exist-2023-task3

This model is a fine-tuned version of microsoft/Multilingual-MiniLM-L12-H384 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3814
• Acc: 47.9045

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: 2e-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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Acc
No log	1.0	97	0.4243	28.9107
No log	2.0	194	0.4149	31.6012
No log	3.0	291	0.4112	32.5061
No log	4.0	388	0.4111	32.6858
No log	5.0	485	0.4049	34.8687
0.416	6.0	582	0.4023	35.9895
0.416	7.0	679	0.4005	36.6499
0.416	8.0	776	0.3978	38.6035
0.416	9.0	873	0.3964	38.5017
0.416	10.0	970	0.3931	40.5065
0.4029	11.0	1067	0.3912	42.2190
0.4029	12.0	1164	0.3891	43.2468
0.4029	13.0	1261	0.3888	42.6855
0.4029	14.0	1358	0.3861	44.5341
0.4029	15.0	1455	0.3851	44.8797
0.3932	16.0	1552	0.3841	46.3287
0.3932	17.0	1649	0.3832	46.5887
0.3932	18.0	1746	0.3820	47.4830
0.3932	19.0	1843	0.3817	47.8015
0.3932	20.0	1940	0.3814	47.9045

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.1+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3