Downloads · 30 days
0
Model-Binding_Site_Detection-650M
SaProtHub/Model-Binding_Site_Detection-650M
Model-Binding_Site_Detection-650M is a machine learning model from SaProtHub. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
Access
Public
Updated Oct 16, 2024
Repo size
12.2 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin12.2 MB · 100%
At a glance
- License
- mit
- Access
- Public
- Created
- May 7, 2024
- Updated
- Oct 16, 2024
- SHA
- 757b8633
Model card
From the Hugging Face model README
Base model: westlake-repl/SaProt_650M_AF2
Task type: residue-level classification
Label 1 means a protein residue belongs to a small molecule binding cavity, 0 means it doesn't.
Dataset: SaProtHub/Dataset-Binding_Site_Detection-ProteinShake
Model input type: SA(Structure-aware) sequence
Performance (on test set): 0.603 Mcc
LoRA config:
- r: 8
- lora_dropout: 0
- lora_alpha: 16
- target_modules: ["query", "key", "value", "intermediate.dense", "output.dense"]
- modules_to_save: ["classifier"]
Training config:
- optimizer:
- class: AdamW
- betas: (0.9, 0.98)
- weight_decay: 0.01
- learning rate: 5e-4
- epoch: 100
- batch size: 64
- precision: 16-mixed
- License
- mit
- Created
- May 7, 2024
- Updated
- Oct 16, 2024