17ad61827dbb3a4ae31215749f132950

This model is a fine-tuned version of openai-community/gpt2-xl on the dair-ai/emotion [split] dataset. It achieves the following results on the evaluation set:

• Loss: 0.3075
• Data Size: 1.0
• Epoch Runtime: 141.3662
• Accuracy: 0.9264
• F1 Macro: 0.8878

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	1.9193	0	6.4005	0.1406	0.1149
No log	1	500	1.5733	0.0078	7.6991	0.3790	0.1722
No log	2	1000	1.2956	0.0156	11.5355	0.5328	0.2940
No log	3	1500	0.7095	0.0312	16.2980	0.7681	0.6782
No log	4	2000	0.4663	0.0625	22.2824	0.8579	0.8047
0.0332	5	2500	0.3607	0.125	30.5097	0.9042	0.8625
0.2108	6	3000	0.2393	0.25	46.8248	0.9098	0.8324
0.0282	7	3500	0.1504	0.5	78.4666	0.9335	0.8956
0.1485	8.0	4000	0.1564	1.0	145.3671	0.9274	0.8768
0.126	9.0	4500	0.1342	1.0	142.7476	0.9274	0.8711
0.0692	10.0	5000	0.2308	1.0	140.2724	0.9274	0.8922
0.0425	11.0	5500	0.2832	1.0	143.7939	0.9219	0.8804
0.0464	12.0	6000	0.2428	1.0	143.7617	0.9284	0.8814
0.0425	13.0	6500	0.3075	1.0	141.3662	0.9264	0.8878

Framework versions

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