Merge pull request #2738 from opentensor/fix/thewhaleking/not-querying-metagraph-at-block

All metagraph subtensor methods now use block

Author: thewhaleking

bittensor/core/metagraph.py

@@ -1,3 +1,4 @@
+import asyncio
 import copy
 import os
 import pickle
@@ -1397,13 +1398,13 @@ async def sync(
 
         # If not a 'lite' version, compute and set weights and bonds for each neuron
         if not lite:
-            await self._set_weights_and_bonds(subtensor=subtensor)
+            await self._set_weights_and_bonds(subtensor=subtensor, block=block)
 
         # Fills in the stake associated attributes of a class instance from a chain response.
         await self._get_all_stakes_from_chain(block=block)
 
         # apply MetagraphInfo data to instance
-        await self._apply_metagraph_info()
+        await self._apply_metagraph_info(block=block)
 
     async def _initialize_subtensor(
         self, subtensor: "AsyncSubtensor"
@@ -1476,9 +1477,7 @@ async def _assign_neurons(
             self.neurons = await subtensor.neurons(block=block, netuid=self.netuid)
         self.lite = lite
 
-    async def _set_weights_and_bonds(
-        self, subtensor: Optional["AsyncSubtensor"] = None
-    ):
+    async def _set_weights_and_bonds(self, subtensor: "AsyncSubtensor", block: int):
         """
         Computes and sets the weights and bonds for each neuron in the metagraph. This method is responsible for
         processing the raw weight and bond data obtained from the network and converting it into a structured format
@@ -1499,6 +1498,7 @@ async def _set_weights_and_bonds(
                 [neuron.weights for neuron in self.neurons],
                 "weights",
                 subtensor,
+                block=block,
             )
         else:
             self.weights = self._process_weights_or_bonds(
@@ -1509,7 +1509,7 @@ async def _set_weights_and_bonds(
             )
 
     async def _process_root_weights(
-        self, data: list, attribute: str, subtensor: "AsyncSubtensor"
+        self, data: list, attribute: str, subtensor: "AsyncSubtensor", block: int
     ) -> Union[NDArray, "torch.nn.Parameter"]:
         """
         Specifically processes the root weights data for the metagraph. This method is similar to :func:`_process_weights_or_bonds`
@@ -1530,8 +1530,10 @@ async def _process_root_weights(
                 self.root_weights = self._process_root_weights(raw_root_weights_data, "weights", subtensor)
         """
         data_array = []
-        n_subnets = await subtensor.get_total_subnets() or 0
-        subnets = await subtensor.get_subnets()
+        n_subnets_, subnets = await asyncio.gather(
+            subtensor.get_total_subnets(block=block), subtensor.get_subnets(block=block)
+        )
+        n_subnets = n_subnets_ or 0
         for item in data:
             if len(item) == 0:
                 if use_torch():
@@ -1612,9 +1614,11 @@ async def _get_all_stakes_from_chain(self, block: int):
         except (SubstrateRequestException, AttributeError) as e:
             logging.debug(e)
 
-    async def _apply_metagraph_info(self):
+    async def _apply_metagraph_info(self, block: int):
         """Retrieves metagraph information for a specific subnet and applies it using a mixin."""
-        metagraph_info = await self.subtensor.get_metagraph_info(self.netuid)
+        metagraph_info = await self.subtensor.get_metagraph_info(
+            self.netuid, block=block
+        )
         if metagraph_info:
             self._apply_metagraph_info_mixin(metagraph_info=metagraph_info)
 
@@ -1711,13 +1715,13 @@ def sync(
 
         # If not a 'lite' version, compute and set weights and bonds for each neuron
         if not lite:
-            self._set_weights_and_bonds(subtensor=subtensor)
+            self._set_weights_and_bonds(subtensor=subtensor, block=block)
 
         # Fills in the stake associated attributes of a class instance from a chain response.
         self._get_all_stakes_from_chain(block=block)
 
         # apply MetagraphInfo data to instance
-        self._apply_metagraph_info()
+        self._apply_metagraph_info(block=block)
 
     def _initialize_subtensor(self, subtensor: "Subtensor") -> "Subtensor":
         """
@@ -1787,7 +1791,7 @@ def _assign_neurons(self, block: int, lite: bool, subtensor: "Subtensor"):
             self.neurons = subtensor.neurons(block=block, netuid=self.netuid)
         self.lite = lite
 
-    def _set_weights_and_bonds(self, subtensor: Optional["Subtensor"] = None):
+    def _set_weights_and_bonds(self, block: int, subtensor: "Subtensor"):
         """
         Computes and sets the weights and bonds for each neuron in the metagraph. This method is responsible for
         processing the raw weight and bond data obtained from the network and converting it into a structured format
@@ -1804,9 +1808,7 @@ def _set_weights_and_bonds(self, subtensor: Optional["Subtensor"] = None):
         """
         if self.netuid == 0:
             self.weights = self._process_root_weights(
-                [neuron.weights for neuron in self.neurons],
-                "weights",
-                subtensor,
+                [neuron.weights for neuron in self.neurons], "weights", subtensor, block
             )
         else:
             self.weights = self._process_weights_or_bonds(
@@ -1817,7 +1819,7 @@ def _set_weights_and_bonds(self, subtensor: Optional["Subtensor"] = None):
             )
 
     def _process_root_weights(
-        self, data: list, attribute: str, subtensor: "Subtensor"
+        self, data: list, attribute: str, subtensor: "Subtensor", block: int
     ) -> Union[NDArray, "torch.nn.Parameter"]:
         """
         Specifically processes the root weights data for the metagraph. This method is similar to :func:`_process_weights_or_bonds`
@@ -1838,8 +1840,8 @@ def _process_root_weights(
                 self.root_weights = self._process_root_weights(raw_root_weights_data, "weights", subtensor)
         """
         data_array = []
-        n_subnets = subtensor.get_total_subnets() or 0
-        subnets = subtensor.get_subnets()
+        n_subnets = subtensor.get_total_subnets(block=block) or 0
+        subnets = subtensor.get_subnets(block=block)
         for item in data:
             if len(item) == 0:
                 if use_torch():
@@ -1920,9 +1922,9 @@ def _get_all_stakes_from_chain(self, block: int):
         except (SubstrateRequestException, AttributeError) as e:
             logging.debug(e)
 
-    def _apply_metagraph_info(self):
+    def _apply_metagraph_info(self, block: int):
         """Retrieves metagraph information for a specific subnet and applies it using a mixin."""
-        metagraph_info = self.subtensor.get_metagraph_info(self.netuid)
+        metagraph_info = self.subtensor.get_metagraph_info(self.netuid, block=block)
         if metagraph_info:
             self._apply_metagraph_info_mixin(metagraph_info=metagraph_info)