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

fix symbolic

Browse source
This commit is contained in:
lipeiji 2025-06-16 20:31:37 +08:00
commit 9f0c6efee5
21 changed files with 5898 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

84
examples/pipelines/data_configs/data_config_test.jsonl
View file

@ -1 +1,83 @@
{"bootcamp_name": "SymbolicRegression", "sample_number": 100, "config_file": "Symbolic_Regression", "bootcamp_cls_name": "SymbolicRegressionbootcamp"} {"bootcamp_name": "SymbolicRegression", "sample_number": 100, "config_file": "Symbolic_Regression", "bootcamp_cls_name": "SymbolicRegressionbootcamp"}
{"bootcamp_name": "cipher", "sample_number": 3, "config_file": "cipher", "bootcamp_cls_name": "Cipherbootcamp"}
{"bootcamp_name": "korLogicDefinitions", "sample_number": 64, "config_file": "korLogicDefinitions", "bootcamp_cls_name": "KorLogicDefinitionsbootcamp"}
{"bootcamp_name": "korLogicCooperativePrinciple", "sample_number": 64, "config_file": "korLogicCooperativePrinciple", "bootcamp_cls_name": "KorLogicCooperativePrinciplebootcamp"}
{"bootcamp_name": "korOperationUnicodeffe0", "sample_number": 64, "config_file": "korOperationUnicodeffe0", "bootcamp_cls_name": "KorOperationUnicodeffe0bootcamp"}
{"bootcamp_name": "korLogicEpistemicLogic", "sample_number": 64, "config_file": "korLogicEpistemicLogic", "bootcamp_cls_name": "KorLogicEpistemicLogicbootcamp"}
{"bootcamp_name": "korLogicDerivativeReasoningOfPropositionalLogic", "sample_number": 64, "config_file": "korLogicDerivativeReasoningOfPropositionalLogic", "bootcamp_cls_name": "KorLogicDerivativeReasoningOfPropositionalLogicbootcamp"}
{"bootcamp_name": "korLogicTruthValueModalPropositions", "sample_number": 64, "config_file": "korLogicTruthValueModalPropositions", "bootcamp_cls_name": "KorLogicTruthValueModalPropositionsbootcamp"}
{"bootcamp_name": "BBEHBuggyTables", "sample_number": 64, "config_file": "BBEHBuggyTables", "bootcamp_cls_name": "BBEHBuggyTablesbootcamp"}
{"bootcamp_name": "BbehWordSorting", "sample_number": 64, "config_file": "BbehWordSorting", "bootcamp_cls_name": "BbehWordSortingbootcamp"}
{"bootcamp_name": "korLogicFigureOfTheSyllogism", "sample_number": 64, "config_file": "korLogicFigureOfTheSyllogism", "bootcamp_cls_name": "KorLogicFigureOfTheSyllogismbootcamp"}
{"bootcamp_name": "korLogicResolution", "sample_number": 64, "config_file": "korLogicResolution", "bootcamp_cls_name": "KorLogicResolutionbootcamp"}
{"bootcamp_name": "thermometers", "sample_number": 64, "config_file": "thermometers", "bootcamp_cls_name": "Thermometersbootcamp"}
{"bootcamp_name": "korPuzzleCalcudoko", "sample_number": 64, "config_file": "korPuzzleCalcudoko", "bootcamp_cls_name": "KorPuzzleCalcudokobootcamp"}
{"bootcamp_name": "korPuzzle24Points", "sample_number": 64, "config_file": "korPuzzle24Points", "bootcamp_cls_name": "KorPuzzle24Pointsbootcamp"}
{"bootcamp_name": "korOperationUnicode2295", "sample_number": 64, "config_file": "korOperationUnicode2295", "bootcamp_cls_name": "KorOperationUnicode2295bootcamp"}
{"bootcamp_name": "nonograms", "sample_number": 64, "config_file": "nonograms", "bootcamp_cls_name": "Nonogramsbootcamp"}
{"bootcamp_name": "bbehboardgameqa", "sample_number": 64, "config_file": "bbeh_boardgame_qa", "bootcamp_cls_name": "Bbehboardgameqabootcamp"}
{"bootcamp_name": "korLogicEquivalenceCalculus", "sample_number": 64, "config_file": "korLogicEquivalenceCalculus", "bootcamp_cls_name": "KorLogicEquivalenceCalculusbootcamp"}
{"bootcamp_name": "dominosa", "sample_number": 64, "config_file": "dominosa", "bootcamp_cls_name": "Dominosabootcamp"}
{"bootcamp_name": "korPuzzleCryptoMath", "sample_number": 64, "config_file": "korPuzzleCryptoMath", "bootcamp_cls_name": "KorPuzzleCryptoMathbootcamp"}
{"bootcamp_name": "korLogicAnalogicalReasoning", "sample_number": 64, "config_file": "korLogicAnalogicalReasoning", "bootcamp_cls_name": "KorLogicAnalogicalReasoningbootcamp"}
{"bootcamp_name": "korOperationUnicode25bd", "sample_number": 64, "config_file": "korOperationUnicode25bd", "bootcamp_cls_name": "KorOperationUnicode25bdbootcamp"}
{"bootcamp_name": "korPuzzleWordBrainTeasers", "sample_number": 64, "config_file": "korPuzzleWordBrainTeasers", "bootcamp_cls_name": "KorPuzzleWordBrainTeasersbootcamp"}
{"bootcamp_name": "korOperationUnicode25a1", "sample_number": 64, "config_file": "korOperationUnicode25a1", "bootcamp_cls_name": "KorOperationUnicode25a1bootcamp"}
{"bootcamp_name": "kakurasu", "sample_number": 64, "config_file": "kakurasu", "bootcamp_cls_name": "Kakurasubootcamp"}
{"bootcamp_name": "korLogicLogicalMethodsForExploringCauseAndEffectRelationships", "sample_number": 64, "config_file": "korLogicLogicalMethodsForExploringCauseAndEffectRelationships", "bootcamp_cls_name": "KorLogicLogicalMethodsForExploringCauseAndEffectRelationshipsbootcamp"}
{"bootcamp_name": "binairo", "sample_number": 64, "config_file": "binairo", "bootcamp_cls_name": "Binairobootcamp"}
{"bootcamp_name": "korLogicInductionParadox", "sample_number": 64, "config_file": "korLogicInductionParadox", "bootcamp_cls_name": "KorLogicInductionParadoxbootcamp"}
{"bootcamp_name": "galaxies", "sample_number": 64, "config_file": "galaxies", "bootcamp_cls_name": "Galaxiesbootcamp"}
{"bootcamp_name": "BbehMultistepArithmetic", "sample_number": 64, "config_file": "BbehMultistepArithmetic", "bootcamp_cls_name": "BbehMultistepArithmeticbootcamp"}
{"bootcamp_name": "korLogicPropositionalLogicFormalization", "sample_number": 64, "config_file": "korLogicPropositionalLogicFormalization", "bootcamp_cls_name": "KorLogicPropositionalLogicFormalizationbootcamp"}
{"bootcamp_name": "korOperationUnicode25cb", "sample_number": 64, "config_file": "korOperationUnicode25cb", "bootcamp_cls_name": "KorOperationUnicode25cbbootcamp"}
{"bootcamp_name": "LightUp", "sample_number": 64, "config_file": "Light_Up", "bootcamp_cls_name": "LightUpbootcamp"}
{"bootcamp_name": "starbattle", "sample_number": 64, "config_file": "starbattle", "bootcamp_cls_name": "Starbattlebootcamp"}
{"bootcamp_name": "korOperationUnicode25b3", "sample_number": 64, "config_file": "korOperationUnicode25b3", "bootcamp_cls_name": "KorOperationUnicode25b3bootcamp"}
{"bootcamp_name": "sudoku", "sample_number": 64, "config_file": "sudoku", "bootcamp_cls_name": "Sudokubootcamp"}
{"bootcamp_name": "korLogicSpeechActs", "sample_number": 64, "config_file": "korLogicSpeechActs", "bootcamp_cls_name": "KorLogicSpeechActsbootcamp"}
{"bootcamp_name": "korOperationUnicode25a0", "sample_number": 64, "config_file": "korOperationUnicode25a0", "bootcamp_cls_name": "KorOperationUnicode25a0bootcamp"}
