bbabc55a8bea9f4bcc8e2e5a45fe328b

This model is a fine-tuned version of albert/albert-xxlarge-v2 on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 1.1757
• Data Size: 1.0
• Epoch Runtime: 34.5248
• Accuracy: 0.7464
• F1 Macro: 0.7879
• Rouge1: 0.7464
• Rouge2: 0.0
• Rougel: 0.7464
• Rougelsum: 0.7464

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	1.9144	0	2.9832	0.2727	0.1296	0.2720	0.0	0.2720	0.2727
No log	1	178	1.6159	0.0078	3.3691	0.3097	0.1621	0.3104	0.0	0.3097	0.3089
No log	2	356	1.5238	0.0156	3.7020	0.3388	0.1912	0.3388	0.0	0.3388	0.3388
No log	3	534	1.6766	0.0312	4.3784	0.3040	0.1833	0.3047	0.0	0.3040	0.3033
No log	4	712	1.2156	0.0625	5.4700	0.5050	0.3304	0.5050	0.0	0.5057	0.5050
No log	5	890	1.2470	0.125	7.2786	0.3956	0.2404	0.3963	0.0	0.3963	0.3963
0.0829	6	1068	1.1592	0.25	11.2324	0.5007	0.3454	0.5014	0.0	0.5007	0.5004
1.0181	7	1246	0.9354	0.5	18.9063	0.6349	0.4529	0.6364	0.0	0.6357	0.6349
0.752	8.0	1424	0.7258	1.0	34.9917	0.7244	0.5663	0.7251	0.0	0.7251	0.7244
0.6424	9.0	1602	0.6150	1.0	34.7189	0.7592	0.7783	0.7599	0.0	0.7599	0.7592
0.5411	10.0	1780	0.7055	1.0	34.5493	0.7472	0.7854	0.7472	0.0	0.7472	0.7472
0.4524	11.0	1958	0.5633	1.0	34.5720	0.7713	0.8076	0.7720	0.0	0.7713	0.7713
0.2773	12.0	2136	0.8556	1.0	34.6120	0.7450	0.7811	0.7450	0.0	0.7457	0.7457
0.1953	13.0	2314	0.8655	1.0	34.4858	0.75	0.7919	0.7504	0.0	0.7504	0.75
0.1454	14.0	2492	1.0329	1.0	34.5014	0.7528	0.7931	0.7528	0.0	0.7528	0.7528
0.0921	15.0	2670	1.1757	1.0	34.5248	0.7464	0.7879	0.7464	0.0	0.7464	0.7464

Framework versions

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