1. Did you load and preprocess the dataset before model training?

✅ Yes
❌ No

Explanation:
Preprocessing prepares data into the correct format and ensures the model trains on clean, consistent inputs.

---

2. Did you split the dataset into training and validation (or test) sets?

✅ Yes
❌ No

Explanation:
Splitting data allows accurate performance evaluation and reduces overfitting risks.

---

3. Did you define your Keras model architecture with at least one hidden layer?

✅ Yes
❌ No

Explanation:
A hidden layer lets the model learn meaningful patterns beyond simple linear relationships.

---

4. Did you compile your model with an appropriate optimizer and loss function?

✅ Yes
❌ No

Explanation:
Using the right optimizer and loss ensures proper gradient updates and stable convergence.

---

5. Did you implement callbacks such as checkpoint_cb or ModelCheckpoint?

✅ Yes
❌ No

Explanation:
Callbacks save the best model, prevent overfitting, and stop training when necessary.

---

6. Did you save and export your trained model (e.g., .model.keras)?

✅ Yes
❌ No

Explanation:
Saving the trained model allows reuse without retraining and supports deployment.

---

7. Did you visualize training and validation loss and accuracy?

✅ Yes
❌ No

Explanation:
Plots reveal training behavior, helping detect underfitting, overfitting, or instability.

---

8. Did you complete all tasks and download the JupyterLab notebook?

✅ Yes
❌ No

Explanation:
Downloading the notebook is required for submission and preserves your full work.

---

🧾 Summary Table

---