Make BotAx registries string-based by class name

Summary: Make `ACQUISITION_REGISTRY` and other modular [BotAx registries](https://www.internalfb.com/intern/diffusion/FBS/browsefile/master/fbcode/ax/ax/storage/botorch_modular_registry.py?lines=73-100%2C57-72) string-based by class name

Reviewed By: lena-kashtelyan

Differential Revision: D25376150

fbshipit-source-id: 1bed59bf8fd2c6fa7d7e4e10312f4ee7bb31380d

ax/storage/botorch_modular_registry.py

@@ -52,57 +52,57 @@
 # to `CLASS_TO_REGISTRY` and `CLASS_TO_REVERSE_REGISTRY` in this file.
 
 """
-Mapping of modular Ax `Acquisition` classes to ints.
+Mapping of modular Ax `Acquisition` classes to class name strings.
 """
-ACQUISITION_REGISTRY: Dict[Type[Acquisition], int] = {
-    Acquisition: 0,
-    KnowledgeGradient: 1,
-    MultiFidelityKnowledgeGradient: 2,
-    MaxValueEntropySearch: 3,
-    MultiFidelityMaxValueEntropySearch: 4,
+ACQUISITION_REGISTRY: Dict[Type[Acquisition], str] = {
+    Acquisition: "Acquisition",
+    KnowledgeGradient: "KnowledgeGradient",
+    MaxValueEntropySearch: "MaxValueEntropySearch",
+    MultiFidelityKnowledgeGradient: "MultiFidelityKnowledgeGradient",
+    MultiFidelityMaxValueEntropySearch: "MultiFidelityMaxValueEntropySearch",
 }
 
 
 """
-Mapping of BoTorch `Model` classes to ints.
+Mapping of BoTorch `Model` classes to class name strings.
 """
-MODEL_REGISTRY: Dict[Type[Model], int] = {
-    FixedNoiseGP: 0,
-    SingleTaskGP: 1,
-    FixedNoiseMultiFidelityGP: 2,
-    SingleTaskMultiFidelityGP: 3,
-    ModelListGP: 4,
-    FixedNoiseMultiTaskGP: 5,
-    MultiTaskGP: 6,
+MODEL_REGISTRY: Dict[Type[Model], str] = {
+    FixedNoiseGP: "FixedNoiseGP",
+    FixedNoiseMultiFidelityGP: "FixedNoiseMultiFidelityGP",
+    FixedNoiseMultiTaskGP: "FixedNoiseMultiTaskGP",
+    ModelListGP: "ModelListGP",
+    MultiTaskGP: "MultiTaskGP",
+    SingleTaskGP: "SingleTaskGP",
+    SingleTaskMultiFidelityGP: "SingleTaskMultiFidelityGP",
 }
 
 
 """
-Mapping of Botorch `AcquisitionFunction` classes to ints.
+Mapping of Botorch `AcquisitionFunction` classes to class name strings.
 """
-ACQUISITION_FUNCTION_REGISTRY: Dict[Type[AcquisitionFunction], int] = {
-    qExpectedImprovement: 0,
-    qNoisyExpectedImprovement: 1,
-    qKnowledgeGradient: 2,
-    qMultiFidelityKnowledgeGradient: 3,
-    qMaxValueEntropy: 4,
-    qMultiFidelityMaxValueEntropy: 5,
+ACQUISITION_FUNCTION_REGISTRY: Dict[Type[AcquisitionFunction], str] = {
+    qExpectedImprovement: "qExpectedImprovement",
+    qKnowledgeGradient: "qKnowledgeGradient",
+    qMaxValueEntropy: "qMaxValueEntropy",
+    qMultiFidelityKnowledgeGradient: "qMultiFidelityKnowledgeGradient",
+    qMultiFidelityMaxValueEntropy: "qMultiFidelityMaxValueEntropy",
+    qNoisyExpectedImprovement: "qNoisyExpectedImprovement",
 }
 
 
 """
-Mapping of BoTorch `MarginalLogLikelihood` classes to ints.
+Mapping of BoTorch `MarginalLogLikelihood` classes to class name strings.
 """
-MLL_REGISTRY: Dict[Type[MarginalLogLikelihood], int] = {
-    ExactMarginalLogLikelihood: 0,
-    SumMarginalLogLikelihood: 1,
+MLL_REGISTRY: Dict[Type[MarginalLogLikelihood], str] = {
+    ExactMarginalLogLikelihood: "ExactMarginalLogLikelihood",
+    SumMarginalLogLikelihood: "SumMarginalLogLikelihood",
 }
 
 
 """
 Overarching mapping from encoded classes to registry map.
 """
-CLASS_TO_REGISTRY: Dict[Any, Dict[Type[Any], int]] = {
+CLASS_TO_REGISTRY: Dict[Any, Dict[Type[Any], str]] = {
     Acquisition: ACQUISITION_REGISTRY,
     AcquisitionFunction: ACQUISITION_FUNCTION_REGISTRY,
     MarginalLogLikelihood: MLL_REGISTRY,
@@ -113,30 +113,30 @@
 """
 Reverse registries for decoding.
 """
-REVERSE_ACQUISITION_REGISTRY: Dict[int, Type[Acquisition]] = {
+REVERSE_ACQUISITION_REGISTRY: Dict[str, Type[Acquisition]] = {
     v: k for k, v in ACQUISITION_REGISTRY.items()
 }
 
 
-REVERSE_MODEL_REGISTRY: Dict[int, Type[Model]] = {
+REVERSE_MODEL_REGISTRY: Dict[str, Type[Model]] = {
     v: k for k, v in MODEL_REGISTRY.items()
 }
 
 
-REVERSE_ACQUISITION_FUNCTION_REGISTRY: Dict[int, Type[AcquisitionFunction]] = {
+REVERSE_ACQUISITION_FUNCTION_REGISTRY: Dict[str, Type[AcquisitionFunction]] = {
     v: k for k, v in ACQUISITION_FUNCTION_REGISTRY.items()
 }
 
 
-REVERSE_MLL_REGISTRY: Dict[int, Type[MarginalLogLikelihood]] = {
+REVERSE_MLL_REGISTRY: Dict[str, Type[MarginalLogLikelihood]] = {
     v: k for k, v in MLL_REGISTRY.items()
 }
 
 
 """
 Overarching mapping from encoded classes to reverse registry map.
 """
-CLASS_TO_REVERSE_REGISTRY: Dict[Any, Dict[int, Type[Any]]] = {
+CLASS_TO_REVERSE_REGISTRY: Dict[Any, Dict[str, Type[Any]]] = {
     Acquisition: REVERSE_ACQUISITION_REGISTRY,
     AcquisitionFunction: REVERSE_ACQUISITION_FUNCTION_REGISTRY,
     MarginalLogLikelihood: REVERSE_MLL_REGISTRY,
@@ -146,7 +146,6 @@
 
 def register_acquisition(acq_class: Type[Acquisition]) -> None:
     """Add a custom acquisition class to the SQA and JSON registries."""
-    ACQUISITION_REGISTRY = CLASS_TO_REGISTRY[Acquisition]
-    index = len(ACQUISITION_REGISTRY)
-    CLASS_TO_REGISTRY[Acquisition].update({acq_class: index})
-    CLASS_TO_REVERSE_REGISTRY[Acquisition].update({index: acq_class})
+    class_name = acq_class.__name__
+    CLASS_TO_REGISTRY[Acquisition].update({acq_class: class_name})
+    CLASS_TO_REVERSE_REGISTRY[Acquisition].update({class_name: acq_class})