Downloads · 30 days
19
6% of all-time downloads
MolMetaLM-base-1B
wudejian789/MolMetaLM-base-1B
MolMetaLM-base-1B is a text generation model from wudejian789. Use it when you need the model to write or continue text.
Using 1B+ smiles (PubChem+ZINC+ChEMBL) for pretraining.
All-time downloads
314
Public
Parameters
100M
201 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors201 MB · 100%
At a glance
- Task
- Text Generation
- Model type
- llama
- Access
- Public
- Created
- Oct 12, 2025
- Updated
- Oct 26, 2025
- SHA
- 77124502
Try a prompt
Model card
From the Hugging Face model README
Using 1B+ smiles (PubChem+ZINC+ChEMBL) for pretraining.
refer to https://github.com/CSUBioGroup/MolMetaLM for more details.
Usage
Prepare tokenizer and model
from transformers import AutoTokenizer, AutoModel
tokenizer = AutoTokenizer.from_pretrained('wudejian789/MolMetaLM-base-1B')
model = AutoModel.from_pretrained('wudejian789/MolMetaLM-base-1B')
Obtain molecular representations from SMILES
smi = "COc1cc2c(cc1OC)CC([NH3+])C2"
tokenized_smi = tokenizer(" ".join(list(smi)), return_token_type_ids=False,
return_tensors='pt', max_length=512, padding='longest', truncation=True)
emb_smi = model(**tokenized_smi).last_hidden_state
print(emb_smi.shape) # batch size, seq length, embedding size
- Task
- Text Generation
- Type
- llama
- Created
- Oct 12, 2025
- Updated
- Oct 26, 2025