MBERTbase_REDv2

This model is a fine-tuned version of google-bert/bert-base-multilingual-cased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3320
• F1: 0.5326
• Roc Auc: 0.7058
• Accuracy: 0.4383

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: 1e-05
• 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
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
No log	1.0	64	0.4206	0.0	0.5	0.0
No log	2.0	128	0.3876	0.0901	0.5328	0.0589
No log	3.0	192	0.3599	0.2983	0.5993	0.2081
No log	4.0	256	0.3434	0.3808	0.6365	0.2965
No log	5.0	320	0.3360	0.4182	0.6474	0.3204
No log	6.0	384	0.3267	0.4638	0.6703	0.3646
No log	7.0	448	0.3259	0.5033	0.6945	0.3959
0.3376	8.0	512	0.3226	0.5140	0.6978	0.4217
0.3376	9.0	576	0.3248	0.5099	0.6959	0.4199
0.3376	10.0	640	0.3252	0.5230	0.6988	0.4162
0.3376	11.0	704	0.3258	0.5211	0.7027	0.4217
0.3376	12.0	768	0.3308	0.5214	0.6998	0.4309
0.3376	13.0	832	0.3304	0.5305	0.7052	0.4383
0.3376	14.0	896	0.3318	0.5297	0.7054	0.4309
0.3376	15.0	960	0.3320	0.5326	0.7058	0.4383

Framework versions

• Transformers 4.45.1
• Pytorch 2.4.0
• Datasets 3.0.1
• Tokenizers 0.20.0