made BatchScheduler an expression. removed engine from its params and…

… using bind decorator instead to bind to engines __call__ method. updated readme and package.json

README.md

@@ -27,18 +27,11 @@ class TestExpression(Expression):
     def forward(self, input, kwargs):
     return self.Symbol(input).query("Summarize this input", kwargs)
 
-
-#Set up your engine (this example uses a mock engine)
-from your_engine_module import MockGPTXChatEngine
-engine = MockGPTXChatEngine()
-EngineRepository.register("neurosymbolic", engine_instance=engine)
-
-
 #Prepare your inputs
 inputs = ["test1", "test2", "test3"]
 
 #Create and run the BatchScheduler
-scheduler = BatchScheduler(TestExpression, num_workers=2, engine=engine, dataset=inputs)
+scheduler = BatchScheduler(TestExpression, num_workers=2, dataset=inputs, batch_size=2)
 results = scheduler.run()
 
 

package.json

@@ -1,16 +1,16 @@
 {
     "version": "0.0.1",
     "name": "AynurAda/symbatcher",
-    "description": "<Project Description>",
+    "description": "the symbolicai batch scheduler for parallel execution of expressions and batched llm calls",
     "expressions": [
         {
             "module": "src/func",
-            "type": "MyExpression"
+            "type": "BatchScheduler"
         }
     ],
     "run": {
         "module": "src/func",
-        "type": "MyExpression"
+        "type": "BatchScheduler"
     },
     "dependencies": []
-}
+}s

src/func.py

@@ -2,12 +2,13 @@
 import logging
 import threading
 from typing import Any, Callable, Dict, List, Optional, Union
-
+from symai.core_ext import bind
 from symai import Expression
 
 lgr = logging.getLogger()
 lgr.setLevel(logging.CRITICAL)
 
+
 class BatchScheduler:
     """
     A class for scheduling and executing batch operations with Expressions from symbolicai.
@@ -16,19 +17,17 @@ class BatchScheduler:
     utilizing multiple workers and an external engine for processing.
     """
 
-    def __init__(self, expr: Expression, num_workers: int, engine: Callable, dataset: List[Any], batch_size: int = 5):
+    def __init__(self, expr: Expression, num_workers: int, dataset: List[Any], batch_size: int = 5):
         """
         Initialize the BatchScheduler for symbolicai Expressions.
 
         Args:
             expr (Expression): The symbolicai Expression to be executed.
             num_workers (int): The number of worker threads to use.
-            engine (Callable): The engine function for processing batches of Expression results.
             dataset (List[Any]): The list of data points to process through the Expression.
             batch_size (int, optional): The size of each batch. Defaults to 5.
         """
         self.num_workers: int = num_workers
-        self.engine: Callable = engine
         self.dataset: List[Any] = dataset
         self.results: Dict[Any, Any] = {}
         self.arguments: List[Any] = []
@@ -40,7 +39,11 @@ def __init__(self, expr: Expression, num_workers: int, engine: Callable, dataset
         self.llm_response_ready: Dict[int, threading.Event] = {}
         self.pending_tasks: int = len(self.dataset)
         self.expr: Expression = expr()
-
+
+    @bind(engine="neurosymbolic", property="__call__")
+    def engine(self):
+        pass
+
     def single_expression(self, data_point: Any) -> Any:
         """
         Execute the symbolicai Expression for a single data point.