jialicheng/unlearn_samsum_t5-small_neggrad_4_42

samsum_42

This model is a fine-tuned version of google/t5-v1_1-small on the samsum dataset. It achieves the following results on the evaluation set:

• Loss: 2.1147
• Rouge1: 39.0474
• Rouge2: 16.9299
• Rougel: 32.8695
• Rougelsum: 36.1983
• Gen Len: 24.7604
• Test Rougel: 32.8695
• Df Rougel: 32.4563
• Unlearn Overall Rougel: 0.7066
• Unlearn Time: 254.2657

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: 1e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum	Gen Len	Overall Rougel	Unlearn Overall Rougel	Time
No log	1.0	19	2.1121	38.8443	16.9802	32.5031	36.0402	20.8081	0.6641	0.6641	-1
No log	2.0	38	2.1128	38.9733	16.9624	32.6111	36.146	22.4988	0.5832	0.5832	-1
No log	3.0	57	2.1147	39.0474	16.9299	32.4563	36.1983	24.7604	0.7066	0.7066	-1
No log	4.0	76	2.1195	38.7283	16.6506	32.1497	35.8362	27.3496	0.6566	0.6566	-1
No log	5.0	95	2.1226	38.3644	16.4887	31.8308	35.5077	29.0428	0.6913	0.6913	-1

Framework versions

• Transformers 4.39.3
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.15.2