protBERTbfd_AAV2_regressor

This model is a fine-tuned version of Rostlab/prot_bert_bfd on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0327
• Mse: 0.0327
• Rmse: 0.1808
• Mae: 0.0618
• R2: 0.8691
• Smape: 101.2324

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 64
• total_train_batch_size: 4096
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 6

Training results

Training Loss	Epoch	Step	Validation Loss	Mse	Rmse	Mae	R2	Smape
No log	1.0	58	0.0985	0.0985	0.3138	0.1707	0.6057	102.5806
No log	2.0	116	0.0689	0.0689	0.2625	0.1432	0.7242	112.9846
No log	3.0	174	0.0400	0.0400	0.1999	0.0859	0.8399	102.6132
No log	4.0	232	0.0402	0.0402	0.2005	0.0745	0.8389	103.3228
No log	5.0	290	0.0337	0.0337	0.1836	0.0665	0.8650	101.0925
No log	6.0	348	0.0327	0.0327	0.1808	0.0618	0.8691	101.2324

Framework versions

• Transformers 4.21.0
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1