084	0.8797	0.8585	0.8474	0.8526
0.205	20.0	2440	0.3078	0.8797	0.8572	0.8499	0.8534

Framework versions

• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1