{"bootcamp_name": "korLogicStatisticalReasoning", "sample_number": 64, "config_file": "korLogicStatisticalReasoning", "bootcamp_cls_name": "KorLogicStatisticalReasoningbootcamp"}
{"bootcamp_name": "aquarium", "sample_number": 64, "config_file": "aquarium", "bootcamp_cls_name": "Aquariumbootcamp"}
{"bootcamp_name": "korLogicEnumerativeInductiveReasoning", "sample_number": 64, "config_file": "korLogicEnumerativeInductiveReasoning", "bootcamp_cls_name": "KorLogicEnumerativeInductiveReasoningbootcamp"}
{"bootcamp_name": "minesweeper", "sample_number": 64, "config_file": "minesweeper", "bootcamp_cls_name": "Minesweeperbootcamp"}
{"bootcamp_name": "futoshiki", "sample_number": 64, "config_file": "futoshiki", "bootcamp_cls_name": "Futoshikibootcamp"}
{"bootcamp_name": "BbehWebOfLies", "sample_number": 64, "config_file": "BbehWebOfLies", "bootcamp_cls_name": "BbehWebOfLiesbootcamp"}
{"bootcamp_name": "korOperationUnicode2605", "sample_number": 64, "config_file": "korOperationUnicode2605", "bootcamp_cls_name": "KorOperationUnicode2605bootcamp"}
{"bootcamp_name": "korPuzzleLogicPuzzle", "sample_number": 64, "config_file": "korPuzzleLogicPuzzle", "bootcamp_cls_name": "KorPuzzleLogicPuzzlebootcamp"}
{"bootcamp_name": "BbehGeometricShapes", "sample_number": 64, "config_file": "BbehGeometricShapes", "bootcamp_cls_name": "BbehGeometricShapesbootcamp"}
{"bootcamp_name": "korPuzzleWordRootsAndAffixes", "sample_number": 64, "config_file": "korPuzzleWordRootsAndAffixes", "bootcamp_cls_name": "KorPuzzleWordRootsAndAffixesbootcamp"}
{"bootcamp_name": "korPuzzleSkyscrapers", "sample_number": 64, "config_file": "korPuzzleSkyscrapers", "bootcamp_cls_name": "KorPuzzleSkyscrapersbootcamp"}
{"bootcamp_name": "korOperationUnicode25cf", "sample_number": 64, "config_file": "korOperationUnicode25cf", "bootcamp_cls_name": "KorOperationUnicode25cfbootcamp"}
{"bootcamp_name": "korLogicPredicateLogicFormalization", "sample_number": 64, "config_file": "korLogicPredicateLogicFormalization", "bootcamp_cls_name": "KorLogicPredicateLogicFormalizationbootcamp"}
{"bootcamp_name": "korPuzzleArrowMaze", "sample_number": 64, "config_file": "korPuzzleArrowMaze", "bootcamp_cls_name": "KorPuzzleArrowMazebootcamp"}
{"bootcamp_name": "bbehshuffobject", "sample_number": 64, "config_file": "bbeh_shuff_object", "bootcamp_cls_name": "Bbehshuffobjectbootcamp"}
{"bootcamp_name": "korPuzzleKukurasu", "sample_number": 64, "config_file": "korPuzzleKukurasu", "bootcamp_cls_name": "KorPuzzleKukurasubootcamp"}
{"bootcamp_name": "korLogicDynamicLogic", "sample_number": 64, "config_file": "korLogicDynamicLogic", "bootcamp_cls_name": "KorLogicDynamicLogicbootcamp"}
{"bootcamp_name": "korLogicDisjunctiveNormalFormAndConjunctiveNormalForm", "sample_number": 64, "config_file": "korLogicDisjunctiveNormalFormAndConjunctiveNormalForm", "bootcamp_cls_name": "KorLogicDisjunctiveNormalFormAndConjunctiveNormalFormbootcamp"}
{"bootcamp_name": "korPuzzleWordscapes", "sample_number": 64, "config_file": "korPuzzleWordscapes", "bootcamp_cls_name": "KorPuzzleWordscapesbootcamp"}
{"bootcamp_name": "bbehobjectproperties", "sample_number": 64, "config_file": "bbeh_object_properties", "bootcamp_cls_name": "Bbehobjectpropertiesbootcamp"}
{"bootcamp_name": "stitches", "sample_number": 64, "config_file": "stitches", "bootcamp_cls_name": "Stitchesbootcamp"}
{"bootcamp_name": "korPuzzleWordLadder", "sample_number": 64, "config_file": "korPuzzleWordLadder", "bootcamp_cls_name": "KorPuzzleWordLadderbootcamp"}
{"bootcamp_name": "BbehDyckLanguages", "sample_number": 64, "config_file": "BbehDyckLanguages", "bootcamp_cls_name": "BbehDyckLanguagesbootcamp"}
{"bootcamp_name": "calcudoku", "sample_number": 64, "config_file": "calcudoku", "bootcamp_cls_name": "Calcudokubootcamp"}
{"bootcamp_name": "korOperationUnicodeffe1", "sample_number": 64, "config_file": "korOperationUnicodeffe1", "bootcamp_cls_name": "KorOperationUnicodeffe1bootcamp"}
{"bootcamp_name": "korPuzzleCampsite", "sample_number": 64, "config_file": "korPuzzleCampsite", "bootcamp_cls_name": "KorPuzzleCampsitebootcamp"}
{"bootcamp_name": "korOperationUnicode20ac", "sample_number": 64, "config_file": "korOperationUnicode20ac", "bootcamp_cls_name": "KorOperationUnicode20acbootcamp"}
{"bootcamp_name": "korLogicPropositionalLogicConcepts", "sample_number": 64, "config_file": "korLogicPropositionalLogicConcepts", "bootcamp_cls_name": "KorLogicPropositionalLogicConceptsbootcamp"}
{"bootcamp_name": "arc", "sample_number": 64, "config_file": "arc", "bootcamp_cls_name": "Arcbootcamp"}
{"bootcamp_name": "korPuzzleWordSearch", "sample_number": 64, "config_file": "korPuzzleWordSearch", "bootcamp_cls_name": "KorPuzzleWordSearchbootcamp"}
{"bootcamp_name": "cryptomath", "sample_number": 64, "config_file": "crypto_math", "bootcamp_cls_name": "Cryptomathbootcamp"}
{"bootcamp_name": "arrowmaze", "sample_number": 64, "config_file": "arrowmaze", "bootcamp_cls_name": "Arrowmazebootcamp"}
{"bootcamp_name": "korLogicTemporalPropositions", "sample_number": 64, "config_file": "korLogicTemporalPropositions", "bootcamp_cls_name": "KorLogicTemporalPropositionsbootcamp"}
{"bootcamp_name": "bbehbooleanexpressions", "sample_number": 64, "config_file": "bbeh_boolean_expressions", "bootcamp_cls_name": "Bbehbooleanexpressionsbootcamp"}
{"bootcamp_name": "tents", "sample_number": 64, "config_file": "tents", "bootcamp_cls_name": "Tentsbootcamp"}
{"bootcamp_name": "bbehobjectcounting", "sample_number": 64, "config_file": "bbeh_object_counting", "bootcamp_cls_name": "Bbehobjectcountingbootcamp"}
{"bootcamp_name": "game24", "sample_number": 64, "config_file": "game24", "bootcamp_cls_name": "Game24bootcamp"}
{"bootcamp_name": "maze", "sample_number": 64, "config_file": "maze", "bootcamp_cls_name": "Mazebootcamp"}
{"bootcamp_name": "korOperationUnicode25ce", "sample_number": 64, "config_file": "korOperationUnicode25ce", "bootcamp_cls_name": "KorOperationUnicode25cebootcamp"}
{"bootcamp_name": "campsite", "sample_number": 64, "config_file": "campsite", "bootcamp_cls_name": "Campsitebootcamp"}
{"bootcamp_name": "korLogicCanonicalPropositions", "sample_number": 64, "config_file": "korLogicCanonicalPropositions", "bootcamp_cls_name": "KorLogicCanonicalPropositionsbootcamp"}
{"bootcamp_name": "korPuzzleConnectWords", "sample_number": 64, "config_file": "korPuzzleConnectWords", "bootcamp_cls_name": "KorPuzzleConnectWordsbootcamp"}
{"bootcamp_name": "korLogicFormalFallacies", "sample_number": 64, "config_file": "korLogicFormalFallacies", "bootcamp_cls_name": "KorLogicFormalFallaciesbootcamp"}
{"bootcamp_name": "BBEHHyperbaton", "sample_number": 64, "config_file": "BBEHHyperbaton", "bootcamp_cls_name": "BBEHHyperbatonbootcamp"}

