Regression_albert_NOaug_MSEloss

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: 0.4715
• Mse: 0.4715
• Mae: 0.6001
• R2: 0.1320
• Accuracy: 0.4737

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: 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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Mae	R2	Accuracy
No log	1.0	33	0.2966	0.2966	0.4630	0.1139	0.7568
No log	2.0	66	0.2679	0.2679	0.4039	0.1995	0.7568
No log	3.0	99	0.4088	0.4088	0.5125	-0.2213	0.5405
No log	4.0	132	0.4331	0.4331	0.5399	-0.2939	0.4865
No log	5.0	165	0.3699	0.3699	0.4317	-0.1053	0.6757
No log	6.0	198	0.3456	0.3456	0.4117	-0.0325	0.6216
No log	7.0	231	0.3371	0.3371	0.4155	-0.0072	0.6757
No log	8.0	264	0.3261	0.3261	0.3811	0.0256	0.7297
No log	9.0	297	0.2312	0.2312	0.2705	0.3092	0.8108
No log	10.0	330	0.3194	0.3194	0.3681	0.0457	0.6757
No log	11.0	363	0.3638	0.3638	0.4124	-0.0870	0.6757
No log	12.0	396	0.3101	0.3101	0.3630	0.0734	0.7027
No log	13.0	429	0.2762	0.2762	0.3221	0.1748	0.7568
No log	14.0	462	0.2970	0.2970	0.3376	0.1126	0.7297
No log	15.0	495	0.3185	0.3185	0.3532	0.0483	0.7297

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3