import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

repo_id = "mryufei/llm4mat-sft-pi1m"
subfolder = "merged"

tokenizer = AutoTokenizer.from_pretrained(repo_id, subfolder=subfolder)
model = AutoModelForCausalLM.from_pretrained(
    repo_id,
    subfolder=subfolder,
    torch_dtype=torch.bfloat16,
    device_map="auto",
)

prompt = "Question: Is this molecule active?\nAnswer:"
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
with torch.no_grad():
    output = model.generate(**inputs, max_new_tokens=128, temperature=0.1)
print(tokenizer.decode(output[0], skip_special_tokens=True))

Notes

Training/evaluation scripts are from this project workflow.
Recommended prompt format: follow the benchmark prompt template used for this dataset.