esm2_t12_35M-lora-binding-sites_2024-04-25_14-35-31

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: 0.3589
• Accuracy: 0.8457

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: 0.0005701568055793089
• train_batch_size: 64
• eval_batch_size: 64
• seed: 8893
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6703	1.0	24	0.6807	0.5820
0.6449	2.0	48	0.6703	0.5820
0.6659	3.0	72	0.6458	0.5977
0.6432	4.0	96	0.6612	0.6328
0.6322	5.0	120	0.6051	0.6523
0.6176	6.0	144	0.6062	0.6504
0.4904	7.0	168	0.5762	0.6777
0.4426	8.0	192	0.5784	0.6953
0.6014	9.0	216	0.5497	0.7148
0.4484	10.0	240	0.5399	0.7227
0.552	11.0	264	0.5142	0.7480
0.3581	12.0	288	0.4395	0.7930
0.3604	13.0	312	0.4201	0.8066
0.2733	14.0	336	0.4107	0.8262
0.2539	15.0	360	0.4373	0.8008
0.3538	16.0	384	0.3954	0.8301
0.4363	17.0	408	0.3852	0.8320
0.3433	18.0	432	0.3735	0.8418
0.2758	19.0	456	0.3685	0.8438
0.2073	20.0	480	0.3860	0.8262
0.3578	21.0	504	0.3689	0.8301
0.3114	22.0	528	0.3626	0.8418
0.3296	23.0	552	0.3621	0.8438
0.276	24.0	576	0.3602	0.8457
0.2583	25.0	600	0.3622	0.8457
0.1917	26.0	624	0.3597	0.8477
0.3588	27.0	648	0.3603	0.8477
0.219	28.0	672	0.3606	0.8438
0.3091	29.0	696	0.3586	0.8457
0.2235	30.0	720	0.3589	0.8457

Framework versions

• PEFT 0.10.0
• Transformers 4.39.3
• Pytorch 2.2.1
• Datasets 2.16.1
• Tokenizers 0.15.2