LEDv2_ACLsum_all_aspects

This model is a fine-tuned version of allenai/led-base-16384 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.3828
• Rouge1: 0.2845
• Rouge2: 0.0873
• Rougel: 0.2305
• Rougelsum: 0.2301
• Gen Len: 20.6767
• Bleu: 0.0452
• Precisions: 0.1101
• Brevity Penalty: 0.6455
• Length Ratio: 0.6956
• Translation Length: 4912.0
• Reference Length: 7062.0
• Precision: 0.8834
• Recall: 0.869
• F1: 0.8761
• Hashcode: roberta-large_L17_no-idf_version=0.3.12(hug_trans=4.52.4)

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len	Bleu	Precisions	Brevity Penalty	Length Ratio	Translation Length	Reference Length	Precision	Recall	F1	Hashcode
No log	1.0	19	5.4890	0.2347	0.0448	0.1818	0.1822	20.2933	0.0202	0.0737	0.6604	0.7067	4991.0	7062.0	0.8765	0.8606	0.8684	roberta-large_L17_no-idf_version=0.3.12(hug_trans=4.52.4)
No log	2.0	38	3.9349	0.2814	0.0821	0.2216	0.2214	20.53	0.0399	0.1062	0.6332	0.6863	4847.0	7062.0	0.8835	0.868	0.8756	roberta-large_L17_no-idf_version=0.3.12(hug_trans=4.52.4)
No log	3.0	57	3.5177	0.2824	0.0842	0.2239	0.2236	20.5667	0.0429	0.1081	0.6487	0.698	4929.0	7062.0	0.8836	0.8686	0.876	roberta-large_L17_no-idf_version=0.3.12(hug_trans=4.52.4)
No log	4.0	76	3.3828	0.2845	0.0873	0.2305	0.2301	20.6767	0.0452	0.1101	0.6455	0.6956	4912.0	7062.0	0.8834	0.869	0.8761	roberta-large_L17_no-idf_version=0.3.12(hug_trans=4.52.4)

Framework versions

• Transformers 4.52.4
• Pytorch 2.7.0+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1