jialicheng/unlearn_samsum_t5-base_neggrad_8_42

samsum_42

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

• Loss: 1.8469
• Rouge1: 39.6521
• Rouge2: 20.5195
• Rougel: 35.2303
• Rougelsum: 36.9399
• Gen Len: 11.4548
• Test Rougel: 35.2303
• Df Rougel: 35.8104
• Unlearn Overall Rougel: 0.2099
• Unlearn Time: 502.7173

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: 8
• 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	37	1.5319	46.4888	23.3914	40.8182	43.0601	17.5134	-0.1262	-0.1262	-1
No log	2.0	74	1.8469	39.6521	20.5195	35.8104	36.9399	11.4548	0.2099	0.2099	-1
No log	3.0	111	2.4451	36.3205	17.8091	33.5109	33.8967	10.2910	0.0407	0.0407	-1
No log	4.0	148	3.2318	32.629	15.2243	31.5303	30.7888	9.5440	-0.3471	-0.3471	-1
No log	5.0	185	3.7401	31.519	14.3311	30.2925	29.8397	9.3056	-0.1495	-0.1495	-1

Framework versions

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