Backport post_training CLI import fix to 0.3.x (#306)

## Description

I'm backporting [this
fix](#293)
to `release-0.3.x`, as discussed in
#288.

Pretty much I've cherry-picked the commit from
#293 and
then synced with `uv`.

## Notes to the Reviewers/Maintainers

I see in the docs there are
[guidelines](https://github.com/llamastack/llama-stack-client-python/blob/main/CONTRIBUTING.md#publishing-and-releases)
for releases. However I'd like some guidance if there are any required
steps I should take.

---------

Co-authored-by: Ashwin Bharambe <[email protected]>
Co-authored-by: stainless-app[bot] <142633134+stainless-app[bot]@users.noreply.github.com>

Author: 3 people

src/llama_stack_client/lib/cli/post_training/post_training.py

@@ -9,7 +9,10 @@
 import click
 from rich.console import Console
 
-from llama_stack_client.types.post_training_supervised_fine_tune_params import AlgorithmConfigParam, TrainingConfig
+from llama_stack_client.types.alpha.post_training_supervised_fine_tune_params import (
+    AlgorithmConfigParam,
+    TrainingConfig,
+)
 
 from ..common.utils import handle_client_errors