flan-t5-large-reaction-extraction

This model is a fine-tuned version of google/flan-t5-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8736
• Rouge1: 34.4249
• Rouge2: 21.4943
• Rougel: 33.4902
• Rougelsum: 33.466
• Gen Len: 28.1622

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 4
• 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: linear
• num_epochs: 5.0

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len
1.0236	1.0	2885	0.9221	24.983	14.6971	24.5305	24.5497	42.0762
0.8177	2.0	5771	0.8503	30.0616	18.1568	29.5486	29.5537	35.5198
0.6784	3.0	8657	0.8324	32.537	20.2178	31.7925	31.75	27.7862
0.5961	4.0	11543	0.8330	32.9769	20.8184	32.2179	32.2036	33.6372
0.4985	5.0	14425	0.8736	34.4249	21.4943	33.4902	33.466	28.1622

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu117
• Datasets 2.6.1
• Tokenizers 0.13.3