4503e95fd680c15e4c326aeb4de34f5b

This model is a fine-tuned version of openai-community/gpt2 on the contemmcm/trec dataset. It achieves the following results on the evaluation set:

• Loss: 0.2228
• Data Size: 1.0
• Epoch Runtime: 13.3883
• Accuracy: 0.9708
• F1 Macro: 0.9733

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	8.5490	0	1.1257	0.1646	0.0471
No log	1	170	3.4189	0.0078	1.4279	0.1583	0.0632
No log	2	340	2.3919	0.0156	1.3600	0.1229	0.0747
No log	3	510	1.8343	0.0312	1.6752	0.3042	0.1506
No log	4	680	1.4796	0.0625	2.0758	0.4125	0.2799
0.1228	5	850	0.8520	0.125	3.0188	0.7167	0.5934
0.1228	6	1020	0.3624	0.25	4.5066	0.8896	0.7495
0.3583	7	1190	0.2102	0.5	7.4094	0.9458	0.9214
0.2065	8.0	1360	0.1629	1.0	13.3485	0.9646	0.9389
0.1459	9.0	1530	0.1273	1.0	13.4659	0.9688	0.9643
0.0722	10.0	1700	0.1999	1.0	13.4825	0.9729	0.9761
0.0481	11.0	1870	0.1570	1.0	13.4167	0.975	0.9780
0.0274	12.0	2040	0.1936	1.0	13.3379	0.9667	0.9557
0.019	13.0	2210	0.2228	1.0	13.3883	0.9708	0.9733

Framework versions

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