final-ft__bert-base-spanish-wwm-cased__70k-ultrasounds

This model is a fine-tuned version of dccuchile/bert-base-spanish-wwm-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5227

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: 2e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	276	0.9156
No log	2.0	552	0.7658
No log	3.0	828	0.7083
0.9708	4.0	1104	0.6710
0.9708	5.0	1380	0.6380
0.9708	6.0	1656	0.6119
0.9708	7.0	1932	0.6025
0.6612	8.0	2208	0.5752
0.6612	9.0	2484	0.5708
0.6612	10.0	2760	0.5680
0.6612	11.0	3036	0.5425
0.6013	12.0	3312	0.5635
0.6013	13.0	3588	0.5386
0.6013	14.0	3864	0.5476
0.6013	15.0	4140	0.5458
0.5675	16.0	4416	0.5345
0.5675	17.0	4692	0.5350
0.5675	18.0	4968	0.5280
0.5675	19.0	5244	0.5260
0.5549	20.0	5520	0.5227

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1