05 - Türkiye

[erwangranger]

This is the thread where you can paste your notes based on the mission presented.

[syntaxsdev]

Experienced this issue:

AttributeError                            Traceback (most recent call last)
Cell In[23], line 9
      6 Path(artifact_path).mkdir(parents=True, exist_ok=True)
      8 # Save the model to ONNX format
----> 9 onnx_model, _ = tf2onnx.convert.from_keras(
     10     model, 
     11     input_signature=[tf.TensorSpec([None, X_train_scaled.shape[1]], tf.float32, name='input')]
     12 )
     13 onnx.save(onnx_model, f"models/{version}/baklava/1/model.onnx")
     15 # Save the scaler

File /opt/app-root/lib64/python3.9/site-packages/tf2onnx/convert.py:442, in from_keras(model, input_signature, opset, custom_ops, custom_op_handlers, custom_rewriter, inputs_as_nchw, outputs_as_nchw, extra_opset, shape_override, target, large_model, output_path, optimizers)
    438 if get_tf_version() < Version("2.0"):
    439     return _from_keras_tf1(model, opset, custom_ops, custom_op_handlers, custom_rewriter, inputs_as_nchw,
    440                            outputs_as_nchw, extra_opset, shape_override, target, large_model, output_path)
--> 442 old_out_names = _rename_duplicate_keras_model_names(model)
    443 from tensorflow.python.keras.saving import saving_utils as _saving_utils # pylint: disable=import-outside-toplevel
    445 # let tensorflow do the checking if model is a valid model

File /opt/app-root/lib64/python3.9/site-packages/tf2onnx/convert.py:331, in _rename_duplicate_keras_model_names(model)
    325 """
    326 In very rare cases, keras has a bug where it will give multiple outputs the same name.
    327 We must edit the model or the TF trace will fail. Returns old_out_names (or None if no edit was made).
    328 IMPORTANT: model may be edited. Assign model.output_names to old_out_names to restore.
    329 """
    330 old_out_names = None
--> 331 if model.output_names and len(set(model.output_names)) != len(model.output_names):
    332     # In very rare cases, keras has a bug where it will give multiple outputs the same name
    333     # We must edit the model or the TF trace will fail
    334     old_out_names = model.output_names
    335     used_names = set()

AttributeError: 'Sequential' object has no attribute 'output_names'```