mt5-base-bias-aware-v2

This model is a fine-tuned version of google/mt5-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7158

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: 48
• eval_batch_size: 48
• seed: 42
• optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
0.9290	1.0	1593	0.7520
0.7912	2.0	3186	0.6916
0.6925	3.0	4779	0.6674
0.6328	4.0	6372	0.6530
0.5876	5.0	7965	0.6518
0.5512	6.0	9558	0.6486
0.5192	7.0	11151	0.6481
0.4837	8.0	12744	0.6568
0.4569	9.0	14337	0.6597
0.4498	10.0	15930	0.6719
0.4183	11.0	17523	0.6749
0.4106	12.0	19116	0.6812
0.3903	13.0	20709	0.6894
0.3725	14.0	22302	0.6962
0.3683	15.0	23895	0.6999
0.3617	16.0	25488	0.7060
0.3462	17.0	27081	0.7104
0.3505	18.0	28674	0.7125
0.3445	19.0	30267	0.7138
0.3416	20.0	31860	0.7158

Framework versions

• Transformers 5.3.0
• Pytorch 2.10.0+cu128
• Datasets 4.8.3
• Tokenizers 0.22.2