sst-t5-base

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

• Loss: 0.0185
• Mse: 0.0185

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: 32
• 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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Mse
No log	1.0	267	0.0196	0.0196
0.0237	2.0	534	0.0179	0.0179
0.0237	3.0	801	0.0174	0.0174
0.0133	4.0	1068	0.0182	0.0182
0.0133	5.0	1335	0.0181	0.0181
0.0101	6.0	1602	0.0180	0.0180
0.0101	7.0	1869	0.0183	0.0183
0.0083	8.0	2136	0.0188	0.0188
0.0083	9.0	2403	0.0185	0.0186
0.0067	10.0	2670	0.0187	0.0187
0.0067	11.0	2937	0.0184	0.0184
0.0057	12.0	3204	0.0186	0.0186
0.0057	13.0	3471	0.0194	0.0194
0.005	14.0	3738	0.0175	0.0176
0.0045	15.0	4005	0.0182	0.0182
0.0045	16.0	4272	0.0183	0.0183
0.0041	17.0	4539	0.0187	0.0187
0.0041	18.0	4806	0.0186	0.0186
0.0038	19.0	5073	0.0188	0.0188
0.0038	20.0	5340	0.0185	0.0185

Framework versions

• Transformers 4.37.0
• Pytorch 1.13.1+cu117
• Datasets 2.15.0
• Tokenizers 0.15.2