ffb45f1766b20fedcd9cc555be6f1cb4

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

• Loss: 0.1887
• Data Size: 1.0
• Epoch Runtime: 37.8911
• Accuracy: 0.9299
• F1 Macro: 0.8752

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	1.8431	0	2.0977	0.0333	0.0107
No log	1	500	1.6775	0.0078	2.5831	0.2908	0.0751
No log	2	1000	1.5636	0.0156	2.8792	0.3488	0.0862
No log	3	1500	1.2361	0.0312	3.5151	0.5580	0.2286
No log	4	2000	0.7845	0.0625	4.7528	0.7787	0.6109
0.0551	5	2500	0.4919	0.125	7.2789	0.8382	0.7833
0.3807	6	3000	0.2524	0.25	11.7743	0.9138	0.8701
0.0411	7	3500	0.2618	0.5	21.2407	0.9158	0.8738
0.2232	8.0	4000	0.1663	1.0	39.5274	0.9269	0.8842
0.1578	9.0	4500	0.1847	1.0	39.7992	0.9259	0.8824
0.1543	10.0	5000	0.1882	1.0	37.6587	0.9325	0.8935
0.1249	11.0	5500	0.1787	1.0	37.2834	0.9315	0.8930
0.1313	12.0	6000	0.1887	1.0	37.8911	0.9299	0.8752

Framework versions

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