92f6fbfcc4260255a370e19a1904e6a7

This model is a fine-tuned version of albert/albert-xxlarge-v1 on the nyu-mll/glue dataset. It achieves the following results on the evaluation set:

• Loss: 0.3309
• Data Size: 1.0
• Epoch Runtime: 808.7512
• Accuracy: 0.9066
• F1 Macro: 0.8997
• Rouge1: 0.9066
• Rouge2: 0.0
• Rougel: 0.9068
• Rougelsum: 0.9067

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	0.7730	0	21.2963	0.3684	0.2704	0.3686	0.0	0.3684	0.3687
0.5466	1	11370	0.4057	0.0078	27.8213	0.8223	0.8006	0.8223	0.0	0.8224	0.8223
0.3782	2	22740	0.3309	0.0156	34.0036	0.8573	0.8447	0.8573	0.0	0.8572	0.8572
0.3355	3	34110	0.3291	0.0312	46.2379	0.8632	0.8546	0.8633	0.0	0.8633	0.8632
0.2846	4	45480	0.3066	0.0625	70.5556	0.8665	0.8526	0.8664	0.0	0.8664	0.8665
0.2721	5	56850	0.2758	0.125	119.5706	0.8792	0.8684	0.8792	0.0	0.8792	0.8792
0.2594	6	68220	0.2835	0.25	217.6568	0.8805	0.8741	0.8805	0.0	0.8805	0.8804
0.224	7	79590	0.2551	0.5	414.8921	0.8911	0.8864	0.8911	0.0	0.8913	0.8912
0.2154	8.0	90960	0.2373	1.0	805.9334	0.9072	0.9010	0.9072	0.0	0.9073	0.9072
0.1597	9.0	102330	0.2453	1.0	809.3369	0.9098	0.9042	0.9098	0.0	0.9098	0.9098
0.1087	10.0	113700	0.2975	1.0	817.5793	0.9103	0.9039	0.9103	0.0	0.9104	0.9103
0.0957	11.0	125070	0.3163	1.0	806.9757	0.9088	0.9027	0.9088	0.0	0.9090	0.9088
0.0979	12.0	136440	0.3309	1.0	808.7512	0.9066	0.8997	0.9066	0.0	0.9068	0.9067

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1