ysy20020107 commited on
Commit
b00728b
·
verified ·
1 Parent(s): e899846

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +35 -25
README.md CHANGED
@@ -1,39 +1,49 @@
 
1
 
2
- ---
3
 
4
- base_model: westlake-repl/SaProt_35M_AF2
5
 
6
- library_name: peft
7
 
8
- ---
9
 
 
10
 
 
11
 
12
- # Model Card for Model-Demo-35M
13
- <slot name='description'>
14
 
15
- ## Task type
16
- Protein-level Regression
17
 
18
- ## Model input type
19
- AA Sequence
20
 
21
- ## LoRA config
22
 
23
- - **r:** 8
24
- - **lora_dropout:** 0.1
25
- - **lora_alpha:** 16
26
- - **target_modules:** ['query', 'intermediate.dense', 'value', 'output.dense', 'key']
27
- - **modules_to_save:** ['classifier']
28
 
29
- ## Training config
30
 
31
- - **optimizer:**
32
- - **class:** AdamW
33
- - **betas:** (0.9, 0.98)
34
- - **weight_decay:** 0.01
35
- - **learning rate:** 0.0005
36
- - **epoch:** 30
37
- - **batch size:** 64
38
- - **precision:** 16-mixed
39
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ base_model: westlake-repl/SaProt_35M_AF2
2
 
3
+ Model Description
4
 
5
+ This model is fine-tuned to predict the mutation effects of *Bacillus subtilis* α-amylase. It takes amino acid sequences as input and performs protein-level regression to predict quantitative enzyme activity values, enabling accurate assessment of how mutations alter enzyme function.
6
 
7
+ Task type: protein-level regression
8
 
9
+ Model input type: Amino acid sequence
10
 
11
+ Dataset
12
 
13
+ The dataset is sourced from van der Flier et al. (2024), available at https://www.sciencedirect.com/science/article/pii/S2001037024002940. It contains a total of 3706 rows, with each sample carrying 1 to 8 mutation sites. The full dataset is randomly split into training, validation, and test sets following an 8:1:1 ratio. The target label is absorbance. A higher absorbance value indicates greater starch degradation range from(), corresponding to stronger detergent activity of the amylase enzyme.
14
 
15
+ Performance (on test set)
 
16
 
17
+ Spearman correlation: 0.76
 
18
 
19
+ Coefficient of determination (R²): 0.69
 
20
 
21
+ LoRA config
22
 
23
+ r: 8
 
 
 
 
24
 
25
+ lora_dropout: 0.1
26
 
27
+ lora_alpha: 16
 
 
 
 
 
 
 
28
 
29
+ target_modules: ["query", "intermediate.dense", "value", "output.dense", "key"]
30
+
31
+ modules_to_save: ["classifier"]
32
+
33
+ Training config
34
+
35
+ optimizer:
36
+
37
+ class: AdamW
38
+
39
+ betas: (0.9, 0.98)
40
+
41
+ weight_decay: 0.01
42
+
43
+ learning rate: 0.0005
44
+
45
+ epoch: 30
46
+
47
+ batch size: 64
48
+
49
+ precision: 16-mixed