InternLM/InternBootcamp
1
0
Fork 0
mirror of https://github.com/InternLM/InternBootcamp.git synced 2026-04-19 12:58:04 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'fix-chem-bootcamp' into 'main'

Browse source 
refactor(bootcamp): simplify InChI and SMILES generation

See merge request lilinyang/internbootcamp!17
This commit is contained in:
陈永康 2025-06-17 05:30:25 +00:00
commit 4ffecaa168
3 changed files with 7 additions and 8 deletions
Download patch file Download diff file Expand all files Collapse all files

7
internbootcamp/bootcamp/ChemStructure2Property/InChI2logPBootCamp.py
View file

@ -20,11 +20,10 @@ class InChI2logPbootcamp(Basebootcamp):
生成一组数字和目标值。
"""
self.InChIGenerator = InChIGenerator(max_atoms=self.max_atoms, min_atoms=self.min_atoms, elements=None, seed=None)
inchis = self.InChIGenerator.generate_n_valid_inchi(10)
inchis = self.InChIGenerator.generate_n_valid_inchi(1)
# print(inchis)
n = random.randint(0, 9)
# print(n)
return inchis[n]
return inchis[0]
def prompt_func(self, InChI) -> str:
@ -76,7 +75,7 @@ class InChI2logPbootcamp(Basebootcamp):
return 0.0 # Error is too large, score is 0
else:
# Linear interpolation: score decreases linearly from 1 to 0 as error goes from 0 to max_relative_error
return 1.0
# return 1.0
return 1 - (relative_error / max_relative_error) * 0.5 ## For RL
if __name__ == "__main__":

4
internbootcamp/bootcamp/ChemStructure2Property/SMILES2logPBootCamp.py
View file

@ -21,7 +21,7 @@ class SMILES2logPbootcamp(InChI2logPbootcamp):
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(10)[random.randint(0, 9)]
return self.SMILESGenerator.generate_n_valid_smiles(1)[0]
def prompt_func(self, SMILES) -> str:
@ -63,5 +63,5 @@ class SMILES2logPbootcamp(InChI2logPbootcamp):
return 0.0 # Error is too large, score is 0
else:
# Linear interpolation: score decreases linearly from 1 to 0 as error goes from 0 to max_relative_error
return 1.0
# return 1.0
return 1 - (relative_error / max_relative_error) * 0.5 ## For RL

4
internbootcamp/libs/chemStructure2Property/ChemStructureGenerator.py
View file

@ -113,7 +113,7 @@ class InChIGenerator:
except Exception as e:
# This can happen if the molecule is somehow malformed even after sanitization,
# or if InChI generation itself encounters an issue (rare).
print(f"Debug: MolToInchi failed: {e} for SMILES: {Chem.MolToSmiles(mol)}")
# print(f"Debug: MolToInchi failed: {e} for SMILES: {Chem.MolToSmiles(mol)}")
pass
return list(valid_inchi_set)
@ -244,7 +244,7 @@ class SMILESGenerator:
valid_smiles_set = set()
total_attempts_overall = 0
print(f"Attempting to generate {n} valid SMILES (min_len={self.min_len}, max_len={self.max_len})...")
# print(f"Attempting to generate {n} valid SMILES (min_len={self.min_len}, max_len={self.max_len})...")
while len(valid_smiles_set) < n:
attempts_for_current_smiles = 0
generated_this_round = False