Qibojit

What's Changed
* Support sparse matrices in cast methods by stavros11 in https://github.com/qiboteam/qibojit/pull/68
* Fix `expm` for sparse matrices on GPU by stavros11 in https://github.com/qiboteam/qibojit/pull/75
* Change `.flatten()` to `.ravel()` to avoid copies by stavros11 in https://github.com/qiboteam/qibojit/pull/74
* Change `.ravel()` back to `.flatten()` for gates by stavros11 in https://github.com/qiboteam/qibojit/pull/79
* New swap implementation for cuQuantum by andrea-pasquale in https://github.com/qiboteam/qibojit/pull/80
* Update cuQuantum version by andrea-pasquale in https://github.com/qiboteam/qibojit/pull/77
* Implement methods for hamiltonians abstraction by andrea-pasquale in https://github.com/qiboteam/qibojit/pull/82
* Use new backend abstractions by stavros11 in https://github.com/qiboteam/qibojit/pull/81
* Fix cuquantum tests by stavros11 in https://github.com/qiboteam/qibojit/pull/83


**Full Changelog**: https://github.com/qiboteam/qibojit/compare/v0.0.4...v0.0.5-rc0

What's Changed
* removing requirements by scarrazza in https://github.com/qiboteam/qibojit/pull/24
* post-install compilation by scarrazza in https://github.com/qiboteam/qibojit/pull/30
* Adding signature with arrays by andrea-pasquale in https://github.com/qiboteam/qibojit/pull/28
* adding scipy dependency by scarrazza in https://github.com/qiboteam/qibojit/pull/32
* Fix template kernels with ROCm by mlazzarin in https://github.com/qiboteam/qibojit/pull/33
* Fix tests and methods with AMD ROCM by mlazzarin in https://github.com/qiboteam/qibojit/pull/36
* Use template kernels for multiqubit gates by mlazzarin in https://github.com/qiboteam/qibojit/pull/37
* Add backend class from qibo by stavros11 in https://github.com/qiboteam/qibojit/pull/35
* Multi-qubit kernel for GPU by stavros11 in https://github.com/qiboteam/qibojit/pull/23
* Multiqubit kernel template by stavros11 in https://github.com/qiboteam/qibojit/pull/21
* using qibo --pre by scarrazza in https://github.com/qiboteam/qibojit/pull/39
* adding minimum qibo version by scarrazza in https://github.com/qiboteam/qibojit/pull/40
* Fix multiqubit GPU ops not called by JitCustomBackend by mlazzarin in https://github.com/qiboteam/qibojit/pull/41
* Fix ``CupyBackend`` crash with 32 qubits by mlazzarin in https://github.com/qiboteam/qibojit/pull/43
* Fix CPU fallback by mlazzarin in https://github.com/qiboteam/qibojit/pull/48
* Add missing stream synchronizations by mlazzarin in https://github.com/qiboteam/qibojit/pull/50
* Implement a custom CPU fallback for qibojit by mlazzarin in https://github.com/qiboteam/qibojit/pull/54
* Fix cast method not changing dtype by mlazzarin in https://github.com/qiboteam/qibojit/pull/55
* removing py3.6 action and requirements by scarrazza in https://github.com/qiboteam/qibojit/pull/56
* Test `one_qubit_base` and `two_qubit_base` directly by stavros11 in https://github.com/qiboteam/qibojit/pull/58
* Adding CuQuantum library by andrea-pasquale in https://github.com/qiboteam/qibojit/pull/57
* cupy/cuquantum platform switcher by stavros11 in https://github.com/qiboteam/qibojit/pull/59
* using qibo master code by scarrazza in https://github.com/qiboteam/qibojit/pull/60
* Use RawKernel without name_expressions to enable CuPy cache by mlazzarin in https://github.com/qiboteam/qibojit/pull/45
* Force C order casting for GPU by stavros11 in https://github.com/qiboteam/qibojit/pull/64
* adding info message concerning compilation by scarrazza in https://github.com/qiboteam/qibojit/pull/66
* Move `supports_multigpu` flag to platforms by stavros11 in https://github.com/qiboteam/qibojit/pull/67
* Update qubits type for cuquantum by stavros11 in https://github.com/qiboteam/qibojit/pull/70
* multi-gpu cupy memory grow fix attempt by scarrazza in https://github.com/qiboteam/qibojit/pull/69
* adding python 3.10 support to CI by scarrazza in https://github.com/qiboteam/qibojit/pull/71
* Improving CuQuantum Implementation by andrea-pasquale in https://github.com/qiboteam/qibojit/pull/72

New Contributors
* andrea-pasquale made their first contribution in https://github.com/qiboteam/qibojit/pull/28
* mlazzarin made their first contribution in https://github.com/qiboteam/qibojit/pull/33

**Full Changelog**: https://github.com/qiboteam/qibojit/compare/v0.0.3...v0.0.4

What's Changed
* removing requirements by scarrazza in https://github.com/qiboteam/qibojit/pull/24
* post-install compilation by scarrazza in https://github.com/qiboteam/qibojit/pull/30
* Adding signature with arrays by andrea-pasquale in https://github.com/qiboteam/qibojit/pull/28
* adding scipy dependency by scarrazza in https://github.com/qiboteam/qibojit/pull/32
* Fix template kernels with ROCm by mlazzarin in https://github.com/qiboteam/qibojit/pull/33
* Fix tests and methods with AMD ROCM by mlazzarin in https://github.com/qiboteam/qibojit/pull/36
* Use template kernels for multiqubit gates by mlazzarin in https://github.com/qiboteam/qibojit/pull/37
* Add backend class from qibo by stavros11 in https://github.com/qiboteam/qibojit/pull/35
* Multi-qubit kernel for GPU by stavros11 in https://github.com/qiboteam/qibojit/pull/23
* Multiqubit kernel template by stavros11 in https://github.com/qiboteam/qibojit/pull/21

New Contributors
* andrea-pasquale made their first contribution in https://github.com/qiboteam/qibojit/pull/28
* mlazzarin made their first contribution in https://github.com/qiboteam/qibojit/pull/33

**Full Changelog**: https://github.com/qiboteam/qibojit/compare/v0.0.3...v0.0.4-rc0

What's Changed
* Add ops required by qibo multi-GPU (19) stavros11
* removing inline (18) scarrazza
* adding numba minimum version (17) scarrazza

Including kernels to wheels for CuPy support.

What's Changed
* Extending CI (14) scarrazza
* adding rocm support (12) scarrazza
* fixing thread seed generation (13) scarrazza
* Fixes for Qibo integration (8) stavros11
* Numba initialization using numpy (11) stavros11
* GPU custom operator for initial state (10) stavros11
* Setup (1) scarrazza
* Add numba CPU custom operators and tests (2) stavros11
* Cupy custom operators for GPU (6) stavros11
* fixing CI (7) scarrazza
* Benchmark script for backend comparison (3) stavros11
* measure frequencies with pure numba (4) scarrazza
* CI setup (5) scarrazza