84
examples/pipelines/data_configs/data_config_train.jsonl
View file

@ -1 +1,83 @@
{"bootcamp_name": "SymbolicRegression", "sample_number": 30000, "config_file": "Symbolic_Regression", "bootcamp_cls_name": "SymbolicRegressionbootcamp"} {"bootcamp_name": "SymbolicRegression", "sample_number": 30000, "config_file": "Symbolic_Regression", "bootcamp_cls_name": "SymbolicRegressionbootcamp"}
{"bootcamp_name": "cipher", "sample_number": 200, "config_file": "cipher", "bootcamp_cls_name": "Cipherbootcamp"}
{"bootcamp_name": "korLogicDisjunctiveNormalFormAndConjunctiveNormalForm", "sample_number": 1000, "config_file": "korLogicDisjunctiveNormalFormAndConjunctiveNormalForm", "bootcamp_cls_name": "KorLogicDisjunctiveNormalFormAndConjunctiveNormalFormbootcamp"}
{"bootcamp_name": "korLogicFigureOfTheSyllogism", "sample_number": 1000, "config_file": "korLogicFigureOfTheSyllogism", "bootcamp_cls_name": "KorLogicFigureOfTheSyllogismbootcamp"}
{"bootcamp_name": "korPuzzle24Points", "sample_number": 1000, "config_file": "korPuzzle24Points", "bootcamp_cls_name": "KorPuzzle24Pointsbootcamp"}
{"bootcamp_name": "stitches", "sample_number": 2000, "config_file": "stitches", "bootcamp_cls_name": "Stitchesbootcamp"}
{"bootcamp_name": "bbehobjectproperties", "sample_number": 2000, "config_file": "bbeh_object_properties", "bootcamp_cls_name": "Bbehobjectpropertiesbootcamp"}
{"bootcamp_name": "thermometers", "sample_number": 2000, "config_file": "thermometers", "bootcamp_cls_name": "Thermometersbootcamp"}
{"bootcamp_name": "BBEHBuggyTables", "sample_number": 2000, "config_file": "BBEHBuggyTables", "bootcamp_cls_name": "BBEHBuggyTablesbootcamp"}
{"bootcamp_name": "korLogicDynamicLogic", "sample_number": 1000, "config_file": "korLogicDynamicLogic", "bootcamp_cls_name": "KorLogicDynamicLogicbootcamp"}
{"bootcamp_name": "futoshiki", "sample_number": 2000, "config_file": "futoshiki", "bootcamp_cls_name": "Futoshikibootcamp"}
{"bootcamp_name": "korLogicStatisticalReasoning", "sample_number": 1000, "config_file": "korLogicStatisticalReasoning", "bootcamp_cls_name": "KorLogicStatisticalReasoningbootcamp"}
{"bootcamp_name": "nonograms", "sample_number": 2000, "config_file": "nonograms", "bootcamp_cls_name": "Nonogramsbootcamp"}
{"bootcamp_name": "bbehboardgameqa", "sample_number": 2000, "config_file": "bbeh_boardgame_qa", "bootcamp_cls_name": "Bbehboardgameqabootcamp"}
{"bootcamp_name": "korPuzzleWordscapes", "sample_number": 1000, "config_file": "korPuzzleWordscapes", "bootcamp_cls_name": "KorPuzzleWordscapesbootcamp"}
{"bootcamp_name": "BbehDyckLanguages", "sample_number": 2000, "config_file": "BbehDyckLanguages", "bootcamp_cls_name": "BbehDyckLanguagesbootcamp"}
{"bootcamp_name": "korOperationUnicode2605", "sample_number": 1000, "config_file": "korOperationUnicode2605", "bootcamp_cls_name": "KorOperationUnicode2605bootcamp"}
{"bootcamp_name": "korLogicResolution", "sample_number": 1000, "config_file": "korLogicResolution", "bootcamp_cls_name": "KorLogicResolutionbootcamp"}
{"bootcamp_name": "korPuzzleCryptoMath", "sample_number": 1000, "config_file": "korPuzzleCryptoMath", "bootcamp_cls_name": "KorPuzzleCryptoMathbootcamp"}
{"bootcamp_name": "korPuzzleWordRootsAndAffixes", "sample_number": 1000, "config_file": "korPuzzleWordRootsAndAffixes", "bootcamp_cls_name": "KorPuzzleWordRootsAndAffixesbootcamp"}
{"bootcamp_name": "korPuzzleSkyscrapers", "sample_number": 1000, "config_file": "korPuzzleSkyscrapers", "bootcamp_cls_name": "KorPuzzleSkyscrapersbootcamp"}
{"bootcamp_name": "cryptomath", "sample_number": 2000, "config_file": "crypto_math", "bootcamp_cls_name": "Cryptomathbootcamp"}
{"bootcamp_name": "korLogicCanonicalPropositions", "sample_number": 1000, "config_file": "korLogicCanonicalPropositions", "bootcamp_cls_name": "KorLogicCanonicalPropositionsbootcamp"}
{"bootcamp_name": "korOperationUnicodeffe0", "sample_number": 1000, "config_file": "korOperationUnicodeffe0", "bootcamp_cls_name": "KorOperationUnicodeffe0bootcamp"}
{"bootcamp_name": "BbehWebOfLies", "sample_number": 1000, "config_file": "BbehWebOfLies", "bootcamp_cls_name": "BbehWebOfLiesbootcamp"}
{"bootcamp_name": "korLogicPropositionalLogicFormalization", "sample_number": 1000, "config_file": "korLogicPropositionalLogicFormalization", "bootcamp_cls_name": "KorLogicPropositionalLogicFormalizationbootcamp"}
{"bootcamp_name": "korLogicDefinitions", "sample_number": 1000, "config_file": "korLogicDefinitions", "bootcamp_cls_name": "KorLogicDefinitionsbootcamp"}
{"bootcamp_name": "minesweeper", "sample_number": 2000, "config_file": "minesweeper", "bootcamp_cls_name": "Minesweeperbootcamp"}
{"bootcamp_name": "BbehWordSorting", "sample_number": 800, "config_file": "BbehWordSorting", "bootcamp_cls_name": "BbehWordSortingbootcamp"}
{"bootcamp_name": "korLogicDerivativeReasoningOfPropositionalLogic", "sample_number": 1000, "config_file": "korLogicDerivativeReasoningOfPropositionalLogic", "bootcamp_cls_name": "KorLogicDerivativeReasoningOfPropositionalLogicbootcamp"}
{"bootcamp_name": "korOperationUnicode2295", "sample_number": 1000, "config_file": "korOperationUnicode2295", "bootcamp_cls_name": "KorOperationUnicode2295bootcamp"}
{"bootcamp_name": "bbehbooleanexpressions", "sample_number": 2000, "config_file": "bbeh_boolean_expressions", "bootcamp_cls_name": "Bbehbooleanexpressionsbootcamp"}
{"bootcamp_name": "starbattle", "sample_number": 2000, "config_file": "starbattle", "bootcamp_cls_name": "Starbattlebootcamp"}
{"bootcamp_name": "bbehobjectcounting", "sample_number": 2000, "config_file": "bbeh_object_counting", "bootcamp_cls_name": "Bbehobjectcountingbootcamp"}
{"bootcamp_name": "arrowmaze", "sample_number": 2000, "config_file": "arrowmaze", "bootcamp_cls_name": "Arrowmazebootcamp"}
{"bootcamp_name": "korPuzzleLogicPuzzle", "sample_number": 1000, "config_file": "korPuzzleLogicPuzzle", "bootcamp_cls_name": "KorPuzzleLogicPuzzlebootcamp"}
{"bootcamp_name": "sudoku", "sample_number": 2000, "config_file": "sudoku", "bootcamp_cls_name": "Sudokubootcamp"}
{"bootcamp_name": "arc", "sample_number": 10000, "config_file": "arc", "bootcamp_cls_name": "Arcbootcamp"}
{"bootcamp_name": "korOperationUnicode25a1", "sample_number": 1000, "config_file": "korOperationUnicode25a1", "bootcamp_cls_name": "KorOperationUnicode25a1bootcamp"}
{"bootcamp_name": "bbehshuffobject", "sample_number": 1000, "config_file": "bbeh_shuff_object", "bootcamp_cls_name": "Bbehshuffobjectbootcamp"}
{"bootcamp_name": "game24", "sample_number": 8000, "config_file": "game24", "bootcamp_cls_name": "Game24bootcamp"}
{"bootcamp_name": "BbehGeometricShapes", "sample_number": 2000, "config_file": "BbehGeometricShapes", "bootcamp_cls_name": "BbehGeometricShapesbootcamp"}
