// Starting with an 8k buffer, double the size of each successive buffer. Smaller buffers
    // quadruple in size until they reach 8k, to minimize the number of small reads for longer
    // streams. Buffers are retained in a deque so that there's no copying between buffers while
    // reading and so all of the bytes in each new allocated buffer are available for reading from
    // the stream.

*   Update `tf.keras` to the Keras 2.1.6 API.
*   Added
    [`tf.keras.layers.CuDNNGRU`](https://www.tensorflow.org/versions/r1.9/api_docs/python/tf/keras/layers/CuDNNGRU)
    and
    [`tf.keras.layers.CuDNNLSTM`](https://www.tensorflow.org/versions/r1.9/api_docs/python/tf/keras/layers/CuDNNLSTM)
    layers.

burgers = get_burgers(2)
```

---

, 🚥 👆 ⚙️ 🗃 👈 💬 👆 👈 👆 💪 🤙 ⚫️ ⏮️ `await`, 👆 💪 ✍ *➡ 🛠️ 🔢* 👈 ⚙️ ⚫️ ⏮️ `async def`, 💖:

```Python hl_lines="2-3"
@app.get('/burgers')
async def read_burgers():
    burgers = await get_burgers(2)
    return burgers
```

### 🌅 📡 ℹ

👆 💪 ✔️ 👀 👈 `await` 💪 🕴 ⚙️ 🔘 🔢 🔬 ⏮️ `async def`.

                _src = _src.deferred;
                int _name_buffers = _src.dec_ndr_long();
                _src.dec_ndr_long();
                int _name_bufferl = _src.dec_ndr_long();
                int _name_bufferi = _src.index;
                _src.advance(2 * _name_bufferl);

                if (name.buffer == null) {
                    if (_name_buffers < 0 || _name_buffers > 0xFFFF) throw new NdrException( NdrException.INVALID_CONFORMANCE );

                int _name_buffers = _src.dec_ndr_long();
                _src.dec_ndr_long();
                int _name_bufferl = _src.dec_ndr_long();
                int _name_bufferi = _src.index;
                _src.advance(2 * _name_bufferl);

                if ( this.name.buffer == null ) {
                    if ( _name_buffers < 0 || _name_buffers > 0xFFFF )

// to the default size.
// Calling b.Reset(b) (that is, resetting a [Reader] to itself) does nothing.
func (b *Reader) Reset(r io.Reader) {
	// If a Reader r is passed to NewReader, NewReader will return r.
	// Different layers of code may do that, and then later pass r
	// to Reset. Avoid infinite recursion in that case.
	if b == r {
		return
	}
	if b.buf == nil {
		b.buf = make([]byte, defaultBufSize)
	}
	b.reset(b.buf, r)
}

<img src="/img/async/concurrent-burgers/concurrent-burgers-01.png" class="illustration">

Dann sind Sie an der Reihe und Sie bestellen zwei sehr schmackhafte Burger für Ihren Schwarm und Sie. 🍔🍔

<img src="/img/async/concurrent-burgers/concurrent-burgers-02.png" class="illustration">

```Python hl_lines="1"
async def get_burgers(number: int):
    # Do some asynchronous stuff to create the burgers
    return burgers
```

...en vez de `def`:

```Python hl_lines="2"
# This is not asynchronous
def get_sequential_burgers(number: int):
    # Do some sequential stuff to create the burgers
    return burgers
```

你和你的恋人一起去快餐店，你排队在后面，收银员从你前面的人接单。😍

<img src="/img/async/concurrent-burgers/concurrent-burgers-01.png" class="illustration">

然后轮到你了，你为你的恋人和你选了两个非常豪华的汉堡。🍔🍔

<img src="/img/async/concurrent-burgers/concurrent-burgers-02.png" class="illustration">

收银员对厨房里的厨师说了一些话，让他们知道他们必须为你准备汉堡（尽管他们目前正在为之前的顾客准备汉堡）。

<img src="/img/async/concurrent-burgers/concurrent-burgers-03.png" class="illustration">

你付钱了。 💸

收银员给你轮到的号码。

#### Burgers parallèles

Imaginons désormais que ce ne sont pas des "burgers concurrents" mais des "burgers parallèles".