39c56382a1dbd7f04d8a094ceafe2fd5

This model is a fine-tuned version of FacebookAI/roberta-base on the nyu-mll/glue [qqp] dataset. It achieves the following results on the evaluation set:

• Loss: 0.3652
• Data Size: 1.0
• Epoch Runtime: 832.4869
• Accuracy: 0.8682
• F1 Macro: 0.8612
• Rouge1: 0.8682
• Rouge2: 0.0
• Rougel: 0.8682
• Rougelsum: 0.8682

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	0.6681	0	29.1760	0.6320	0.3872	0.6318	0.0	0.6319	0.6317
0.5973	1	11370	0.4370	0.0078	35.5348	0.7962	0.7842	0.7962	0.0	0.7963	0.7962
0.4472	2	22740	0.3972	0.0156	41.8248	0.8220	0.8086	0.8220	0.0	0.8221	0.8219
0.4068	3	34110	0.3755	0.0312	53.8164	0.8344	0.8172	0.8344	0.0	0.8345	0.8344
0.3627	4	45480	0.3833	0.0625	77.8531	0.8436	0.8278	0.8435	0.0	0.8435	0.8436
0.3309	5	56850	0.3324	0.125	128.4601	0.8548	0.8398	0.8547	0.0	0.8548	0.8548
0.3204	6	68220	0.3055	0.25	227.7489	0.8698	0.8624	0.8698	0.0	0.8698	0.8698
0.3086	7	79590	0.3167	0.5	434.9633	0.8594	0.8543	0.8595	0.0	0.8595	0.8595
0.3348	8.0	90960	0.3484	1.0	839.5986	0.8614	0.8554	0.8614	0.0	0.8615	0.8616
0.3537	9.0	102330	0.3340	1.0	830.1252	0.8582	0.8470	0.8582	0.0	0.8583	0.8583
0.2946	10.0	113700	0.3652	1.0	832.4869	0.8682	0.8612	0.8682	0.0	0.8682	0.8682

Framework versions

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