{"bootcamp_name": "korPuzzleKukurasu", "sample_number": 1000, "config_file": "korPuzzleKukurasu", "bootcamp_cls_name": "KorPuzzleKukurasubootcamp"}
{"bootcamp_name": "korPuzzleWordSearch", "sample_number": 1000, "config_file": "korPuzzleWordSearch", "bootcamp_cls_name": "KorPuzzleWordSearchbootcamp"}
{"bootcamp_name": "korLogicCooperativePrinciple", "sample_number": 1000, "config_file": "korLogicCooperativePrinciple", "bootcamp_cls_name": "KorLogicCooperativePrinciplebootcamp"}
{"bootcamp_name": "korPuzzleCalcudoko", "sample_number": 1000, "config_file": "korPuzzleCalcudoko", "bootcamp_cls_name": "KorPuzzleCalcudokobootcamp"}
{"bootcamp_name": "BbehMultistepArithmetic", "sample_number": 2000, "config_file": "BbehMultistepArithmetic", "bootcamp_cls_name": "BbehMultistepArithmeticbootcamp"}
{"bootcamp_name": "korLogicSpeechActs", "sample_number": 1000, "config_file": "korLogicSpeechActs", "bootcamp_cls_name": "KorLogicSpeechActsbootcamp"}
{"bootcamp_name": "korPuzzleWordLadder", "sample_number": 1000, "config_file": "korPuzzleWordLadder", "bootcamp_cls_name": "KorPuzzleWordLadderbootcamp"}
{"bootcamp_name": "korPuzzleConnectWords", "sample_number": 1000, "config_file": "korPuzzleConnectWords", "bootcamp_cls_name": "KorPuzzleConnectWordsbootcamp"}
{"bootcamp_name": "korLogicLogicalMethodsForExploringCauseAndEffectRelationships", "sample_number": 1000, "config_file": "korLogicLogicalMethodsForExploringCauseAndEffectRelationships", "bootcamp_cls_name": "KorLogicLogicalMethodsForExploringCauseAndEffectRelationshipsbootcamp"}
{"bootcamp_name": "korPuzzleArrowMaze", "sample_number": 1000, "config_file": "korPuzzleArrowMaze", "bootcamp_cls_name": "KorPuzzleArrowMazebootcamp"}
{"bootcamp_name": "korPuzzleWordBrainTeasers", "sample_number": 1000, "config_file": "korPuzzleWordBrainTeasers", "bootcamp_cls_name": "KorPuzzleWordBrainTeasersbootcamp"}
{"bootcamp_name": "LightUp", "sample_number": 2000, "config_file": "Light_Up", "bootcamp_cls_name": "LightUpbootcamp"}
{"bootcamp_name": "korOperationUnicode25bd", "sample_number": 1000, "config_file": "korOperationUnicode25bd", "bootcamp_cls_name": "KorOperationUnicode25bdbootcamp"}
{"bootcamp_name": "binairo", "sample_number": 2000, "config_file": "binairo", "bootcamp_cls_name": "Binairobootcamp"}
{"bootcamp_name": "korOperationUnicode25b3", "sample_number": 1000, "config_file": "korOperationUnicode25b3", "bootcamp_cls_name": "KorOperationUnicode25b3bootcamp"}
{"bootcamp_name": "korLogicFormalFallacies", "sample_number": 1000, "config_file": "korLogicFormalFallacies", "bootcamp_cls_name": "KorLogicFormalFallaciesbootcamp"}
{"bootcamp_name": "korOperationUnicode25cb", "sample_number": 1000, "config_file": "korOperationUnicode25cb", "bootcamp_cls_name": "KorOperationUnicode25cbbootcamp"}
{"bootcamp_name": "korOperationUnicodeffe1", "sample_number": 1000, "config_file": "korOperationUnicodeffe1", "bootcamp_cls_name": "KorOperationUnicodeffe1bootcamp"}
{"bootcamp_name": "korLogicTemporalPropositions", "sample_number": 1000, "config_file": "korLogicTemporalPropositions", "bootcamp_cls_name": "KorLogicTemporalPropositionsbootcamp"}
{"bootcamp_name": "korLogicEquivalenceCalculus", "sample_number": 1000, "config_file": "korLogicEquivalenceCalculus", "bootcamp_cls_name": "KorLogicEquivalenceCalculusbootcamp"}
{"bootcamp_name": "korLogicEpistemicLogic", "sample_number": 1000, "config_file": "korLogicEpistemicLogic", "bootcamp_cls_name": "KorLogicEpistemicLogicbootcamp"}
{"bootcamp_name": "maze", "sample_number": 2000, "config_file": "maze", "bootcamp_cls_name": "Mazebootcamp"}
{"bootcamp_name": "calcudoku", "sample_number": 2000, "config_file": "calcudoku", "bootcamp_cls_name": "Calcudokubootcamp"}
{"bootcamp_name": "aquarium", "sample_number": 2000, "config_file": "aquarium", "bootcamp_cls_name": "Aquariumbootcamp"}
{"bootcamp_name": "kakurasu", "sample_number": 2000, "config_file": "kakurasu", "bootcamp_cls_name": "Kakurasubootcamp"}
{"bootcamp_name": "korLogicTruthValueModalPropositions", "sample_number": 1000, "config_file": "korLogicTruthValueModalPropositions", "bootcamp_cls_name": "KorLogicTruthValueModalPropositionsbootcamp"}
{"bootcamp_name": "tents", "sample_number": 2000, "config_file": "tents", "bootcamp_cls_name": "Tentsbootcamp"}
{"bootcamp_name": "korLogicInductionParadox", "sample_number": 1000, "config_file": "korLogicInductionParadox", "bootcamp_cls_name": "KorLogicInductionParadoxbootcamp"}
{"bootcamp_name": "korPuzzleCampsite", "sample_number": 1000, "config_file": "korPuzzleCampsite", "bootcamp_cls_name": "KorPuzzleCampsitebootcamp"}
{"bootcamp_name": "korLogicAnalogicalReasoning", "sample_number": 1000, "config_file": "korLogicAnalogicalReasoning", "bootcamp_cls_name": "KorLogicAnalogicalReasoningbootcamp"}
{"bootcamp_name": "campsite", "sample_number": 2000, "config_file": "campsite", "bootcamp_cls_name": "Campsitebootcamp"}
{"bootcamp_name": "korOperationUnicode25ce", "sample_number": 1000, "config_file": "korOperationUnicode25ce", "bootcamp_cls_name": "KorOperationUnicode25cebootcamp"}
{"bootcamp_name": "dominosa", "sample_number": 2000, "config_file": "dominosa", "bootcamp_cls_name": "Dominosabootcamp"}
{"bootcamp_name": "korOperationUnicode20ac", "sample_number": 1000, "config_file": "korOperationUnicode20ac", "bootcamp_cls_name": "KorOperationUnicode20acbootcamp"}
{"bootcamp_name": "korLogicEnumerativeInductiveReasoning", "sample_number": 1000, "config_file": "korLogicEnumerativeInductiveReasoning", "bootcamp_cls_name": "KorLogicEnumerativeInductiveReasoningbootcamp"}
{"bootcamp_name": "korLogicPropositionalLogicConcepts", "sample_number": 1000, "config_file": "korLogicPropositionalLogicConcepts", "bootcamp_cls_name": "KorLogicPropositionalLogicConceptsbootcamp"}
{"bootcamp_name": "korOperationUnicode25cf", "sample_number": 1000, "config_file": "korOperationUnicode25cf", "bootcamp_cls_name": "KorOperationUnicode25cfbootcamp"}
{"bootcamp_name": "korLogicPredicateLogicFormalization", "sample_number": 1000, "config_file": "korLogicPredicateLogicFormalization", "bootcamp_cls_name": "KorLogicPredicateLogicFormalizationbootcamp"}
{"bootcamp_name": "galaxies", "sample_number": 2000, "config_file": "galaxies", "bootcamp_cls_name": "Galaxiesbootcamp"}
{"bootcamp_name": "korOperationUnicode25a0", "sample_number": 1000, "config_file": "korOperationUnicode25a0", "bootcamp_cls_name": "KorOperationUnicode25a0bootcamp"}
{"bootcamp_name": "BBEHHyperbaton", "sample_number": 1000, "config_file": "BBEHHyperbaton", "bootcamp_cls_name": "BBEHHyperbatonbootcamp"}

