cb8517dba1966f6f5aae589719bbd08d

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

• Loss: 0.3577
• Data Size: 1.0
• Epoch Runtime: 1522.7065
• Accuracy: 0.9057
• F1 Macro: 0.8986
• Rouge1: 0.9056
• Rouge2: 0.0
• Rougel: 0.9057
• Rougelsum: 0.9057

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	6.7363	0	59.2099	0.5282	0.5244	0.5282	0.0	0.5282	0.5283
0.9584	1	11370	0.5380	0.0078	70.8794	0.7569	0.7139	0.7568	0.0	0.7569	0.7567
0.4827	2	22740	0.4538	0.0156	82.9487	0.7858	0.7504	0.7857	0.0	0.7857	0.7857
0.4205	3	34110	0.3615	0.0312	105.1014	0.8339	0.8240	0.8340	0.0	0.8339	0.8339
0.3302	4	45480	0.3291	0.0625	151.8743	0.8574	0.8466	0.8573	0.0	0.8574	0.8574
0.3159	5	56850	0.3142	0.125	242.5735	0.8593	0.8544	0.8594	0.0	0.8592	0.8594
0.2745	6	68220	0.2870	0.25	432.3027	0.8726	0.8672	0.8727	0.0	0.8726	0.8726
0.2465	7	79590	0.2571	0.5	787.0146	0.8904	0.8851	0.8905	0.0	0.8905	0.8903
0.224	8.0	90960	0.2465	1.0	1523.3613	0.8988	0.8925	0.8988	0.0	0.8988	0.8988
0.1495	9.0	102330	0.2651	1.0	1547.1681	0.9017	0.8964	0.9018	0.0	0.9018	0.9018
0.089	10.0	113700	0.3347	1.0	1537.6047	0.9027	0.8971	0.9028	0.0	0.9027	0.9027
0.0829	11.0	125070	0.3452	1.0	1528.9348	0.9057	0.8998	0.9058	0.0	0.9058	0.9057
0.0565	12.0	136440	0.3577	1.0	1522.7065	0.9057	0.8986	0.9056	0.0	0.9057	0.9057

Framework versions

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