0e1b91cfd013eb6d337e9b95b21541b0

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

• Loss: 0.4087
• Data Size: 1.0
• Epoch Runtime: 3095.7940
• Accuracy: 0.8933
• F1 Macro: 0.8859
• Rouge1: 0.8933
• Rouge2: 0.0
• Rougel: 0.8934
• Rougelsum: 0.8933

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	0.6996	0	109.6035	0.5399	0.5010	0.5397	0.0	0.5397	0.5396
0.5589	1	11370	0.4391	0.0078	134.3139	0.7921	0.7655	0.7919	0.0	0.7922	0.7921
0.4134	2	22740	0.3949	0.0156	161.1755	0.8239	0.8043	0.8240	0.0	0.8238	0.8238
0.3609	3	34110	0.3829	0.0312	210.3968	0.8357	0.8196	0.8357	0.0	0.8358	0.8357
0.3416	4	45480	0.3396	0.0625	302.3634	0.8532	0.8393	0.8531	0.0	0.8531	0.8531
0.3177	5	56850	0.3101	0.125	486.8587	0.8638	0.8569	0.8638	0.0	0.8640	0.8638
0.2782	6	68220	0.2913	0.25	862.5849	0.8708	0.8634	0.8708	0.0	0.8709	0.8709
0.2615	7	79590	0.2772	0.5	1606.8325	0.8804	0.8729	0.8803	0.0	0.8804	0.8804
0.259	8.0	90960	0.2807	1.0	3085.8652	0.8881	0.8813	0.8882	0.0	0.8882	0.8881
0.1805	9.0	102330	0.2625	1.0	3127.3159	0.8937	0.8874	0.8936	0.0	0.8937	0.8936
0.1258	10.0	113700	0.3247	1.0	3102.6746	0.8966	0.8892	0.8966	0.0	0.8967	0.8966
0.1046	11.0	125070	0.3671	1.0	3128.7897	0.8946	0.8870	0.8946	0.0	0.8946	0.8945
0.0964	12.0	136440	0.3417	1.0	3113.1528	0.8934	0.8872	0.8934	0.0	0.8935	0.8934
0.1002	13.0	147810	0.4087	1.0	3095.7940	0.8933	0.8859	0.8933	0.0	0.8934	0.8933

Framework versions

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