scott156/LongT5XLNSPCCV1

LongT5-XLarge-NSPCC

This model is a fine-tuned version of google/long-t5-tglobal-xl on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6802
• Rouge1: 0.522
• Rouge2: 0.2398
• Rougel: 0.3131
• Rougelsum: 0.3134
• Gen Len: 310.4149

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: 0.0003
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
3.7613	0.9960	188	0.7287	0.4625	0.1767	0.2497	0.2491	375.9149
0.8593	1.9974	377	0.6978	0.4823	0.2081	0.2782	0.2781	364.7021
0.786	2.9987	566	0.6765	0.5097	0.2262	0.2997	0.3004	320.3298
0.7648	4.0	755	0.6783	0.5225	0.2382	0.3115	0.3113	315.383
0.7594	4.9801	940	0.6802	0.522	0.2398	0.3131	0.3134	310.4149

Framework versions

• PEFT 0.10.0
• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1