6
examples/pipelines/puzzle_configs/InChI2logPbootcamp_test.json Normal file
View file

@ -0,0 +1,6 @@
[
{
"max_atoms": 15,
"min_atoms": 3
}
]

6
examples/pipelines/puzzle_configs/InChI2logPbootcamp_train.json Normal file
View file

@ -0,0 +1,6 @@
[
{
"max_atoms": 15,
"min_atoms": 3
}
]

6
examples/pipelines/puzzle_configs/SMILES2logPBootCamp_test.json Normal file
View file

@ -0,0 +1,6 @@
[
{
"min_len": 5,
"max_len": 25
}
]

6
examples/pipelines/puzzle_configs/SMILES2logPBootCamp_train.json Normal file
View file

@ -0,0 +1,6 @@
[
{
"min_len": 5,
"max_len": 25
}
]

32
examples/pipelines/puzzle_configs/earth_dew2humidity_test.json Normal file
View file

@ -0,0 +1,32 @@
[
{
"temperature_range": [
-20,
40
],
"temperature_dewpoint_range": [
0,
10
]
},
{
"temperature_range": [
-15,
35
],
"temperature_dewpoint_range": [
0,
8
]
},
{
"temperature_range": [
-10,
30
],
"temperature_dewpoint_range": [
0,
5
]
}
]

33
examples/pipelines/puzzle_configs/earth_dew2humidity_train.json Normal file
View file

@ -0,0 +1,33 @@
[
{
"temperature_range": [
-20,
40
],
"temperature_dewpoint_range": [
0,
10
]
},
{
"temperature_range": [
-15,
35
],
"temperature_dewpoint_range": [
0,
8
]
},
{
"temperature_range": [
-10,
30
],
"temperature_dewpoint_range": [
0,
5
]
}
]

77
examples/pipelines/puzzle_configs/linearODE_test.json Normal file
View file

@ -0,0 +1,77 @@
[
{
"k_range": [
0.15,
0.9
],
"x0_range": [
0.6,
1.8
],
"t_span": [
0,
6
],
"n_points": 60
},
{
"k_range": [
0.3,
2.0
],
"x0_range": [
0.7,
3.0
],
"t_span": [
0,
12
],
"n_points": 110
},
{
"k_range": [
0.05,
0.4
],
"x0_range": [
0.2,
1.3
],
"t_span": [
0,
8
],
"n_points": 80
},
{
"k_range": [
0.25,
1.2
],
"x0_range": [
0.4,
2.2
],
"t_span": [
0,
4
],
"n_points": 70
},
{
"k_range": [
0.2,
1.1
],
"x0_range": [
0.3,
2.5
],
"t_span": [
0,
9
],
"n_points": 90
}
]

152
examples/pipelines/puzzle_configs/linearODE_train.json Normal file
View file

@ -0,0 +1,152 @@
[
{
"k_range": [
0.1,
1.0
],
"x0_range": [
0.5,
2.0
],
"t_span": [
0,
5
],
"n_points": 50
},
{
"k_range": [
0.2,
1.5
],
"x0_range": [
0.3,
2.5
],
"t_span": [
0,
10
],
"n_points": 100
},
{
"k_range": [
0.05,
0.5
],
"x0_range": [
0.4,
1.5
],
"t_span": [
0,
7
],
"n_points": 75
},
{
"k_range": [
0.1,
0.8
],
"x0_range": [
0.6,
2.0
],
"t_span": [
0,
3
],
"n_points": 40
},
{
"k_range": [
0.3,
1.2
],
"x0_range": [
0.5,
3.0
],
"t_span": [
0,
15
],
"n_points": 120
},
{
"k_range": [
0.2,
0.8
],
"x0_range": [
0.7,
2.5
],
"t_span": [
0,
6
],
"n_points": 55
},
{
"k_range": [
0.5,
1.8
],
"x0_range": [
0.4,
3.0
],
"t_span": [
0,
8
],
"n_points": 90
},
{
"k_range": [
0.1,
0.6
],
"x0_range": [
0.3,
1.8
],
"t_span": [
0,
4
],
"n_points": 60
},
{
"k_range": [
0.15,
1.0
],
"x0_range": [
0.5,
2.3
],
"t_span": [
0,
5
],
"n_points": 70
},
{
"k_range": [
0.3,
1.5
],
"x0_range": [
0.6,
2.8
],
"t_span": [
0,
10
],
"n_points": 120
}
]

29
internbootcamp/bootcamp/ChemStructure2Property/InChI2MRBootCamp.py Executable file
View file

