bart-large-cnn-samsum-rescom-finetuned-resume-summarizer-9-epoch-tweak

This model is a fine-tuned version of Ameer05/model-token-repo on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.4511
• Rouge1: 59.76
• Rouge2: 52.1999
• Rougel: 57.3631
• Rougelsum: 59.3075

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 9
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
No log	0.91	5	2.0185	52.2186	45.4675	49.3152	51.9415
No log	1.91	10	1.6571	60.7728	52.8611	57.3487	60.1676
No log	2.91	15	1.5323	60.5674	52.2246	57.9846	60.073
No log	3.91	20	1.4556	61.2167	53.5087	58.9609	60.893
1.566	4.91	25	1.4632	62.918	55.4544	60.7116	62.6614
1.566	5.91	30	1.4360	60.4173	52.5859	57.8131	59.8864
1.566	6.91	35	1.4361	61.4273	53.9663	59.4445	60.9672
1.566	7.91	40	1.4477	60.3401	52.7276	57.7504	59.8209
0.6928	8.91	45	1.4511	59.76	52.1999	57.3631	59.3075

Framework versions

• Transformers 4.15.0
• Pytorch 1.9.1
• Datasets 1.18.4
• Tokenizers 0.10.3