62570c130d9c027e2ca47babd21cadfa

This model is a fine-tuned version of openai-community/gpt2-medium on the nyu-mll/glue [qnli] dataset. It achieves the following results on the evaluation set:

• Loss: 0.4633
• Data Size: 1.0
• Epoch Runtime: 429.8114
• Accuracy: 0.8899
• F1 Macro: 0.8899
• Rouge1: 0.8899
• Rouge2: 0.0
• Rougel: 0.8901
• Rougelsum: 0.8901

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	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	11.8969	0	8.5703	0.5066	0.3431	0.5062	0.0	0.5064	0.5064
No log	1	3273	0.7073	0.0078	13.2616	0.5860	0.5237	0.5858	0.0	0.5855	0.5858
0.022	2	6546	0.5080	0.0156	15.7701	0.7616	0.7535	0.7616	0.0	0.7614	0.7612
0.5882	3	9819	0.4724	0.0312	22.9579	0.8057	0.8042	0.8053	0.0	0.8059	0.8057
0.4573	4	13092	0.4225	0.0625	35.7555	0.8156	0.8140	0.8154	0.0	0.8158	0.8154
0.3789	5	16365	0.3449	0.125	62.8776	0.8548	0.8546	0.8548	0.0	0.8550	0.8544
0.3552	6	19638	0.3032	0.25	116.2979	0.8796	0.8795	0.8794	0.0	0.8798	0.8796
0.2876	7	22911	0.3125	0.5	221.1460	0.8651	0.8644	0.8649	0.0	0.8653	0.8649
0.2608	8.0	26184	0.2901	1.0	432.1524	0.8820	0.8817	0.8822	0.0	0.8818	0.8822
0.1293	9.0	29457	0.3893	1.0	435.3946	0.8844	0.8842	0.8842	0.0	0.8847	0.8846
0.1096	10.0	32730	0.3718	1.0	434.1044	0.8919	0.8918	0.8919	0.0	0.8923	0.8919
0.0711	11.0	36003	0.4568	1.0	430.7896	0.8965	0.8965	0.8967	0.0	0.8967	0.8963
0.0902	12.0	39276	0.4633	1.0	429.8114	0.8899	0.8899	0.8899	0.0	0.8901	0.8901

Framework versions

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