@ -0,0 +1,29 @@
from rdkit import Chem
from rdkit.Chem import Crippen
from .InChI2logPBootCamp import InChI2logPbootcamp
class InChI2MRBootCamp(InChI2logPbootcamp):
def prompt_func(self, InChI) -> str:
instruction = f"Given the InChI, determine the Molar Refractivity (MR) value of the material. The InChI is: {InChI}"
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, InChI)->bool:
"""
Verify the correction of the solution.
"""
mol = Chem.MolFromInchi(InChI)
true_MR = Crippen.MolMR(mol)
solution_float = float(solution)
# Handle case where true_logp is 0
if true_MR == 0:
return abs(solution_float) <= 0.01 # Just check if solution is close to 0
else:
return abs(true_MR - solution_float)/abs(true_MR) <= 0.01

57
internbootcamp/bootcamp/ChemStructure2Property/InChI2logPBootCamp.py Executable file
View file

@ -0,0 +1,57 @@
from internbootcamp.bootcamp.base import Basebootcamp
from internbootcamp.libs.chemStructure2Property.ChemStructureGenerator import InChIGenerator
from .utils import last_boxed_only_string, remove_boxed
from rdkit import Chem
from rdkit.Chem import Crippen
class InChI2logPbootcamp(Basebootcamp):
def __init__(self, max_atoms=15, min_atoms=3, elements=None, seed=None):
# super.__init__()
self.InChIGenerator = InChIGenerator(max_atoms=max_atoms, min_atoms=min_atoms, elements=elements, seed=seed)
self.tolerance_factor = tolerance_factor # 1 for 1% error consider true, 0.1 for 0.1% error true, 10 for 10% error
def case_generator(self) -> str:
"""
生成一组数字和目标值。
"""
return self.InChIGenerator.generate_n_valid_inchi(1)[0]
def prompt_func(self, InChI) -> str:
instruction = f"Given the InChI, determine the lipophilicity (logP) value of the material. The InChI is: {InChI}"
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
@staticmethod
def extract_output(output):
"""
Extract the output from the solution.
Args:
output: Model output to be processed.
Returns:
The processed output.
"""
output = last_boxed_only_string(output)
if output is None:
return None
return remove_boxed(output)
@classmethod
def _verify_correction(cls, solution, InChI)->bool:
"""
Verify the correction of the solution.
"""
mol = Chem.MolFromInchi(InChI)
true_logp = Crippen.MolLogP(mol)
solution_float = float(solution)
# Handle case where true_logp is 0
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

32
internbootcamp/bootcamp/ChemStructure2Property/SMILES2MRBootCamp.py Executable file
View file

@ -0,0 +1,32 @@
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 .SMILES2logPBootCamp import SMILES2logPBootCamp
class SMILES2MRBootCamp(SMILES2logPBootCamp):
def prompt_func(self, SMILES) -> str:
instruction = f"Given the SMILES, determine the Molar Refractivity (MR) 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_MR = Crippen.MolMR(mol)
solution_float = float(solution)
if true_MR == 0:
return abs(solution_float) <= 0.01 # Just check if solution is close to 0
else:
return abs(true_MR - solution_float)/abs(true_MR) <= 0.01

41
internbootcamp/bootcamp/ChemStructure2Property/SMILES2logPBootCamp.py Executable file
View file

@ -0,0 +1,41 @@
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.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。
"""
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

43
internbootcamp/bootcamp/ChemStructure2Property/utils.py Normal file
View file

@ -0,0 +1,43 @@
def remove_boxed(s):
if "\\boxed " in s:
left = "\\boxed "
assert s[:len(left)] == left
return s[len(left):]
left = "\\boxed{"
assert s[:len(left)] == left
assert s[-1] == "}"
return s[len(left):-1]
def last_boxed_only_string(string):
idx = string.rfind("\\boxed")
if "\\boxed " in string:
return "\\boxed " + string.split("\\boxed ")[-1].split("$")[0]
if idx < 0:
idx = string.rfind("\\fbox")
if idx < 0:
return None
i = idx
right_brace_idx = None
num_left_braces_open = 0
while i < len(string):
if string[i] == "{":
num_left_braces_open += 1
if string[i] == "}":
num_left_braces_open -= 1
if num_left_braces_open == 0:
right_brace_idx = i
break
i += 1
if right_brace_idx is None:
retval = None
else:
retval = string[idx:right_brace_idx + 1]
return retval

7
internbootcamp/bootcamp/__init__.py
View file

@ -30,6 +30,13 @@ from .bbeh_hyperbaton.hyperbaton_default import BBEHHyperbatonbootcamp
from .bbeh_boardgame_qa.bbeh_boardgame_qa import Bbehboardgameqabootcamp from .bbeh_boardgame_qa.bbeh_boardgame_qa import Bbehboardgameqabootcamp
from .bbeh_boolean_expressions.bbeh_boolean_expressions import Bbehbooleanexpressionsbootcamp from .bbeh_boolean_expressions.bbeh_boolean_expressions import Bbehbooleanexpressionsbootcamp
from .ChemStructure2Property.InChI2logPBootCamp import InChI2logPbootcamp
from .ChemStructure2Property.InChI2MRBootCamp import InChI2MRBootCamp
from .ChemStructure2Property.SMILES2logPBootCamp import SMILES2logPBootCamp
from .ChemStructure2Property.SMILES2MRBootCamp import SMILES2MRBootCamp
from .kakurasu.kakurasu import Kakurasubootcamp from .kakurasu.kakurasu import Kakurasubootcamp
from .nonograms.nonograms import Nonogramsbootcamp from .nonograms.nonograms import Nonogramsbootcamp
from .hitori.hitori import Hitoribootcamp from .hitori.hitori import Hitoribootcamp

85
internbootcamp/bootcamp/earth_dew2humidity/earth_dew2humidity.py Normal file
View file

@ -0,0 +1,85 @@
import re
import json
import numpy as np
from internbootcamp.bootcamp.base import Basebootcamp
class earth_dew2humidity(Basebootcamp):
def __init__(
self,
temperature_range=(-20, 40),
temperature_dewpoint_range=(0, 10),
seed=None
):
self.temperature_range, self.temperature_dewpoint_range = temperature_range, temperature_dewpoint_range
if seed is not None:
np.random.seed(seed)
def case_generator(self):
# 1. 随机采样参数 dewpoint 和 temperature
temperature_original = float(np.random.uniform(*self.temperature_range))
dewpoint_original = temperature_original - float(np.random.uniform(*self.temperature_dewpoint_range))
# 2. 计算湿度
dewpoint = dewpoint_original + 273.15
temperature = temperature_original + 273.15
e = 611.2 * np.exp(17.67 * (dewpoint - 273.15) / (dewpoint - 29.65))
e_s = 611.2 * np.exp(17.67 * (temperature - 273.15) / (temperature - 29.65))
rh = e / e_s * 100
return {"dewpoint": dewpoint_original, "temperature": temperature_original, "humidity": float(rh)}
def prompt_func(self, identity) -> str:
dewpoint = identity["dewpoint"]
temperature = identity["temperature"]
return (
f"下面给出露点温度（dewpoint）={dewpoint} (摄氏度)\n温度（temperature）={temperature} (摄氏度)\n"
"请计算湿度，计算公式为：\n"
"dewpoint = dewpoint + 273.15，temperature = temperature + 273.15\n"
"e = 611.2 * np.exp(17.67 * (dewpoint - 273.15) / (dewpoint - 29.65))\n"
"e_s = 611.2 * np.exp(17.67 * (temperature - 273.15) / (temperature - 29.65))\n"
"relative humidity = e / e_s * 100\n"
"只需返回 “relative humidity = ”。"
)
@staticmethod
def extract_output(output: str) -> str:
# 用正则提取“relative humidity = …”右侧的表达式
m = re.search(r"relative humidity\s*=\s*([^\n\r]+)", output)
return m.group(1).strip() if m else None
@classmethod
def _verify_correction(cls, solution: str, identity: dict) -> bool:
# 解析 LLM 给出的系数 c，形如 “c*x”
solution = solution.replace(" ", "")
try:
c = float(solution)
except:
return False
# print(c)
# 验证 c ≈ k
return abs(c - identity["humidity"]) < 1e-2
if __name__ == "__main__":
bootcamp = earth_dew2humidity(seed=123)
# 生成几个样例
examples = [bootcamp.case_generator() for _ in range(3)]
print(examples)
print(bootcamp.prompt_func(examples[0]))
print(bootcamp.extract_output("xxxxx relative humidity = 111222 "))
solution = bootcamp.extract_output("xxxxx relative humidity = 84.79 ")
print(bootcamp._verify_correction(solution, examples[0]))
solution = bootcamp.extract_output("xxxxx relative humidity = 83.79 ")
print(bootcamp._verify_correction(solution, examples[0]))
for identity in examples:
# 构造“模型”返回答案，模拟 LLM 的输出
humidity = identity["humidity"]
sol = f"{humidity:.4f}"
# 调用 Basebootcamp 提供的 verify_score 接口进行验证
score = bootcamp.verify_score(sol, identity, short_threshold=1e-2)
# 打印结果
print(json.dumps({
"identity": identity,
"solution": sol,
"verify_score": score
}, ensure_ascii=False, indent=2))

