d6bc41ea79dad8132c0aafdfa1010c27

This model is a fine-tuned version of google-bert/bert-large-uncased-whole-word-masking on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 1.3777
• Data Size: 1.0
• Epoch Runtime: 20.4556
• Accuracy: 0.7727
• F1 Macro: 0.7952
• Rouge1: 0.7734
• Rouge2: 0.0
• Rougel: 0.7727
• Rougelsum: 0.7727

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	1.9262	0	1.6519	0.0625	0.0670	0.0618	0.0	0.0625	0.0625
No log	1	178	1.5328	0.0078	2.7121	0.4446	0.2522	0.4453	0.0	0.4446	0.4446
No log	2	356	1.0822	0.0156	2.4237	0.6030	0.4175	0.6044	0.0	0.6037	0.6033
No log	3	534	0.9450	0.0312	3.0995	0.5845	0.4152	0.5845	0.0	0.5838	0.5852
No log	4	712	0.7912	0.0625	4.3506	0.7287	0.5708	0.7301	0.0	0.7294	0.7287
No log	5	890	0.7434	0.125	5.9598	0.7202	0.5549	0.7216	0.0	0.7209	0.7202
0.0536	6	1068	0.7149	0.25	8.5057	0.7259	0.5830	0.7266	0.0	0.7266	0.7266
0.5684	7	1246	0.6760	0.5	12.4274	0.7557	0.7678	0.7564	0.0	0.7557	0.7557
0.45	8.0	1424	0.6138	1.0	21.3084	0.7649	0.7896	0.7649	0.0	0.7649	0.7656
0.2713	9.0	1602	0.7694	1.0	20.7010	0.7727	0.8060	0.7727	0.0	0.7727	0.7734
0.1628	10.0	1780	1.0655	1.0	20.5736	0.7415	0.7702	0.7422	0.0	0.7422	0.7422
0.1446	11.0	1958	0.9690	1.0	20.7686	0.7585	0.7943	0.7592	0.0	0.7592	0.7585
0.0998	12.0	2136	1.3777	1.0	20.4556	0.7727	0.7952	0.7734	0.0	0.7727	0.7727

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1