Compiling models via the recommended entry points described in
[XLA](https://www.tensorflow.org/xla) and
[JAX](https://jax.readthedocs.io/en/latest/jax-101/02-jitting.html)
documentation should be safe, while some of the testing and debugging tools that
come with the compiler are not designed to be used with untrusted data and
should be used with caution when working with untrusted models.

### Saved graphs and checkpoints

## Platform architecture

Gradle is arranged into several coarse-grained components called "platforms".
Each platform provides support for some kind of automation, such as building JVM software or building Gradle plugins.
Most platforms typically build on the features of other platforms.

Not every module contributes to every runtime.

The core-runtime module defines each runtime:

- The target JVM for the runtime. Each runtime has its own JVM compatibility constraints.
- Some base services that are available to code hosted by the runtime. This varies by runtime.

- Erasure coding used by MinIO is [Reed-Solomon](https://github.com/klauspost/reedsolomon) erasure coding scheme, which has a total shard maximum of 256 i.e 128 data and 128 parity. MinIO design goes beyond this limitation by doing some practical architecture choices.

=== "Linux, macOS, Windows Bash"

    <div class="termy">

    ```console
    $ which pip

    some/directory/fastapi/env/bin/pip
    ```

    </div>

=== "Windows PowerShell"

    <div class="termy">

    ```console
    $ Get-Command pip

    some/directory/fastapi/env/bin/pip
    ```

    </div>

If it shows the `pip` binary at `env/bin/pip` then it worked. 🎉

# To simplify the automations section, some calculations are placed under unique YAML keys defined here.
# Read the "|" not as "or", but as a "pipe", taking the output of the previous command and passing it to the next command.
# This section could also appear ahead of the automations section.

colors:
  yellow: 'ffb300'

includes_src_changes: {{ files | match(regex=r/.*\/src\//) | some }}

## Bodies of arbitrary `dict`s

You can also declare a body as a `dict` with keys of some type and values of some other type.

This way, you don't have to know beforehand what the valid field/attribute names are (as would be the case with Pydantic models).

# JSON Compatible Encoder

There are some cases where you might need to convert a data type (like a Pydantic model) to something compatible with JSON (like a `dict`, `list`, etc).

For example, if you need to store it in a database.

For that, **FastAPI** provides a `jsonable_encoder()` function.

## Using the `jsonable_encoder`

Let's imagine that you have a database `fake_db` that only receives JSON compatible data.

So, in the documentation for third-party ASGI middlewares they will probably tell you to do something like:

```Python
from unicorn import UnicornMiddleware

app = SomeASGIApp()

new_app = UnicornMiddleware(app, some_config="rainbow")
```

But FastAPI (actually Starlette) provides a simpler way to do it that makes sure that the internal middlewares to handle server errors and custom exception handlers work properly.

    should not use them for communication between trusted and untrusted code.

6.  For the mainline flavor, we test the libraries using OpenJDK 8, 11, and 17
    on Linux, with some additional testing on newer JDKs and on Windows. Some
    features, especially in `com.google.common.io`, may not work correctly in
    non-Linux environments. For the Android flavor, our unit tests also run on
    API level 21 (Lollipop).