29cb06aeb56cbdecf76bdc4c3e179e78

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: 0.5375
• Data Size: 1.0
• Epoch Runtime: 946.5174
• Accuracy: 0.8565
• F1 Macro: 0.8562
• Rouge1: 0.8563
• Rouge2: 0.0
• Rougel: 0.8565
• Rougelsum: 0.8565

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	1.3291	0	6.0977	0.3167	0.2066	0.3170	0.0	0.3167	0.3168
1.0205	1	12271	0.6707	0.0078	13.6475	0.7407	0.7360	0.7408	0.0	0.7409	0.7405
0.4835	2	24542	0.4064	0.0156	20.9804	0.8509	0.8499	0.8508	0.0	0.8509	0.8510
0.4203	3	36813	0.3873	0.0312	35.9227	0.8565	0.8558	0.8566	0.0	0.8565	0.8566
0.4056	4	49084	0.3664	0.0625	65.0683	0.8628	0.8626	0.8626	0.0	0.8627	0.8627
0.3384	5	61355	0.3593	0.125	123.7432	0.8647	0.8641	0.8646	0.0	0.8647	0.8647
0.3534	6	73626	0.3995	0.25	241.1688	0.8552	0.8547	0.8550	0.0	0.8552	0.8554
0.3062	7	85897	0.3659	0.5	476.0455	0.8677	0.8674	0.8674	0.0	0.8675	0.8676
0.252	8.0	98168	0.3545	1.0	948.0369	0.8707	0.8705	0.8704	0.0	0.8705	0.8707
0.2087	9.0	110439	0.3708	1.0	946.9071	0.8664	0.8659	0.8663	0.0	0.8665	0.8664
0.1756	10.0	122710	0.4547	1.0	947.1643	0.8609	0.8604	0.8606	0.0	0.8606	0.8607
0.1215	11.0	134981	0.5158	1.0	945.5557	0.8592	0.8585	0.8589	0.0	0.8592	0.8591
0.1242	12.0	147252	0.5375	1.0	946.5174	0.8565	0.8562	0.8563	0.0	0.8565	0.8565

Framework versions

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