Improve KerasHub doc content

keras-team · Nov 25, 2024 · 68fdde7 · 68fdde7
1 parent 378a0cb
commit 68fdde7
Show file tree

Hide file tree

Showing 13 changed files with 106 additions and 67 deletions.
diff --git a/scripts/hub_master.py b/scripts/hub_master.py
@@ -1,6 +1,6 @@
 BASE_CLASSES = {
     "path": "base_classes/",
-    "title": "Model Base Classes",
+    "title": "Modeling API",
     "toc": True,
     "children": [
         {
@@ -209,7 +209,7 @@
 
 MODELS_MASTER = {
     "path": "models/",
-    "title": "Pretrained Models",
+    "title": "Model Architectures",
     "toc": True,
     "children": [
         {
@@ -491,7 +491,7 @@
         },
         {
             "path": "deeplab_v3/",
-            "title": "DeepLabV3 and DeepLabV3Plus",
+            "title": "DeepLabV3",
             "toc": True,
             "children": [
                 {
@@ -1725,12 +1725,18 @@
         {
             "path": "audio_converter",
             "title": "AudioConverter layer",
-            "generate": ["keras_hub.layers.AudioConverter"],
+            "generate": [
+                "keras_hub.layers.AudioConverter",
+                "keras_hub.layers.AudioConverter.from_preset",
+            ],
         },
         {
             "path": "image_converter",
             "title": "ImageConverter layer",
-            "generate": ["keras_hub.layers.ImageConverter"],
+            "generate": [
+                "keras_hub.layers.ImageConverter",
+                "keras_hub.layers.ImageConverter.from_preset",
+            ],
         },
         {
             "path": "start_end_packer",
@@ -1848,8 +1854,8 @@
     "title": "API documentation",
     "toc": True,
     "children": [
-        MODELS_MASTER,
         BASE_CLASSES,
+        MODELS_MASTER,
         TOKENIZERS_MASTER,
         PREPROCESSING_LAYERS_MASTER,
         MODELING_LAYERS_MASTER,
@@ -1860,32 +1866,32 @@
 
 HUB_GUIDES_MASTER = {
     "path": "guides/",
-    "title": "Guides",
+    "title": "Developer guides",
     "toc": True,
     "children": [
         {
-            "path": "semantic_segmentation_deeplab_v3",
-            "title": "Semantic Segmentation with KerasHub",
+            "path": "upload",
+            "title": "Uploading Models",
         },
         {
-            "path": "transformer_pretraining",
-            "title": "Pretraining a Transformer from scratch with KerasHub",
+            "path": "stable_diffusion_3_in_keras_hub",
+            "title": "Stable Diffusion 3",
         },
         {
-            "path": "upload",
-            "title": "Uploading Models with KerasHub",
+            "path": "segment_anything_in_keras_hub",
+            "title": "Segment Anything",
         },
         {
             "path": "classification_with_keras_hub",
-            "title": "Classification with KerasHub",
+            "title": "Image Classification",
         },
         {
-            "path": "segment_anything_in_keras_hub",
-            "title": "Segment Anything in KerasHub",
+            "path": "semantic_segmentation_deeplab_v3",
+            "title": "Semantic Segmentation",
         },
         {
-            "path": "stable_diffusion_3_in_keras_hub",
-            "title": "Stable Diffusion 3 in KerasHub",
+            "path": "transformer_pretraining",
+            "title": "Pretraining a Transformer from scratch",
         },
     ],
 }
@@ -1902,7 +1908,7 @@
         HUB_API_MASTER,
         {
             "path": "presets/",
-            "title": "Presets",
+            "title": "Pretrained models list",
        },
     ],
 }
diff --git a/scripts/render_presets.py b/scripts/render_presets.py
@@ -102,8 +102,10 @@ def render_table(symbol):
         return None
     for preset in symbol.presets:
         metadata = symbol.presets[preset]["metadata"]
+        url = symbol.presets[preset]["kaggle_handle"]
+        url = url.replace("kaggle://", "https://www.kaggle.com/models/")
         table += (
-            f"{preset} | "
+            f"[{preset}]({url}) | "
             f"{format_param_count(metadata)} | "
             f"{metadata['description']} \n"
         )

diff --git a/scripts/tuner_master.py b/scripts/tuner_master.py
@@ -210,7 +210,7 @@
 
