bart-base-finetuned-steel-news-Environment

This model is a fine-tuned version of facebook/bart-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 5.0878
• Rouge1: 33.7492
• Rouge2: 16.9181
• Rougel: 31.1710
• Rougelsum: 31.1290
• Rouge3: 5.2115

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: 3e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• 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
• lr_scheduler_warmup_steps: 500
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Rouge3
No log	1.0	23	13.6118	23.8890	8.6893	19.6100	19.7700	2.9881
No log	2.0	46	11.2290	23.8890	8.6893	19.6100	19.7700	2.9881
No log	3.0	69	10.0166	23.7559	8.2513	19.3871	19.4579	2.4654
No log	4.0	92	8.0345	27.0003	9.6629	24.1700	24.1187	2.0520
11.4324	5.0	115	4.9696	32.3997	12.4816	29.5265	29.4606	3.2122

Framework versions

• Transformers 4.52.2
• Pytorch 2.6.0+cu124
• Datasets 2.14.4
• Tokenizers 0.21.1