distilbert-base-multilingual-cased_regression_finetuned_dcard

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

• Loss: 0.6389
• Mse: 0.6389
• Mae: 0.5015
• Rmse: 0.7993
• Mape: inf
• R Squared: 0.6036

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 891
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Mae	Rmse	Mape	R Squared
1.0667	1.0	8917	0.9488	0.9488	0.6910	0.9741	inf	0.4113
0.8502	2.0	17834	0.7789	0.7789	0.6072	0.8825	inf	0.5167
0.6093	3.0	26751	0.7659	0.7659	0.5919	0.8751	inf	0.5248
0.5891	4.0	35668	0.7029	0.7029	0.5537	0.8384	inf	0.5639
0.5542	5.0	44585	0.6521	0.6521	0.5156	0.8075	inf	0.5954
0.5475	6.0	53502	0.6414	0.6414	0.5087	0.8009	inf	0.6020
0.4619	7.0	62419	0.6389	0.6389	0.5015	0.7993	inf	0.6036
0.4368	8.0	71336	0.6471	0.6471	0.5014	0.8044	inf	0.5985
0.4106	9.0	80253	0.6568	0.6568	0.5036	0.8104	inf	0.5925

Framework versions

• Transformers 4.39.3
• Pytorch 2.2.1
• Datasets 2.18.0
• Tokenizers 0.15.2