Step NumPy pin to 1.22 in conda build config

[antonwolfy]

Intel 2023.0 release includes NumPy 1.22.
Thus dpctl should switch pinning NumPy version to the same value in config file.

Otherwise it is impossible to correctly build dpctl, when there is no new 2023 compiler in the host environment. This is because the current NumPy 1.21 can't be installed with the new 2023 compiler.
As a result it leads to improper linkage of libDPCTLSyclInterface.so with libsycl.so.6, which came with 2023 compiler only and present in build env, but missing in the host env.

• Have you provided a meaningful PR description?
• Have you added a test, reproducer or referred to an issue with a reproducer?
• Have you tested your changes locally for CPU and GPU devices?
• Have you made sure that new changes do not introduce compiler warnings?
• If this PR is a work in progress, are you filing the PR as a draft?

[antonwolfy]

A verification job array-api-conformity seems failed, because the PR is from forked repository and entire step couldn't post the result as a PR comment.
While the result of conformance tests looks OK (the same as in previous runs):

    MESSAGE: Array API standard conformance tests for dpctl=0.14.1dev0=py310h76be34b_24 ran successfully.
    Passed: 33
    Failed: 801
    Skipped: 280

[github-actions]

Array API standard conformance tests for dpctl=0.14.1dev0=py310h76be34b_24 ran successfully.
Passed: 33
Failed: 801
Skipped: 280