 TUNER_GUIDES_MASTER = {
     "path": "guides/",
-    "title": "Guides",
+    "title": "Developer guides",
     "toc": True,
     "children": [
         {

diff --git a/templates/keras_hub/api/base_classes/index.md b/templates/keras_hub/api/base_classes/index.md
@@ -0,0 +1,12 @@
+# KerasHub Modeling API
+
+The following base classes form the API for working with pretrained models
+through KerasHub. These base classes can be used with the `from_preset()`
+constructor to automatically instantiate a subclass with the correct model
+architecture, e.g.
+`keras_hub.models.TextClassifier.from_preset("bert_base_en", num_classes=2)`.
+
+For the full list of available pretrained model presets shipped directly by the
+Keras team, see the [Pretrained Models](/keras_hub/presets/) page.
+
+{{toc}}
diff --git a/templates/keras_hub/api/index.md b/templates/keras_hub/api/index.md
@@ -1,9 +1,6 @@
 # KerasHub API documentation
 
 KerasHub is a toolbox of modular building blocks ranging from pretrained
-state-of-the-art models, to low-level Transformer Encoder layers. For an
-introduction to the library see the  [KerasHub home page](/keras_hub/). For a
-high-level introduction to the API see our
-[getting started guide](/keras_hub/getting_started/).
+state-of-the-art models, to low-level Transformer Encoder layers.
 
 {{toc}}
diff --git a/templates/keras_hub/api/metrics/index.md b/templates/keras_hub/api/metrics/index.md
@@ -1,5 +1,6 @@
 # KerasHub Metrics
 
-KerasHub metrics are `keras.Metric` subclasses for NLP-specific use cases.
+KerasHub metrics are `keras.Metric` subclasses that are common to computer
+vision and natural language processing workflows.
 
 {{toc}}
diff --git a/templates/keras_hub/api/modeling_layers/index.md b/templates/keras_hub/api/modeling_layers/index.md
@@ -0,0 +1,7 @@
+# KerasHub Modeling Layers
+
+KerasHub modeling layers are give `keras.layers.Layer` implementations for
+building blocks common to pretrained models. They can be used to create a new
+model from scratch, or to extend a pretrained model.
+
+{{toc}}
diff --git a/templates/keras_hub/api/models/index.md b/templates/keras_hub/api/models/index.md
@@ -1,15 +1,13 @@
-# KerasHub Models
+# KerasHub Model Architectures
 
-KerasHub contains end-to-end implementations of popular model architectures.
-These models can be created in two ways:
+The following is a  list of model architectures supported by KerasHub. These
+models can be created in two ways:
 
 - Through the `from_preset()` constructor, which instantiates an object with
-  a pre-trained configurations, vocabularies, and (optionally) weights.
+  a pre-trained configurations, vocabularies, and weights.
 - Through custom configuration controlled by the user.
 
-For the full list of available presets shipped directly by the Keras team, see
-the [Presets page](/keras_hub/presets/)
-
-## API Documentation
+For the full list of available pretrained model presets shipped directly by the
+Keras team, see the [Pretrained Models](/keras_hub/presets/) page.
 
 {{toc}}
diff --git a/templates/keras_hub/api/preprocessing_layers/index.md b/templates/keras_hub/api/preprocessing_layers/index.md
@@ -0,0 +1,7 @@
+# KerasHub Preprocessing Layers
+
+KerasHub preprocessing layers can be used to create custom preprocessing
+pipelines for pretrained models. Preprocessing layers are all compatible with
+`tf.data`, even when running on the `jax` and `torch` backends.
+
+{{toc}}
diff --git a/templates/keras_hub/api/samplers/index.md b/templates/keras_hub/api/samplers/index.md
@@ -0,0 +1,7 @@
+# KerasHub Samplers
+
+KerasHub samplers are used to control the sampling process when generating text
+with generative models. See `keras_hub.models.CausalLM` for the high-level
+generative API.
+
+{{toc}}
diff --git a/templates/keras_hub/index.md b/templates/keras_hub/index.md
@@ -6,23 +6,19 @@
 and fast. The library provides [Keras 3](https://keras.io/keras_3/)
 implementations of popular model architectures, paired with a collection of
 pretrained checkpoints available on [Kaggle Models](https://kaggle.com/models/).
-Models can be use for both training and inference, on any of the TensorFlow,
+Models can be used for both training and inference, on any of the TensorFlow,
 Jax, and Torch backends.
 
 KerasHub is an extension of the core Keras API; KerasHub components are provided
 as `keras.layers.Layer` and `keras.Model` implementations. If you are familiar
 with Keras, congratulations! You already understand most of KerasHub.
 
