mt5-small-task3-dataset1

This model is a fine-tuned version of google/mt5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3494
• Accuracy: 0.156
• Mse: 1.4726
• Log-distance: 0.6559
• S Score: 0.5092

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: 5.6e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Mse	Log-distance	S Score
12.3227	1.0	250	3.0911	0.126	1.6773	0.5862	0.5608
3.0171	2.0	500	1.8496	0.126	1.6805	0.5886	0.5608
2.1379	3.0	750	1.4488	0.126	1.6773	0.5862	0.5608
1.7896	4.0	1000	1.4309	0.126	1.6773	0.5862	0.5608
1.6843	5.0	1250	1.3863	0.136	1.5477	0.5660	0.5764
1.6196	6.0	1500	1.3676	0.142	1.4865	0.6943	0.4700
1.5812	7.0	1750	1.3518	0.14	1.4748	0.6894	0.4728
1.5336	8.0	2000	1.3538	0.148	1.6125	0.7828	0.4220
1.5106	9.0	2250	1.3468	0.172	1.4330	0.6204	0.5484
1.486	10.0	2500	1.3519	0.16	1.4487	0.6414	0.5268
1.4524	11.0	2750	1.3465	0.156	1.3796	0.5703	0.5720
1.4614	12.0	3000	1.3494	0.162	1.4250	0.6270	0.5316
1.4525	13.0	3250	1.3589	0.146	1.4602	0.6592	0.5068
1.4379	14.0	3500	1.3505	0.154	1.4722	0.6524	0.5128
1.4397	15.0	3750	1.3494	0.156	1.4726	0.6559	0.5092

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.15.0
• Tokenizers 0.15.0