b5e705cfab9c79599cda41909cb16cf9

This model is a fine-tuned version of Qwen/Qwen2.5-0.5B on the dim/tldr_news dataset. It achieves the following results on the evaluation set:

• Loss: 4.4398
• Data Size: 1.0
• Epoch Runtime: 34.4935
• Accuracy: 0.7294
• F1 Macro: 0.7652
• Rouge1: 0.7301
• Rouge2: 0.0
• Rougel: 0.7301
• Rougelsum: 0.7294

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	24.9409	0	3.6137	0.2955	0.1584	0.2955	0.0	0.2955	0.2955
No log	1	178	26.7087	0.0078	4.4667	0.3430	0.2173	0.3430	0.0	0.3438	0.3430
No log	2	356	12.6671	0.0156	4.3271	0.6243	0.4747	0.6243	0.0	0.6243	0.6236
No log	3	534	5.0187	0.0312	5.3210	0.6428	0.5852	0.6428	0.0	0.6435	0.6428
No log	4	712	5.4231	0.0625	6.4204	0.6094	0.4646	0.6108	0.0	0.6108	0.6094
No log	5	890	2.8235	0.125	8.5103	0.7294	0.5792	0.7301	0.0	0.7301	0.7301
0.3804	6	1068	3.2914	0.25	12.1875	0.6903	0.7307	0.6896	0.0	0.6903	0.6903
2.2573	7	1246	3.2131	0.5	19.8656	0.7159	0.7164	0.7159	0.0	0.7159	0.7152
1.5679	8.0	1424	3.0587	1.0	36.1422	0.7308	0.7435	0.7308	0.0	0.7312	0.7308
0.9795	9.0	1602	4.4398	1.0	34.4935	0.7294	0.7652	0.7301	0.0	0.7301	0.7294

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.2.0
• Tokenizers 0.22.1