-See our [Getting Started guide](/guides/keras_hub/getting_started)
-to start learning our API. We welcome
-[contributions](https://github.com/keras-team/keras-hub/issues/1835).
-
 ---
 ## Quick links
 
 * [Getting started with KerasHub](/keras_hub/getting_started/)
-* [Guides](/keras_hub/guides)
-* [API reference](/keras_hub/api/)
+* [Developer guides](/keras_hub/guides)
+* [API documentation](/keras_hub/api/)
 * [KerasHub on GitHub](https://github.com/keras-team/keras-hub)
 * [KerasHub models on Kaggle](https://www.kaggle.com/organizations/keras/models)
 * [Pretrained model list](/keras_hub/presets/)
@@ -43,62 +39,65 @@ our nightly package.
 pip install --upgrade keras-hub-nightly
 ```
 
-Note that currently, installing KerasHub will always pull in TensorFlow for use
-of the `tf.data` API for preprocessing. Even when pre-processing with `tf.data`,
-training can still happen on any backend. Read
-[Getting started with Keras](https://keras.io/getting_started/) for more
-information on installing Keras 3 and compatibility with different frameworks.
+Currently, installing KerasHub will always pull in TensorFlow for use of the
+`tf.data` API for preprocessing. When pre-processing with `tf.data`, training
+can still happen on any backend.
 
-**Note:** We recommend using KerasHub with TensorFlow 2.16 or later, as TF 2.16
-packages Keras 3 by default.
+Visit the [core Keras getting started page](https://keras.io/getting_started/)
+for more information on installing Keras 3, accelerator support, and
+compatibility with different frameworks.
 
 ---
 ## Quickstart
 
-Below is a quick example using ResNet to predict an image, and BERT to train a
-classifier:
+Choose a backend:
 
 ```python
+# Set a backend.
 import os
 os.environ["KERAS_BACKEND"] = "jax"  # Or "tensorflow" or "torch"!
+```
+
+Import KerasHub and other libraries:
 
+```python
+# Import libraries.
 import keras
 import keras_hub
 import numpy as np
 import tensorflow_datasets as tfds
+```
+
+Load a resnet model and use it to predict a label for an image:
 
-# Load a ResNet model.
+```python
 classifier = keras_hub.models.ImageClassifier.from_preset(
     "resnet_50_imagenet",
     activation="softmax",
 )
-# Predict a label for a single image.
-image_url = "https://upload.wikimedia.org/wikipedia/commons/a/aa/California_quail.jpg"
-image_path = keras.utils.get_file(origin=image_url)
-image = keras.utils.load_img(image_path)
-batch = np.array([image])
-preds = classifier.predict(batch)
+url = "https://upload.wikimedia.org/wikipedia/commons/a/aa/California_quail.jpg"
+path = keras.utils.get_file(origin=url)
+image = keras.utils.load_img(path)
+preds = classifier.predict(np.array([image]))
 print(keras_hub.utils.decode_imagenet_predictions(preds))
+```
+
+Load a Bert model and fine-tune it on IMDb movie reviews:
 
-# Load a BERT model.
+```python
 classifier = keras_hub.models.BertClassifier.from_preset(
     "bert_base_en_uncased",
     activation="softmax",
     num_classes=2,
 )
-
-# Fine-tune on IMDb movie reviews.
 imdb_train, imdb_test = tfds.load(
     "imdb_reviews",
     split=["train", "test"],
     as_supervised=True,
     batch_size=16,
 )
 classifier.fit(imdb_train, validation_data=imdb_test)
-# Predict two new examples.
-preds = classifier.predict(
-    ["What an amazing movie!", "A total waste of my time."]
-)
+preds = classifier.predict(["What an amazing movie!", "A total waste of time."])
 print(preds)
 ```
 

diff --git a/templates/keras_hub/presets/index.md b/templates/keras_hub/presets/index.md
@@ -1,3 +1,5 @@
+# KerasHub pretrained models
+
 Below, we list all presets available in the KerasHub library. For more detailed
 usage, browse the docstring for a particular class. For an in depth introduction
 to our API, see the [getting started guide](/keras_hub/getting_started/).

diff --git a/templates/keras_tuner/index.md b/templates/keras_tuner/index.md
@@ -14,8 +14,8 @@ experiment with new search algorithms.
 ## Quick links
 
 * [Getting started with KerasTuner](/keras_tuner/getting_started/)
-* [Guides](/keras_tuner/guides)
-* [API reference](/keras_tuner/api/)
+* [Developer guides](/keras_tuner/guides)
+* [API documentation](/keras_tuner/api/)
 * [KerasTuner on GitHub](https://github.com/keras-team/keras-tuner)
 
 
@@ -74,7 +74,8 @@ tuner.search(x_train, y_train, epochs=5, validation_data=(x_val, y_val))
 best_model = tuner.get_best_models()[0]
 ```
 
-To learn more about KerasTuner, check out [this starter guide](https://keras.io/guides/keras_tuner/getting_started/).
+To learn more about KerasTuner, check out the
+[getting stated guide](https://keras.io/guides/keras_tuner/getting_started/).
 
 
 ---