78
internbootcamp/bootcamp/linearODE/linearODE.py Normal file
View file

@ -0,0 +1,78 @@
import re
import json
import numpy as np
from scipy.integrate import odeint
from internbootcamp.bootcamp.base import Basebootcamp
class LinearODEBootcamp(Basebootcamp):
def __init__(
self,
k_range=(0.1, 1.0),
x0_range=(0.5, 2.0),
t_span=(0, 5),
n_points=50,
seed=None
):
self.k_range, self.x0_range = k_range, x0_range
self.t0, self.t1 = t_span
self.n_points = n_points
if seed is not None:
np.random.seed(seed)
def case_generator(self):
# 1. 随机采样参数 k 和初始值 x0
k = float(np.random.uniform(*self.k_range))
x0 = float(np.random.uniform(*self.x0_range))
# 2. 构造时间序列并模拟 dx/dt = -k * x
t = np.linspace(self.t0, self.t1, self.n_points).tolist()
def model(x, t_val):
return -k * x
x = odeint(model, x0, t).flatten().tolist()
return {"t": t, "x": x, "k": k}
def prompt_func(self, identity) -> str:
# 将 (t, x) 对格式化为提示
points = ", ".join(f"({t:.2f}, {x:.2f})"
for t, x in zip(identity["t"], identity["x"]))
return (
f"下面给出变量 x(t) 的观测数据点：\n{points}\n\n"
"请找出其满足的微分方程，形式为：dx/dt = f(x)。\n"
"只需返回 “dx/dt = <表达式>”。"
)
@staticmethod
def extract_output(output: str) -> str:
# 用正则提取“dx/dt = …”右侧的表达式
m = re.search(r"dx/dt\s*=\s*([^\n\r]+)", output)
return m.group(1).strip() if m else None
@classmethod
def _verify_correction(cls, solution: str, identity: dict) -> bool:
# 解析 LLM 给出的系数 c，形如 “c*x”
sol = solution.replace(" ", "")
match = re.fullmatch(r"([\-0-9\.eE]+)\*x", sol)
if not match:
return False
c = float(match.group(1))
# 验证 c ≈ -k
return abs(c + identity["k"]) < 1e-2
if __name__ == "__main__":
bootcamp = LinearODEBootcamp(seed=123)
# 生成几个样例
examples = [bootcamp.case_generator() for _ in range(3)]
for identity in examples:
# 构造“模型”返回答案，模拟 LLM 的输出
coeff = -identity["k"]
sol = f"{coeff:.4f}*x"
# 调用 Basebootcamp 提供的 verify_score 接口进行验证
score = bootcamp.verify_score(sol, identity, short_threshold=1e-2)
# 打印结果
print(json.dumps({
"identity": identity,
"solution": sol,
"verify_score": score
}, ensure_ascii=False, indent=2))

4755
internbootcamp/bootcamp/med_calculator/med_calculator.json Normal file
View file

File diff suppressed because it is too large Load diff

8
internbootcamp/bootcamp/med_calculator/med_calculator.py
View file

@ -1,7 +1,11 @@
import json import json
import math import math
import random import random
<<<<<<< HEAD
from bootcamp import Basebootcamp from bootcamp import Basebootcamp
=======
from internbootcamp.bootcamp.base import Basebootcamp
>>>>>>> remotes/origin/main
def remove_boxed(s): def remove_boxed(s):
if "\\boxed " in s: if "\\boxed " in s:
@ -154,7 +158,11 @@ class Medcalculatorbootcamp(Basebootcamp): # 医学计算器类
return cases return cases
if __name__ == '__main__': if __name__ == '__main__':
<<<<<<< HEAD
bootcamp = MedCalculatorbootcamp(seed=42) bootcamp = MedCalculatorbootcamp(seed=42)
=======
bootcamp = MedCalculatorSandbox(seed=42)
>>>>>>> remotes/origin/main
identity = bootcamp.case_generator() identity = bootcamp.case_generator()
print(f'{identity = }') print(f'{identity = }')

279
internbootcamp/libs/chemStructure2Property/ChemStructureGenerator.py Executable file
View file

