- Classiq is now supported on Google Colab.
- Change the default authentication URL from classiq.eu.auth0.com to auth.classiq.io. The Classiq IDE and Python SDK update automatically.
- Register names of arithmetic operations are no longer customizable. See Arithmetic and sub-pages for the up-to-date interfaces.
- The default URL for the backend is changed from classiquantum.com to api.classiq.io. The Classiq IDE and Python SDK update automatically.
- Remove the
StatePreparation. See StatePreparation.
- Remove "dirty" arithmetic uncomputation options. See Arithmetic Expressions.
Synthesis and execution flow has been updated to work on textual inputs.
get_modelmethod as been added to
Modelin order to obtain the textual serialized underlying model.
- Model synthesis is now done by
serialized_quantum_program = synthesize(serialized_model)instead of
model.synthesize(). See synthesis.
- Execution of the quantum algorithms is now done by
execute(serialized_quantum_program). See execution
- The execution flow is now incorporated as part of the model.
In order to add it to the
Modelinstance use the
- The execution flow is now incorporated as part of the model. In order to add it to the
- Setting additional preferences and constraints for the synthesis and execution engines is done by the dedicated functions (that work on the pre synthesis serialized model):
set_preferences(serialized_model, preferences)for setting the synthesis preferences.
set_constraints(serialized_model, constraints)for setting the synthesis constraints.
set_execution_preferences(serialized_model, execution_preferences)for setting the execution preferences.
Application suites have been changed to work with the new flow by adding specific model constructor function for each application.
construct_chemistry_modelfunction creates a model for solving ground state problems with VQE.
construct_combinatorial_optimization_modelfunction creates a model for solving combinatorial optimization problems with QAOA.
construct_grover_modelfunction creates a model for using the "Grover Search" algorithm.
construct_qsvm_modelfunction creates a model for running the QSVM algorithm.
construct_finance_modelfunction creates a model for solving finance problems.
- Multiple Execution: Users can select up to five hardware backends for simultaneous execution.
- Generated model in the backend for VQE, finance and Grover: A “generate model” button has been added which generates the model based on the form values in the IDE.
- Modifications to the home screen: fast execution page, registration page.
- Modification to application nav bar: Added “learn” section with links to tutorial and Python SDK installation guide.
- The results page now supports multiple results per run.
- The execution page was simplified. There are now much fewer execution options on the page, and they are instead handled automatically via model options.
- The Examples menu changed location and its interface and will be extended soon.
- Hardware providers can now be selected from a dropdown list on the synthesis page.
- Output formats selected from the dropdown in the Synthesis page can now be exported via the Circuit page.
- Correctly assign qubits for sliced inputs in the main function.
- Introduce new amplitude loading functions. See PiecewiseLinearRotationAmplitudeLoading and PiecewiseLinearAmplitudeLoading.
- Add introduction tutorials. See tutorials.
- Deprecate light mode.
- Add auto completion when searching. Use the right arrow key to accept the suggestion.
- Highlight the searched keywords when entering a page using the search widget.