357d350da953a816816f5d421631ebca

This model is a fine-tuned version of distilbert/distilbert-base-uncased-distilled-squad on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 1.2177
• Data Size: 1.0
• Epoch Runtime: 6.5230
• Accuracy: 0.7457
• F1 Macro: 0.7818
• Rouge1: 0.7457
• Rouge2: 0.0
• Rougel: 0.7450
• Rougelsum: 0.7450

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	1.6564	0	0.9976	0.0185	0.0082	0.0185	0.0	0.0185	0.0185
No log	1	178	1.5166	0.0078	2.3233	0.2429	0.0819	0.2429	0.0	0.2436	0.2429
No log	2	356	1.4090	0.0156	1.2717	0.4446	0.2390	0.4453	0.0	0.4453	0.4439
No log	3	534	1.1799	0.0312	1.7685	0.6477	0.4803	0.6484	0.0	0.6484	0.6477
No log	4	712	0.9568	0.0625	1.9427	0.6612	0.4935	0.6619	0.0	0.6619	0.6612
No log	5	890	0.7724	0.125	2.2570	0.6911	0.5139	0.6925	0.0	0.6911	0.6911
0.0616	6	1068	0.6960	0.25	2.5959	0.7266	0.6262	0.7273	0.0	0.7269	0.7273
0.562	7	1246	0.6171	0.5	3.9627	0.7585	0.7734	0.7592	0.0	0.7592	0.7578
0.4835	8.0	1424	0.6270	1.0	6.2698	0.7578	0.7858	0.7571	0.0	0.7585	0.7578
0.3017	9.0	1602	0.7640	1.0	6.1325	0.7493	0.7881	0.75	0.0	0.75	0.7486
0.2016	10.0	1780	0.9923	1.0	6.2195	0.7294	0.7717	0.7301	0.0	0.7298	0.7294
0.0952	11.0	1958	1.2177	1.0	6.5230	0.7457	0.7818	0.7457	0.0	0.7450	0.7450

Framework versions

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