b57323785e5127b9c935063d5688fde0

This model is a fine-tuned version of distilbert/distilbert-base-multilingual-cased on the dair-ai/emotion [split] dataset. It achieves the following results on the evaluation set:

• Loss: 0.2704
• Data Size: 1.0
• Epoch Runtime: 16.9117
• Accuracy: 0.9194
• F1 Macro: 0.8741

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro
No log	0	0	1.8121	0	1.1530	0.2177	0.1022
No log	1	500	1.6309	0.0078	1.3270	0.3317	0.1219
No log	2	1000	1.5782	0.0156	1.5345	0.3488	0.0862
No log	3	1500	1.5358	0.0312	2.0691	0.3579	0.0975
No log	4	2000	1.2502	0.0625	2.5937	0.5524	0.2308
0.0727	5	2500	0.7393	0.125	3.4916	0.7581	0.5244
0.4968	6	3000	0.3629	0.25	5.4156	0.8760	0.8189
0.0464	7	3500	0.2426	0.5	9.5759	0.8982	0.8473
0.1687	8.0	4000	0.2034	1.0	17.4262	0.9158	0.8636
0.1511	9.0	4500	0.1969	1.0	17.2475	0.9214	0.8715
0.1112	10.0	5000	0.2097	1.0	16.9975	0.9274	0.8865
0.1074	11.0	5500	0.2528	1.0	16.7826	0.9189	0.8708
0.1268	12.0	6000	0.2661	1.0	17.0161	0.9168	0.8730
0.0732	13.0	6500	0.2704	1.0	16.9117	0.9194	0.8741

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1