rasa init doesn't work

Rasa Open Source version

2.8.6

Rasa SDK version

2.8.2

Rasa X version

None

Python version

3.8

What operating system are you using?

Cloud (AWS / GCP / AZURE)

What happened?

Tries rasa init on a newly created GCP instance (Ubuntu 20.04) and tried to train. An error occurred. My command history:

   1  sudo apt update
   2  sudo apt install python3-dev python3-pip
   3  python3 -m venv ./venv
   4  pip3 install -U pip
   5  python3 -m venv ./venv
   6  sudo apt install python3.8-venv
   7  python3 -m venv ./venv
   8  source ./venv/bin/activate
   9  pip3 install rasa
  10  mkdir bot
  11  cd bot
  12  rasa --version
  13  rasa init

Command / Request

rasa init

Relevant log output

qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqk
x Rasa Open Source reports anonymous usage telemetry to help improve the product x
x for all its users.                                                             x
x                                                                                x
x If you'd like to opt-out, you can use `rasa telemetry disable`.                x
x To learn more, check out https://rasa.com/docs/rasa/telemetry/telemetry.       x
mqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqj
Welcome to Rasa! 🤖

To get started quickly, an initial project will be created.
If you need some help, check out the documentation at https://rasa.com/docs/rasa.
Now let's start! 👇🏽

? Please enter a path where the project will be created [default: current directory]                              
Created project directory at '/home/rumesh_m_s16/bot'.
Finished creating project structure.
? Do you want to train an initial model? 💪🏽 Yes                                                                 
Training an initial model...
The configuration for pipeline and policies was chosen automatically. It was written into the config file at './config.yml'.
Training NLU model...
/home/rumesh_m_s16/venv/lib/python3.8/site-packages/tensorflow_addons/utils/ensure_tf_install.py:53: UserWarning: Tensorflow Addons supports using Python ops for all Tensorflow versions above or equal to 2.3.0 and strictly below 2.6.0 (nightly versions are not supported). 
 The versions of TensorFlow you are currently using is 2.6.0 and is not supported. 
Some things might work, some things might not.
If you were to encounter a bug, do not file an issue.
If you want to make sure you're using a tested and supported configuration, either change the TensorFlow version or the TensorFlow Addons's version. 
You can find the compatibility matrix in TensorFlow Addon's readme:
https://github.com/tensorflow/addons
  warnings.warn(
Traceback (most recent call last):
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/utils/common.py", line 296, in run_in_loop
    result = loop.run_until_complete(f)
  File "uvloop/loop.pyx", line 1456, in uvloop.loop.Loop.run_until_complete
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/model_training.py", line 108, in train_async
    return await _train_async_internal(
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/model_training.py", line 288, in _train_async_internal
    await _do_training(
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/model_training.py", line 334, in _do_training
    model_path = await _train_nlu_with_validated_data(
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/model_training.py", line 758, in _train_nlu_with_validated_data
    await rasa.nlu.train.train(
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/nlu/train.py", line 96, in train
    trainer = Trainer(
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/nlu/model.py", line 164, in __init__
    components.validate_requirements(cfg.component_names)
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/nlu/components.py", line 58, in validate_requirements
    from rasa.nlu import registry
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/nlu/registry.py", line 12, in <module>
    from rasa.nlu.classifiers.diet_classifier import DIETClassifier
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/nlu/classifiers/diet_classifier.py", line 24, in <module>
    from rasa.utils.tensorflow.models import RasaModel, TransformerRasaModel
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/utils/tensorflow/models.py", line 47, in <module>
    from rasa.utils.tensorflow.temp_keras_modules import TmpKerasModel
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/utils/tensorflow/temp_keras_modules.py", line 26, in <module>
    class TmpKerasModel(tf.keras.models.Model):
  File "/home/rumesh_m_s16/venv/lib/python3.8/site-packages/rasa/utils/tensorflow/temp_keras_modules.py", line 37, in TmpKerasModel
    @training.enable_multi_worker
AttributeError: module 'tensorflow.python.keras.engine.training' has no attribute 'enable_multi_worker'

[kalkbrennerei]

Hey @rumeshmadhusanka thanks for letting us know! 💯 Since we don’t support tensorflow 2.6 yet, pip3 install rasa should install a lower version of tensorflow. We have a PR up to fix this!

You can probably also manually install tensorflow 2.3.4 and it should work

[kalkbrennerei]

Hi! On my end if I try to run pip install rasa==2.8.9 in a freshly created virtual environment it installs the specified rasa version as it should (pip version 21.3). Have you followed the installation instructions from here?