fb81a0f973c8e3b8f74608b52cb0ea37

This model is a fine-tuned version of albert/albert-base-v2 on the nyu-mll/glue dataset. It achieves the following results on the evaluation set:

• Loss: 0.4882
• Data Size: 1.0
• Epoch Runtime: 119.1311
• Accuracy: 0.8489
• F1 Macro: 0.8488
• Rouge1: 0.8487
• Rouge2: 0.0
• Rougel: 0.8489
• Rougelsum: 0.8487

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	0.6956	0	2.6638	0.5107	0.3566	0.5105	0.0	0.5107	0.5103
No log	1	3273	0.6928	0.0078	3.8673	0.4943	0.3308	0.4947	0.0	0.4943	0.4944
0.0112	2	6546	0.6616	0.0156	4.5946	0.6009	0.5240	0.6011	0.0	0.6006	0.6006
0.5675	3	9819	0.5216	0.0312	6.3566	0.7954	0.7946	0.7958	0.0	0.7952	0.7954
0.5119	4	13092	0.4519	0.0625	9.8708	0.8167	0.8167	0.8169	0.0	0.8168	0.8167
0.4421	5	16365	0.4387	0.125	18.9666	0.7971	0.7945	0.7972	0.0	0.7972	0.7974
0.467	6	19638	0.3684	0.25	34.1663	0.8415	0.8415	0.8414	0.0	0.8415	0.8414
0.3921	7	22911	0.3910	0.5	60.6173	0.8267	0.8253	0.8265	0.0	0.8267	0.8268
0.3501	8.0	26184	0.3467	1.0	127.3543	0.8574	0.8572	0.8573	0.0	0.8575	0.8574
0.2784	9.0	29457	0.3494	1.0	121.4642	0.8507	0.8503	0.8506	0.0	0.8511	0.8509
0.1784	10.0	32730	0.3531	1.0	123.8021	0.8540	0.8540	0.8540	0.0	0.8540	0.8540
0.1709	11.0	36003	0.4465	1.0	119.9891	0.8496	0.8495	0.8496	0.0	0.85	0.8494
0.1311	12.0	39276	0.4882	1.0	119.1311	0.8489	0.8488	0.8487	0.0	0.8489	0.8487

Framework versions

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