419d7b2aacaac42665f05deb6b171e0f

This model is a fine-tuned version of FacebookAI/roberta-large on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 0.7011
• Data Size: 1.0
• Epoch Runtime: 30.6839
• Accuracy: 0.7734
• F1 Macro: 0.8099
• Rouge1: 0.7741
• Rouge2: 0.0
• Rougel: 0.7741
• Rougelsum: 0.7741

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	1.8205	0	2.4714	0.0185	0.0073	0.0185	0.0	0.0185	0.0185
No log	1	178	1.5122	0.0078	2.9117	0.2763	0.1122	0.2763	0.0	0.2756	0.2756
No log	2	356	1.4152	0.0156	3.3986	0.2834	0.1341	0.2841	0.0	0.2841	0.2834
No log	3	534	0.8161	0.0312	4.4051	0.7031	0.5437	0.7038	0.0	0.7031	0.7031
No log	4	712	0.7809	0.0625	5.4730	0.7436	0.5846	0.7447	0.0	0.7436	0.7436
No log	5	890	0.7273	0.125	7.2879	0.7337	0.5730	0.7351	0.0	0.7344	0.7337
0.0524	6	1068	0.6326	0.25	10.5516	0.7557	0.6942	0.7564	0.0	0.7564	0.7557
0.5716	7	1246	0.6684	0.5	16.6750	0.7436	0.7150	0.7443	0.0	0.7443	0.7443
0.496	8.0	1424	0.6485	1.0	31.1960	0.7656	0.7786	0.7663	0.0	0.7656	0.7656
0.3959	9.0	1602	0.7267	1.0	29.6157	0.7621	0.7977	0.7621	0.0	0.7628	0.7621
0.2627	10.0	1780	0.7011	1.0	30.6839	0.7734	0.8099	0.7741	0.0	0.7741	0.7741

Framework versions

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