deployment section

code/deployment/fastapi/.gitignore

@@ -0,0 +1 @@
+.vercel

code/deployment/fastapi/client.py

@@ -0,0 +1,22 @@
+import requests
+import tensorflow as tf
+from tensorflow import keras
+
+# Import the Fashion MNIST dataset
+fashion_mnist = keras.datasets.fashion_mnist
+(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()
+
+# scale the values to 0.0 to 1.0
+train_images = train_images / 255.0
+test_images = test_images / 255.0
+
+# reshape for feeding into the model
+train_images = train_images.reshape(train_images.shape[0], 28, 28, 1)
+test_images = test_images.reshape(test_images.shape[0], 28, 28, 1)
+
+
+url = "http://localhost:8000"
+req = tf.constant(test_images[:1], dtype=tf.float32)
+
+resp = requests.post(f"{url}/predict", json={"instances": req.numpy().tolist()}).json()
+print(resp)

code/deployment/fastapi/main.py

@@ -0,0 +1,49 @@
+import numpy as np
+import tensorflow as tf
+from fastapi import FastAPI
+from fastapi.responses import JSONResponse
+from pydantic import BaseModel
+
+class_names = [
+    "T-shirt/top",
+    "Trouser",
+    "Pullover",
+    "Dress",
+    "Coat",
+    "Sandal",
+    "Shirt",
+    "Sneaker",
+    "Bag",
+    "Ankle boot",
+]
+
+
+class PredictInput(BaseModel):
+    instances: list
+
+
+app = FastAPI()
+
+model = tf.keras.models.load_model("./model/1.keras")
+
+
+@app.get("/")
+def hello_world():
+    return "Hello, world!"
+
+
+@app.post("/predict")
+def predict(input: PredictInput):
+    instances = tf.constant(input.instances, dtype=tf.float32)
+    predictions = model(instances)
+    response = {
+        "predictions": predictions.numpy().tolist(),
+        "message": f"The model thought this was a {class_names[np.argmax(predictions)]} (class {np.argmax(predictions)})",
+    }
+    return JSONResponse(content=response)
+
+
+if __name__ == "__main__":
+    import uvicorn
+
+    uvicorn.run(app)

code/deployment/fastapi/requirements.txt

@@ -0,0 +1,51 @@
+absl-py==2.1.0
+annotated-types==0.7.0
+anyio==4.4.0
+astunparse==1.6.3
+certifi==2024.8.30
+charset-normalizer==3.3.2
+click==8.1.7
+fastapi==0.112.2
+flatbuffers==24.3.25
+gast==0.6.0
+google-pasta==0.2.0
+grpcio==1.66.1
+h11==0.14.0
+h5py==3.11.0
+idna==3.8
+keras==3.5.0
+libclang==18.1.1
+Markdown==3.7
+markdown-it-py==3.0.0
+MarkupSafe==2.1.5
+mdurl==0.1.2
+ml-dtypes==0.4.0
+namex==0.0.8
+numpy==1.26.4
+opt-einsum==3.3.0
+optree==0.12.1
+packaging==24.1
+pandas==2.2.2
+protobuf==4.25.4
+pydantic==2.8.2
+pydantic_core==2.20.1
+Pygments==2.18.0
+python-dateutil==2.9.0.post0
+pytz==2024.1
+requests==2.32.3
+rich==13.8.0
+setuptools==74.0.0
+six==1.16.0
+sniffio==1.3.1
+starlette==0.38.2
+tensorboard==2.17.1
+tensorboard-data-server==0.7.2
+tensorflow==2.17.0
+termcolor==2.4.0
+typing_extensions==4.12.2
+tzdata==2024.1
+urllib3==2.2.2
+uvicorn==0.30.6
+Werkzeug==3.0.4
+wheel==0.44.0
+wrapt==1.16.0

code/deployment/fastapi/train.py

@@ -0,0 +1,89 @@
+import os
+
+import tensorflow as tf
+from tensorflow import keras
+
+print("TensorFlow version: {}".format(tf.__version__))
+
+# Import the Fashion MNIST dataset
+fashion_mnist = keras.datasets.fashion_mnist
+(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()
+
+# scale the values to 0.0 to 1.0
+train_images = train_images / 255.0
+test_images = test_images / 255.0
+
+# reshape for feeding into the model
+train_images = train_images.reshape(train_images.shape[0], 28, 28, 1)
+test_images = test_images.reshape(test_images.shape[0], 28, 28, 1)
+
+class_names = [
+    "T-shirt/top",
+    "Trouser",
+    "Pullover",
+    "Dress",
+    "Coat",
+    "Sandal",
+    "Shirt",
+    "Sneaker",
+    "Bag",
+    "Ankle boot",
+]
+
+print("train_images.shape: {}, of {}".format(train_images.shape, train_images.dtype))
+print("test_images.shape: {}, of {}".format(test_images.shape, test_images.dtype))
+
+# Train and evaluate the model
+model = keras.Sequential(
+    [
+        keras.layers.Conv2D(
+            input_shape=(28, 28, 1),
+            filters=8,
+            kernel_size=3,
+            strides=2,
+            activation="relu",
+            name="Conv1",
+        ),
+        keras.layers.Flatten(),
+        keras.layers.Dense(10, name="Dense"),
+    ]
+)
+model.summary()
+
+testing = False
+epochs = 5
+
+model.compile(
+    optimizer="adam",
+    loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
+    metrics=[keras.metrics.SparseCategoricalAccuracy()],
+)
+model.fit(train_images, train_labels, epochs=epochs)
+print("Model trained!")
+
+# Fetch the Keras session and save the model
+MODEL_DIR = "model"
+version = 1
+model_format = ".keras"
+model_path = os.path.join(MODEL_DIR, str(version) + model_format)
+print("model_path = {}\n".format(model_path))
+
+tf.keras.models.save_model(
+    model,
+    model_path,
+    overwrite=True,
+    include_optimizer=True,
+    save_format=None,
+)
+print("Model saved!")
+
+# Test saved model
+model = tf.keras.models.load_model(model_path)
+print(model.summary())
+
+sample_input = tf.constant(test_images[:1], dtype=tf.float32)
+output = model(sample_input)
+print(output)
+
+test_loss, test_acc = model.evaluate(test_images, test_labels)
+print("\nTest accuracy: {}".format(test_acc))

