9eed42cdea42f806fc6d9968b1c55312

This model is a fine-tuned version of albert/albert-xlarge-v1 on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 1.4705
• Data Size: 1.0
• Epoch Runtime: 17.5951
• Accuracy: 0.2486
• F1 Macro: 0.0796
• Rouge1: 0.2486
• Rouge2: 0.0
• Rougel: 0.2486
• Rougelsum: 0.2479

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	1.6535	0	1.6755	0.2365	0.1453	0.2365	0.0	0.2358	0.2372
No log	1	178	1.5140	0.0078	3.1886	0.3168	0.1728	0.3168	0.0	0.3168	0.3168
No log	2	356	1.4675	0.0156	2.0007	0.3842	0.2385	0.3842	0.0	0.3849	0.3849
No log	3	534	1.1224	0.0312	2.3157	0.4517	0.3452	0.4510	0.0	0.4524	0.4510
No log	4	712	0.9170	0.0625	2.9576	0.6626	0.5110	0.6634	0.0	0.6630	0.6630
No log	5	890	0.7540	0.125	3.8821	0.7017	0.5872	0.7031	0.0	0.7024	0.7024
0.0623	6	1068	0.7350	0.25	5.8063	0.7081	0.5722	0.7088	0.0	0.7081	0.7088
1.3378	7	1246	1.4583	0.5	9.5836	0.2486	0.0796	0.2486	0.0	0.2486	0.2479
1.4399	8.0	1424	1.4651	1.0	17.6396	0.2756	0.0864	0.2752	0.0	0.2749	0.2756
1.4469	9.0	1602	1.4534	1.0	17.5320	0.2756	0.0864	0.2752	0.0	0.2749	0.2756
1.4332	10.0	1780	1.4705	1.0	17.5951	0.2486	0.0796	0.2486	0.0	0.2486	0.2479

Framework versions

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