![François Chollet on Twitter: "4) You can automatically retrieve the gradients of the weights of a layer by calling it inside a GradientTape. Using these gradients, you can update the weights of François Chollet on Twitter: "4) You can automatically retrieve the gradients of the weights of a layer by calling it inside a GradientTape. Using these gradients, you can update the weights of](https://pbs.twimg.com/media/D1Y-tSmUwAAAixK.jpg:large)
François Chollet on Twitter: "4) You can automatically retrieve the gradients of the weights of a layer by calling it inside a GradientTape. Using these gradients, you can update the weights of
François Chollet on Twitter: "4) You can automatically retrieve the gradients of the weights of a layer by calling it inside a GradientTape. Using these gradients, you can update the weights of
![python - Why does my model work with `tf.GradientTape()` but fail when using `keras.models.Model.fit()` - Stack Overflow python - Why does my model work with `tf.GradientTape()` but fail when using `keras.models.Model.fit()` - Stack Overflow](https://i.postimg.cc/brc7LgVK/Gradient-Tape-OK.png)
python - Why does my model work with `tf.GradientTape()` but fail when using `keras.models.Model.fit()` - Stack Overflow
![python - Why does my model work with `tf.GradientTape()` but fail when using `keras.models.Model.fit()` - Stack Overflow python - Why does my model work with `tf.GradientTape()` but fail when using `keras.models.Model.fit()` - Stack Overflow](https://i.stack.imgur.com/MMIQa.png)