code/deployment/fastapi/vercel.json

@@ -0,0 +1,14 @@
+{
+  "builds": [
+    {
+      "src": "main.py",
+      "use": "@vercel/python"
+    }
+  ],
+  "routes": [
+    {
+      "src": "/(.*)",
+      "dest": "main.py"
+    }
+  ]
+}

code/deployment/fasthtml/.gitignore

@@ -0,0 +1,2 @@
+.vercel
+.sesskey

code/deployment/fasthtml/README.md

@@ -0,0 +1,40 @@
+# FastHTML Boilerplate
+
+Deploy your [FastHTML](https://fastht.ml/) project to Vercel with zero configuration.
+
+[![Deploy with Vercel](https://vercel.com/button)](https://vercel.com/new/clone?repository-url=https://github.com/vercel/vercel/tree/main/examples/fasthtml&template=fasthtml)
+
+_Live Example: https://fasthtml-template.vercel.app_
+
+Visit the [FastHTML documentation](https://docs.fastht.ml/) to learn more.
+
+## Getting Started
+
+Install the required dependencies:
+
+```bash
+pip install -r requirements.txt
+```
+
+## Running Locally
+
+Start the development server on http://0.0.0.0:5001
+
+```bash
+python main.py
+```
+
+When you make changes to your project, the server will automatically reload.
+
+## Deploying to Vercel
+
+Deploy your project to Vercel with the following command:
+
+```bash
+npm install -g vercel
+vercel --prod
+```
+
+Or `git push` to your repostory with our [git integration](https://vercel.com/docs/deployments/git).
+
+To view the source code for this template, [visit the example repository](https://github.com/vercel/vercel/tree/main/examples/fasthtml).

code/deployment/fasthtml/main.py

@@ -0,0 +1,85 @@
+from os import getenv
+
+from openai import OpenAI
+
+from fasthtml.common import *
+
+# Set up the app, including daisyui and tailwind for the chat component
+hdrs = (
+    picolink,
+    Script(src="https://cdn.tailwindcss.com"),
+    Link(
+        rel="stylesheet",
+        href="https://cdn.jsdelivr.net/npm/daisyui@4.11.1/dist/full.min.css",
+    ),
+)
+app = FastHTML(hdrs=hdrs, cls="p-4 max-w-lg mx-auto")
+
+# Set up a chat model (https://claudette.answer.ai/)
+client = OpenAI(
+    base_url="https://openrouter.ai/api/v1",
+    api_key=getenv("OPENROUTER_API_KEY"),
+)
+model = "mistralai/mistral-7b-instruct:free"
+
+
+# Chat message component (renders a chat bubble)
+def ChatMessage(msg, user):
+    bubble_class = "chat-bubble-primary" if user else "chat-bubble-secondary"
+    chat_class = "chat-end" if user else "chat-start"
+    return Div(cls=f"chat {chat_class}")(
+        Div("user" if user else "assistant", cls="chat-header"),
+        Div(msg, cls=f"chat-bubble {bubble_class}"),
+        Hidden(msg, name="messages"),
+    )
+
+
+# The input field for the user message. Also used to clear the
+# input field after sending a message via an OOB swap
+def ChatInput():
+    return Input(
+        name="msg",
+        id="msg-input",
+        placeholder="Type a message",
+        cls="input input-bordered w-full",
+        hx_swap_oob="true",
+    )
+
+
+# The main screen
+@app.get
+def index():
+    page = Form(hx_post=send, hx_target="#chatlist", hx_swap="beforeend")(
+        Div(id="chatlist", cls="chat-box h-[73vh] overflow-y-auto"),
+        Div(cls="flex space-x-2 mt-2")(
+            Group(ChatInput(), Button("Send", cls="btn btn-primary"))
+        ),
+    )
+    return Titled("Chatbot Demo", page)
+
+
+# Handle the form submission
+@app.post
+def send(msg: str, messages: list[str] = None):
+    if not messages:
+        messages = []
+    messages.append(msg.rstrip())
+    completion = client.chat.completions.create(
+        model=model,
+        messages=[
+            {
+                "role": "user",
+                "content": msg.rstrip(),
+            },
+        ],
+    )
+    return (
+        ChatMessage(msg, True),  # The user's message
+        ChatMessage(
+            completion.choices[0].message.content.rstrip(), False
+        ),  # The chatbot's response
+        ChatInput(),
+    )  # And clear the input field via an OOB swap
+
+
+serve()

code/deployment/fasthtml/requirements.txt

@@ -0,0 +1,3 @@
+openai==1.42.0
+python-fasthtml==0.4.4
+uvicorn==0.30.1

docs/_toc.yml

@@ -119,6 +119,12 @@ parts:
       - file: gen_ai/intro_to_llm
       - file: gen_ai/local_llms
 
+  - caption: Deployment
+    chapters:
+      - file: deployment/intro_to_deployment
+      - file: deployment/deploy_python_apps
+      - file: deployment/deploy_ml_apps
+
   - caption: Reference
     chapters:
       - file: reference/ai_domains