belisards/azmn-posicao

modelos

This model is a fine-tuned version of belisards/congretimbau on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5239
• Accuracy: 0.8254
• F1: 0.7442
• Recall: 0.7267
• Precision: 0.7727

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: 32
• eval_batch_size: 32
• seed: 5151
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Recall	Precision
0.5681	1.0323	32	0.5508	0.75	0.4286	0.5	0.375
0.5233	2.0645	64	0.5138	0.7381	0.5146	0.5317	0.5897
0.4339	3.0968	96	0.4529	0.7917	0.6875	0.6706	0.7240
0.3907	4.1290	128	0.4087	0.8393	0.7683	0.75	0.7970
0.2166	5.1613	160	0.4054	0.8452	0.7867	0.7778	0.7976
0.14	6.1935	192	0.4474	0.8274	0.7716	0.7738	0.7696
0.0673	7.2258	224	0.5118	0.8393	0.7726	0.7579	0.7932

Framework versions

• Transformers 4.45.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.1