e05c0c2f29cbfc8d239fbd48041bb9f5

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

• Loss: 1.5896
• Data Size: 0.25
• Epoch Runtime: 27.2599
• Accuracy: 0.2908
• F1 Macro: 0.0751

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	1.7637	0	3.7052	0.1421	0.0503
No log	1	500	1.6117	0.0078	4.6221	0.3488	0.0862
No log	2	1000	1.5769	0.0156	5.6798	0.3488	0.0862
No log	3	1500	1.6113	0.0312	7.8976	0.3488	0.0862
No log	4	2000	1.5925	0.0625	11.2247	0.2908	0.0751
0.0862	5	2500	1.5912	0.125	16.6001	0.2908	0.0751
1.5992	6	3000	1.5896	0.25	27.2599	0.2908	0.0751

Framework versions

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