Regression_Albert

This model is a fine-tuned version of albert-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0459
• Mse: 0.0459
• Mae: 0.1675
• R2: 0.9763
• Accuracy: 1.0

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Mae	R2	Accuracy
No log	1.0	7	1.4379	1.4379	1.1107	-0.3492	0.0
No log	2.0	14	1.2159	1.2159	1.0476	-0.1409	0.1429
No log	3.0	21	1.7679	1.7679	1.1233	-0.6588	0.4286
No log	4.0	28	1.7069	1.7069	1.1072	-0.6015	0.1429
No log	5.0	35	1.4438	1.4438	0.9771	-0.3547	0.5714
No log	6.0	42	1.0275	1.0275	0.7910	0.0359	0.4286
No log	7.0	49	0.7649	0.7649	0.7080	0.2823	0.4286
No log	8.0	56	0.6584	0.6584	0.7083	0.3823	0.2857
No log	9.0	63	0.5064	0.5064	0.6108	0.5248	0.4286
No log	10.0	70	0.3638	0.3638	0.5078	0.6586	0.4286
No log	11.0	77	0.2660	0.2660	0.4352	0.7504	0.5714
No log	12.0	84	0.1570	0.1570	0.3323	0.8527	0.7143
No log	13.0	91	0.1953	0.1953	0.3863	0.8168	0.4286
No log	14.0	98	0.2230	0.2230	0.4033	0.7908	0.7143
No log	15.0	105	0.0578	0.0578	0.1935	0.9458	1.0
No log	16.0	112	0.0504	0.0504	0.1701	0.9527	1.0
No log	17.0	119	0.0466	0.0466	0.1713	0.9563	1.0
No log	18.0	126	0.0173	0.0173	0.1148	0.9837	1.0
No log	19.0	133	0.0417	0.0417	0.1811	0.9609	1.0
No log	20.0	140	0.0899	0.0899	0.1895	0.9156	0.8571
No log	21.0	147	0.0571	0.0571	0.1599	0.9465	0.8571
No log	22.0	154	0.0247	0.0247	0.1478	0.9768	1.0
No log	23.0	161	0.0201	0.0201	0.1261	0.9812	1.0
No log	24.0	168	0.0178	0.0178	0.1262	0.9833	1.0
No log	25.0	175	0.0172	0.0172	0.1208	0.9838	1.0

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.9.0
• Tokenizers 0.13.2