@ -0,0 +1,279 @@
import random
from rdkit import Chem
from rdkit import RDLogger
from rdkit.Chem.inchi import MolToInchi
from rdkit.Chem import Crippen
class InChIGenerator:
def __init__(self, max_atoms=15, min_atoms=3, elements=None,
seed=None):
RDLogger.DisableLog('rdApp.*')
random.seed(42) if seed is None else random.seed(seed)
self.max_atoms = max_atoms
self.min_atoms = min_atoms
if elements is None:
self.elements = ['C', 'N', 'O', 'S', 'P', 'F', 'Cl', 'Br', 'I']
else:
self.elements = elements
def generate_random_molecule_rdkit(self):
"""
Generates a random molecule using RDKit's RWMol.
Can optionally try to assign random stereochemistry.
"""
rw_mol = Chem.RWMol() # Editable molecule
num_atoms_to_add = random.randint(self.min_atoms, self.max_atoms)
if num_atoms_to_add == 0:
return None
# Add the first atom
atom_symbol = random.choice(self.elements)
rw_mol.AddAtom(Chem.Atom(atom_symbol))
# Add subsequent atoms and connect them
for i in range(1, num_atoms_to_add):
if not rw_mol.GetNumAtoms(): break
existing_atom_idx = random.randrange(rw_mol.GetNumAtoms())
new_atom_symbol = random.choice(self.elements)
new_atom_idx = rw_mol.AddAtom(Chem.Atom(new_atom_symbol))
bond_type = random.choice([Chem.BondType.SINGLE, Chem.BondType.SINGLE, Chem.BondType.DOUBLE, Chem.BondType.TRIPLE])
rw_mol.AddBond(existing_atom_idx, new_atom_idx, bond_type)
# Attempt to form rings
if rw_mol.GetNumAtoms() > 2:
num_rings_to_try = random.randint(0, rw_mol.GetNumAtoms() // 3)
for _ in range(num_rings_to_try):
if rw_mol.GetNumAtoms() < 2: break
atom_indices = list(range(rw_mol.GetNumAtoms()))
if len(atom_indices) < 2: break
idx1, idx2 = random.sample(atom_indices, 2)
if rw_mol.GetBondBetweenAtoms(idx1, idx2) is None:
rw_mol.AddBond(idx1, idx2, Chem.BondType.SINGLE) # Usually single for new rings
try:
mol = rw_mol.GetMol()
Chem.SanitizeMol(mol) # Crucial: checks valency, aromaticity, etc.
if mol.GetNumAtoms() > 0:
# It might create non-physical or conflicting assignments.
# InChI will represent whatever stereo is defined.
Chem.AssignStereochemistryFrom3D(mol) # If 3D coords were present (not here)
# Or, more directly, find potential chiral centers and assign randomly
chiral_centers = Chem.FindMolChiralCenters(mol, includeUnassigned=True)
for center_idx, stereo_val in chiral_centers:
if stereo_val == '?': # Unassigned
atom = mol.GetAtomWithIdx(center_idx)
if random.choice([True, False]):
atom.SetChiralTag(Chem.ChiralType.CHI_TETRAHEDRAL_CW)
else:
atom.SetChiralTag(Chem.ChiralType.CHI_TETRAHEDRAL_CCW)
# Re-sanitize after modifying chiral tags might be good practice
Chem.SanitizeMol(mol)
return mol
except Exception as e:
# print(f"Debug: RDKit molecule construction/sanitization failed: {e}")
return None
def generate_n_valid_inchi(self, n):
"""
Generates N valid, unique InChI strings.
kwargs_for_mol_gen are passed to generate_random_molecule_rdkit.
"""
valid_inchi_set = set()
total_attempts_overall = 0
while len(valid_inchi_set) < n:
attempts_for_current_inchi = 0
generated_this_round = False
while not generated_this_round:
total_attempts_overall += 1
attempts_for_current_inchi += 1
mol = self.generate_random_molecule_rdkit()
if mol:
try:
inchi_string = MolToInchi(mol)
mol = Chem.MolFromInchi(inchi_string)
logp = Crippen.MolLogP(mol)
if inchi_string and inchi_string not in valid_inchi_set:
valid_inchi_set.add(inchi_string)
generated_this_round = True
break # Found one
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)}")
pass
return list(valid_inchi_set)
class SMILESGenerator:
def __init__(self, min_len=5, max_len=25,
seed=None):
RDLogger.DisableLog('rdApp.*')
random.seed(42) if seed is None else random.seed(seed)
self.min_len = min_len
self.max_len = max_len
def is_valid_smiles(self, smi):
"""Checks if a SMILES string is valid using RDKit."""
if not smi:
return False
mol = Chem.MolFromSmiles(smi, sanitize=False) # Parse without sanitization first
if mol is None:
return False
try:
Chem.SanitizeMol(mol)
return True
except Exception as e:
return False
def generate_random_smiles_candidate(self):
"""
Generates a random string that might be a SMILES string.
This is a VERY naive generator and will produce many invalid SMILES.
"""
# A basic set of SMILES characters
# More comprehensive: C,c,N,n,O,o,S,s,P,p,F,Cl,Br,I,B,Si,Se,*,[nH],[nH+],[cH-],...
# Also: -,=,#,$,:,.,(,),[,],%,0-9 (for ring closures and isotopes/charges)
atom_chars = ['C', 'N', 'O', 'S', 'F', 'Cl', 'Br', 'I', 'P']
aromatic_chars = ['c', 'n', 'o', 's']
bond_chars = ['-', '=', '#']
branch_chars = ['(', ')']
ring_digits = [str(i) for i in range(1, 10)] # 1-9
# More complex elements like charges, isotopes, chiral centers are harder to randomize simply
# For simplicity, we'll stick to a subset.
all_chars = atom_chars + aromatic_chars + bond_chars + branch_chars + ring_digits
length = random.randint(self.min_len, self.max_len)
candidate = ""
candidate += random.choice(atom_chars + aromatic_chars)
open_parentheses = 0
open_rings = {}
for _ in range(length - 1):
choices = []
weights = []
choices.extend(atom_chars + aromatic_chars)
weights.extend([10] * (len(atom_chars) + len(aromatic_chars)))
if candidate and (candidate[-1].isalpha() or candidate[-1] == ')' or candidate[-1].isdigit()):
choices.extend(bond_chars)
weights.extend([5] * len(bond_chars))
if candidate and (candidate[-1].isalpha() or candidate[-1] == ')' or candidate[-1].isdigit()):
choices.append('(')
weights.append(3)
if open_parentheses > 0:
choices.append(')')
weights.append(3)
# Ring closures
if candidate and (candidate[-1].isalpha() or candidate[-1] == ')'):
# Try to close an open ring
open_ring_digits = [d for d, status in open_rings.items() if status == 'open']
if open_ring_digits and random.random() < 0.5: # 50% chance to close an open ring
digit_to_close = random.choice(open_ring_digits)
choices.append(digit_to_close)
weights.append(5)
else: # Try to open a new ring
available_digits = [d for d in ring_digits if d not in open_rings or open_rings[d] == 'closed']
if available_digits:
choices.extend(available_digits)
weights.extend([2] * len(available_digits))
if not choices: # Fallback if no valid options (e.g., after certain bonds)
chosen_char = random.choice(atom_chars)
else:
chosen_char = random.choices(choices, weights=weights, k=1)[0]
# Update state
if chosen_char == '(':
open_parentheses += 1
elif chosen_char == ')':
if open_parentheses > 0:
open_parentheses -= 1
else:
continue # Don't add a closing parenthesis if none are open
elif chosen_char in ring_digits:
if chosen_char not in open_rings or open_rings[chosen_char] == 'closed':
open_rings[chosen_char] = 'open'
elif open_rings[chosen_char] == 'open':
open_rings[chosen_char] = 'closed'
candidate += chosen_char
# Attempt to close any remaining open parentheses
candidate += ')' * open_parentheses
# Attempt to close any remaining open rings (very crudely)
for digit, status in open_rings.items():
if status == 'open':
# Find a suitable place to close it - this is hard without graph info
# For now, just append another atom and the digit if possible
if candidate and (candidate[-1].isalpha() or candidate[-1] == ')'):
if random.random() < 0.7 and len(candidate) < self.max_len -1 : # Add another atom then close
candidate += random.choice(atom_chars) + digit
else: # Just append the digit (might be invalid)
candidate += digit
return candidate
def generate_n_valid_smiles(self, n):
"""Generates N valid, unique (canonical) SMILES strings."""
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})...")
while len(valid_smiles_set) < n:
attempts_for_current_smiles = 0
generated_this_round = False
while not generated_this_round:
total_attempts_overall += 1
attempts_for_current_smiles += 1
candidate = self.generate_random_smiles_candidate()
if self.is_valid_smiles(candidate):
mol = Chem.MolFromSmiles(candidate) # Re-parse to be sure and for canonicalization
if mol: # Should be true if is_valid_smiles passed
canonical_smi = Chem.MolToSmiles(mol, isomericSmiles=True, canonical=True)
try:
mol = Chem.MolFromSmiles(canonical_smi)
logp = Crippen.MolLogP(mol)
if canonical_smi not in valid_smiles_set:
valid_smiles_set.add(canonical_smi)
# print(f"Generated ({len(valid_smiles_set)}/{n}): {canonical_smi} (after {attempts_for_current_smiles} attempts for this one, {total_attempts_overall} total)")
generated_this_round = True
break # Found one, move to the next
except Exception as e:
pass
return list(valid_smiles_set)
if __name__ == "__main__":
aInChIGenerator = InChIGenerator(max_atoms=15, min_atoms=3)
inchi_10 = aInChIGenerator.generate_n_valid_inchi(10)
print(inchi_10)
aSMILESGenerator = SMILESGenerator(min_len=5, max_len=25)
smiles_10 = aSMILESGenerator.generate_n_valid_smiles(10)
print(smiles_10)