backward pass working on long graph

.gitignore

@@ -82,6 +82,8 @@ target/
 profile_default/
 ipython_config.py
 
+dev.py
+
 # pyenv
 #   For a library or package, you might want to ignore these files since the code is
 #   intended to run in multiple environments; otherwise, check them in:

README.md

@@ -4,10 +4,20 @@
 
 Newest ML framework that you propbaly don't need, <br>
 this is really autograd engine backed by numpy<br><br>
+>> `tinytorch.py` shall always remain under 1000 lines. if not we will revert commit
 
-#### why this exists
-Bcs I was bored 
+
+
+$$
+f(x) =x^3+x
+$$
+<p align="center">
+  <img src="images/image-1.png" alt="Alt text" width="70%">
+</p>
 
+<p align="center">
+  <img src="images/image.png" alt="Alt text">
+</p>
 
 #### Visulization 
 If you want to see your computation graph run visulize.py
@@ -17,15 +27,19 @@ requirements
 pip install graphviz
 sudo apt-get install -y graphviz # IDK what to do for windows I use wsl
 ```
+<p align="center">
+  <img src="images/image-2.png"  width="50%" >
+</p>
+
+
+#### why this exists
+Bcs I was bored
 
+### DEV BLOG
+Part 1: [pythonstuff/build-tensors](https://www.pythonstuff.com/blog/buinging%20own%20autograd%20engine%20tinytorch)
 
 #### powerlevel
 1.0 -  karpathy [micrograd](https://github.com/karpathy/micrograd) (really simple, not much you can do with it) <br>
 3.14 -  [tinytorch](https://github.com/joey00072/nanograd) (simpile and you can do lot of things with it) <= ❤️ <br>
 69 - [tinygrad](https://github.com/tinygrad/tinygrad) (no longer simple you can do lot more)<br>
 ∞  -  [pytorch](https://github.com/pytorch/pytorch) (goat library, that makes gpu go burrr)<br>
-
-
-
-### DEV BLOG
-Part 1: [pythonstuff/build-tensors](https://www.pythonstuff.com/blog/buinging%20own%20autograd%20engine%20tinytorch)

tinytorch.py

@@ -24,6 +24,9 @@ def __repr__(self):
         return f"tensor({self.data})"
 
     def backward(self,grad=None):
+        if self._ctx is None:
+            return 
+
         if grad is None:
             grad = Tensor([1.])
             self.grad = grad
@@ -34,8 +37,11 @@ def backward(self,grad=None):
         grads = op.backward(self._ctx,grad)
 
         for tensor,grad in zip(child_nodes,grads):
-            tensor.grad = grad
-
+            if tensor.grad is None:
+                tensor.grad = Tensor(np.zeros_like(self.data))
+            tensor.grad += grad
+            tensor.backward(grad)
+
 
 class Function:
     def __init__(self, op, *args):
@@ -51,7 +57,7 @@ def forward(x, y):
     @staticmethod
     def backward(ctx, grad):
         x, y = ctx.args
-        return Tensor([1]), Tensor([1])
+        return Tensor([1])*grad, Tensor([1]) *grad
 
 
 class Mul:
@@ -62,28 +68,21 @@ def forward(x, y):
     @staticmethod
     def backward(ctx, grad):
         x, y = ctx.args
-        return Tensor(y.data), Tensor(x.data)  #  dz/dx, dz/dy
+        return Tensor(y.data)*grad, Tensor(x.data)*grad #  dz/dx, dz/dy
+
+
 
 
 if __name__ == "__main__":
-    x = Tensor([8])
-    y = Tensor([5])
 
-    print("Add")
-    z = x + y
-    print(z)
-    z.backward()
-    print(f"x: {x} , grad {x.grad}")
-    print(f"y: {y} , grad {y.grad}")
-    print("="*100)
+    def f(x):
+        return x*x*x + x
 
-    print("Mul")
-    z = x * y
-    print(z)
-    z.backward()
-    print(f"x: {x} , grad {x.grad}")
-    print(f"y: {y} , grad {y.grad}")
-    print("="*100)
