esm2_t12_35M_UR50D-finetuned-rep7868aav2-v0

This model is a fine-tuned version of facebook/esm2_t12_35M_UR50D on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0513
• Spearmanr: 0.7389

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: 8
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Spearmanr
0.118	1.0	1180	0.1154	0.3185
0.1156	2.0	2360	0.1109	0.3383
0.1143	3.0	3540	0.1162	0.3194
0.1192	4.0	4720	0.1111	0.2974
0.1147	5.0	5900	0.1125	0.4043
0.1196	6.0	7080	0.1116	0.1580
0.1171	7.0	8260	0.1114	0.2923
0.1177	8.0	9440	0.1106	0.3592
0.1126	9.0	10620	0.1105	0.3724
0.1152	10.0	11800	0.1135	0.4947
0.1159	11.0	12980	0.1082	0.5113
0.0953	12.0	14160	0.0820	0.6096
0.0798	13.0	15340	0.0688	0.6442
0.074	14.0	16520	0.0710	0.6738
0.0704	15.0	17700	0.0816	0.6736
0.0678	16.0	18880	0.0596	0.7142
0.0599	17.0	20060	0.0689	0.7187
0.0568	18.0	21240	0.0566	0.7308
0.0534	19.0	22420	0.0518	0.7340
0.0522	20.0	23600	0.0513	0.7389

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1