db480d464c15fbbb3e78d1d2771bf203

This model is a fine-tuned version of studio-ousia/luke-large-lite on the dair-ai/emotion [split] dataset. It achieves the following results on the evaluation set:

• Loss: 1.5614
• Data Size: 1.0
• Epoch Runtime: 94.0786
• Accuracy: 0.3488
• F1 Macro: 0.0862

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	1.7826	0	3.7784	0.0791	0.0244
No log	1	500	1.6039	0.0078	4.6853	0.2908	0.0751
No log	2	1000	1.6056	0.0156	5.9231	0.3488	0.0862
No log	3	1500	1.5582	0.0312	7.7229	0.3488	0.0862
No log	4	2000	1.7412	0.0625	11.0322	0.4390	0.2054
0.081	5	2500	1.1406	0.125	17.2653	0.5927	0.2447
1.195	6	3000	0.9361	0.25	28.7498	0.6774	0.3673
0.1097	7	3500	0.5756	0.5	52.1193	0.8347	0.5822
0.5342	8.0	4000	0.5511	1.0	95.2725	0.8503	0.7056
0.5663	9.0	4500	0.5202	1.0	93.6776	0.8503	0.7053
0.6283	10.0	5000	0.5979	1.0	94.5637	0.8362	0.6896
1.5501	11.0	5500	1.5435	1.0	93.8162	0.2984	0.0995
1.5722	12.0	6000	1.5611	1.0	93.2845	0.3488	0.0862
1.5744	13.0	6500	1.5614	1.0	94.0786	0.3488	0.0862

Framework versions

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