+    x = Tensor([1.2])
 
+    z = f(x)
+    z.backward()
+    print(f"X: {x} grad: {x.grad}")
+
 
 

vis/Digraph.gv

@@ -1,9 +1,21 @@
 digraph {
-	139935156534576 [label="Tensor: [13] "]
-	139935156533664 [label="Context: Add"]
-	139935156534576 -> 139935156533664
-	139935156533664 -> 139935160270656
-	139935160270656 [label="Tensor: [8] "]
-	139935156533664 -> 139935156533232
-	139935156533232 [label="Tensor: [5] "]
+	140589163168192 [label="Tensor: ([2.928]  grad: [1.]) "]
+	140589163168528 [label="Context: Add"]
+	140589163168192 -> 140589163168528
+	140589163168528 -> 140589163168480
+	140589163168480 [label="Tensor: ([1.728]  grad: [1.]) "]
+	140589163168912 [label="Context: Mul"]
+	140589163168480 -> 140589163168912
+	140589163168912 -> 140589163167856
+	140589163167856 [label="Tensor: ([1.44]  grad: [1.2]) "]
+	140589163167952 [label="Context: Mul"]
+	140589163167856 -> 140589163167952
+	140589163167952 -> 140589555449664
+	140589555449664 [label="Tensor: ([1.2]  grad: [5.32]) "]
+	140589163167952 -> 140589555449664
+	140589555449664 [label="Tensor: ([1.2]  grad: [5.32]) "]
+	140589163168912 -> 140589555449664
+	140589555449664 [label="Tensor: ([1.2]  grad: [5.32]) "]
+	140589163168528 -> 140589555449664
+	140589555449664 [label="Tensor: ([1.2]  grad: [5.32]) "]
 }

visulize.py

@@ -1,4 +1,5 @@
 import graphviz
+import matplotlib.pyplot as plt
 from tinytorch import *
 
 G = graphviz.Digraph(format="png")
@@ -7,7 +8,7 @@
 
 def visit_nodes(G: graphviz.Digraph, node: Tensor):
     uid = str(id(node))
-    G.node(uid, f"Tensor: {str(node.data) } ")
+    G.node(uid, f"Tensor: ({str(node.data) }  { 'grad: '+str(node.grad.data) if node.grad is not None else ''}) ")
     if node._ctx:
         ctx_uid = str(id(node._ctx))
         G.node(ctx_uid, f"Context: {str(node._ctx.op.__name__)}")
@@ -17,12 +18,51 @@ def visit_nodes(G: graphviz.Digraph, node: Tensor):
             visit_nodes(G, child)
 
 
+
+
+def f(x):
+    return x * x * x + x
+# Defining the function to plot the given function and its derivative using the custom Tensor class
+def plot_function_and_derivative():
+
+
+    # Values for x ranging from -3 to 3
+    x_values_custom = np.linspace(-3, 3, 100)
+    y_values_custom = []
+    derivative_values_custom = []
+
+    # Using the custom Tensor class to calculate the function and its derivative for each x value
+    for x_val in x_values_custom:
+        x_tensor = Tensor([x_val])
+        y_tensor = f(x_tensor)
+        y_tensor.backward()
+        y_values_custom.append(y_tensor.data[0])
+        derivative_values_custom.append(x_tensor.grad.data[0])
+
+    # Plotting the original function and its derivative using the custom implementation
+    plt.plot(x_values_custom, y_values_custom, label="f(x) = x^3 + x (custom)")
+    plt.plot(x_values_custom, derivative_values_custom, label="f'(x) = 3x^2 + 1 (custom)")
+
+    plt.xlabel('x')
+    plt.ylabel('y')
+    plt.title('Plot of the Function and its Derivative (Custom Implementation)')
+    plt.legend()
+    plt.grid(True)
+    plt.show()
+
+
+
+
 if __name__ == "__main__":
-    x = Tensor([8])
-    y = Tensor([5])
-    z = x + y
+    plot_function_and_derivative()
+
+    x = Tensor([1.2])
+    z = f(x)
+    z.backward()
     visit_nodes(G, z)
     G.render(directory="vis", view=True)
-    print(z)
+    print(f"Z:{x} grad:{x.grad}")
 
-    print(len(G.body))
+
+
+