Initial phase of adding PyTorch support for data analysis. Added conversion nodes.

modules/pytorch_wrapper/pt_float_create.py

@@ -0,0 +1,47 @@
+from typing import Any, Dict
+import torch
+import ast
+
+
+class PtFloatCreate:
+    """
+    Creates a PyTorch tensor with 32-bit floating point precision 
+    using values entered in the text field.
+    
+    category: PyTorch wrapper
+    """
+    
+    @classmethod
+    def INPUT_TYPES(cls) -> Dict[str, Any]:
+        """
+        Defines the input types for the function.
+
+        Returns:
+            Dict[str, Any]: A dictionary specifying required input types.
+        """
+        return {
+            "required": {
+                "data": ("STRING", {"multiline": True}),
+            }
+        }
+
+    RETURN_TYPES: tuple = ("TENSOR",)
+    FUNCTION: str = "f"
+    CATEGORY: str = "PyTorch wrapper"
+
+    def f(self, data: str) -> tuple:
+        """
+        Creates a PyTorch tensor using values entered in the text field.
+
+        Args:
+            data (str): String value in the text field.
+
+        Returns:
+            tuple: A tuple containing the tensor.
+        """
+        # Convert string to Python list
+        list_data = ast.literal_eval(data)
+
+        # Convert list to PyTorch tensor
+        tensor = torch.tensor(list_data, dtype=torch.float32)
+        return (tensor,)