# No module named tf
**URL:**
**Author:**  [@eskayML](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2FeskayML)%5C
**Post date:** [October 9, 2022, 7:32am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fno-module-named-tf%2F24173%2F1 "2022-10-09T07:32:59Z")
Please I need help
My app [Here](https://huggingface.co/spaces/eskayML/cat-and-dog-classifier) is giving an error of no module named tf which I quite weird since I already installed tensorflow
---
**Author:**  [@nightfury](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fnightfury)%5C
**Post date:** [October 10, 2022, 7:52am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fno-module-named-tf%2F24173%2F2 "2022-10-10T07:52:37Z")
ValueError: Unknown layer: Functional
```auto
File "/home/user/app/app.py", line 31, in
model = load_model('cat_and_dog_classifier.h5')File "/home/user/.local/lib/python3.8/site-packages/tensorflow/python/keras/saving/save.py", line 184, in load_model
return hdf5_format.load_model_from_hdf5(filepath, custom_objects, compile)
```
the reason for this error is cause of use of different versions of Keras & TensorFlow… like in for training the model the version might have ‘2.3.0’ for Tensorflow & ‘2.4.3’ for Keras and now while accessing the saved model(.h5) are using another version of Keras & TensorFlow.
So for the solution it is best to re-trained model from scratch with upgraded versions or downgrades your TF&Keras to the same version as on which model is trained.
---
**Author:**  [@nightfury](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fnightfury)%5C
**Post date:** [October 10, 2022, 7:56am UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fno-module-named-tf%2F24173%2F3 "2022-10-10T07:56:34Z")
You may even use following code if you are using older TF and trying to unpack h5 file packed by newer TF
```auto
your_model = tf.keras.models.load_model(
path_to_h5,
custom_objects={'Functional':tf.keras.models.Model})
```
Follow below discussion for more details (hope it would be helpful):
>