File size: 1,776 Bytes
2d62678
 
 
 
 
 
8d08d2f
 
 
 
 
 
 
 
 
 
7e55a0d
8d08d2f
 
 
 
 
 
 
 
 
 
 
 
 
 
9129738
 
 
 
 
 
 
8d08d2f
3c47826
 
8d08d2f
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
---
license: mit
metrics:
- accuracy
tags:
- chemistry
---
# Molecular BERT Pretrained Using ChEMBL Database

This model has been pretrained based on the methodology outlined in the paper [Pushing the Boundaries of Molecular Property Prediction for Drug Discovery with Multitask Learning BERT Enhanced by SMILES Enumeration](https://spj.science.org/doi/10.34133/research.0004). While the original model was initially trained using custom code, it has been adapted for use within the Hugging Face Transformers framework in this project.

## Model Details
The model architecture utilized is based on BERT. Here are the key configuration details:

```
BertConfig(
    vocab_size=70,
    hidden_size=256,
    num_hidden_layers=8,
    num_attention_heads=8,
    intermediate_size=1024,
    hidden_act="gelu",
    hidden_dropout_prob=0.1,
    attention_probs_dropout_prob=0.1,
    max_position_embeddings=max_seq_len,
    type_vocab_size=1,
    pad_token_id=tokenizer_pretrained.vocab["[PAD]"],
    position_embedding_type="absolute"
)
```

- Optimizer: AdamW
- Learning rate: 1e-4
- Learning rate scheduler: False
- Epochs: 50
- AMP: True
- GPU: Single Nvidia RTX 3090

## Pretraining Database
The model was pretrained using data from the ChEMBL database, specifically version 33. You can download the database from [ChEMBL](https://ftp.ebi.ac.uk/pub/databases/chembl/ChEMBLdb/latest/). 
Additionally, the dataset is available on the Hugging Face Datasets Hub and can be accessed at [Hugging Face Datasets - ChEMBL_v33_pretraining](https://huggingface.co/datasets/jonghyunlee/ChEMBL_v33_pretraining/viewer/default/train).

## Performance
The accuracy score achieved by the pretrained model is 0.9672. The testing dataset used for evaluation constitutes 10% of the ChEMBL dataset.