mirror of
https://github.com/InternLM/InternBootcamp.git
synced 2026-04-19 12:58:04 +00:00
324d068f8d
- Rename puzzle configuration files for consistency (e.g., InChI2logP_test.json) - Standardize class names to PascalCase (e.g., InChI2MRBootCamp -> InChI2MRbootCamp) - Improve code structure in various bootcamp modules for better readability and maintainability - Update import statements and file references to reflect new naming conventions - Enhance setup.py to include rdkit dependency
44 lines
1.8 KiB
Python
Executable file
44 lines
1.8 KiB
Python
Executable file
from internbootcamp.bootcamp.base import Basebootcamp
|
|
from internbootcamp.libs.chemStructure2Property.ChemStructureGenerator import SMILESGenerator
|
|
from .utils import last_boxed_only_string, remove_boxed
|
|
from rdkit import Chem
|
|
from rdkit.Chem import Crippen
|
|
|
|
from .InChI2logPBootCamp import InChI2logPbootcamp
|
|
|
|
class SMILES2logPbootcamp(InChI2logPbootcamp):
|
|
def __init__(self,min_len=5, max_len=25,
|
|
seed=None):
|
|
# super.__init__()
|
|
self.min_len = min_len
|
|
self.max_len = max_len
|
|
# self.SMILESGenerator = SMILESGenerator(min_len=min_len, max_len=max_len, seed=seed)
|
|
|
|
def case_generator(self) -> str:
|
|
"""
|
|
生成一组数字和目标值。
|
|
"""
|
|
self.SMILESGenerator = SMILESGenerator(min_len=self.min_len, max_len=self.max_len, seed=None)
|
|
return self.SMILESGenerator.generate_n_valid_smiles(1)[0]
|
|
|
|
def prompt_func(self, SMILES) -> str:
|
|
|
|
instruction = f"Given the SMILES, determine the lipophilicity (logP) value of the material. The SMILES is: {SMILES}"
|
|
instruction_following = """Let's think step by step and output the final answer within \\boxed{}.The final answer should be one float number. For example "Final Answer: \\boxed{afloat}"."""
|
|
|
|
prompt = instruction + '\n' + instruction_following
|
|
return prompt
|
|
|
|
|
|
@classmethod
|
|
def _verify_correction(cls, solution, SMILES)->bool:
|
|
"""
|
|
Verify the correction of the solution.
|
|
"""
|
|
mol = Chem.MolFromSmiles(SMILES)
|
|
true_logp = Crippen.MolLogP(mol)
|
|
solution_float = float(solution)
|
|
if true_logp == 0:
|
|
return abs(solution_float) <= 0.01 # Just check if solution is close to 0
|
|
else:
|
|
return abs(true_logp - solution_float)/abs(true_logp) <= 0.01
|