Merge branch 'master' into docstring

lightly-ai · Oct 8, 2024 · ce3277a · ce3277a
2 parents fe78a14 + 820b20f
commit ce3277a
Show file tree

Hide file tree

Showing 57 changed files with 915 additions and 196 deletions.
diff --git a/README.md b/README.md
diff --git a/docs/source/getting_started/lightly_at_a_glance.rst b/docs/source/getting_started/lightly_at_a_glance.rst
@@ -34,6 +34,7 @@ Let's now load an image dataset and create a PyTorch dataloader.
 .. code-block:: python
 
     import torch
+    import lightly.data as data
 
     # Create a dataset from your image folder.
     dataset = data.LightlyDataset(

diff --git a/docs/source/tutorials/structure_your_input.rst b/docs/source/tutorials/structure_your_input.rst
@@ -6,7 +6,7 @@ Tutorial 1: Structure Your Input
 The modern-day open-source ecosystem has changed a lot over the years, and there are now
 many viable options for data pipelining. The `torchvision.data <https://pytorch.org/vision/main/datasets.html>`_ submodule provides a robust implementation for most use cases,
 and the `Hugging Face Hub <https://hf.co>`_ has emerged as a growing collection of datasets that span a variety of domains and tasks.
-It you want to use your own data, the ability to quickly create datasets and dataloaders is of prime importance.
+If you want to use your own data, the ability to quickly create datasets and dataloaders is of prime importance.
 
 In this tutorial, we will provide a brief overview of the `LightlyDataset <https://docs.lightly.ai/self-supervised-learning/lightly.data.html#lightly.data.dataset.LightlyDataset>`_
 and go through examples of using datasets from various open-source libraries such as PyTorch and

diff --git a/examples/notebooks/pytorch/aim.ipynb b/examples/notebooks/pytorch/aim.ipynb
@@ -129,16 +129,38 @@
    "cell_type": "code",
    "execution_count": null,
    "id": "8",
-   "metadata": {},
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
    "outputs": [],
    "source": [
     "transform = AIMTransform()\n",
-    "# we ignore object detection annotations by setting target_transform to return 0\n",
+    "# we ignore object detection annotations by setting target_transform to return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def target_transform(t):\n",
+    "    return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "10",
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "dataset = torchvision.datasets.VOCDetection(\n",
     "    \"datasets/pascal_voc\",\n",
     "    download=True,\n",
     "    transform=transform,\n",
-    "    target_transform=lambda t: 0,\n",
+    "    target_transform=target_transform,\n",
     ")\n",
     "# or create a dataset from a folder containing images or videos:\n",
     "# dataset = LightlyDataset(\"path/to/folder\")"
@@ -147,7 +169,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "9",
+   "id": "11",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -163,7 +185,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "10",
+   "id": "12",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -174,7 +196,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "11",
+   "id": "13",
    "metadata": {},
    "outputs": [],
    "source": [

diff --git a/examples/notebooks/pytorch/dino.ipynb b/examples/notebooks/pytorch/dino.ipynb
@@ -139,13 +139,33 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "transform = DINOTransform()\n",
+    "transform = DINOTransform()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "11",
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "# we ignore object detection annotations by setting target_transform to return 0\n",
+    "def target_transform(t):\n",
+    "    return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "12",
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "dataset = torchvision.datasets.VOCDetection(\n",
     "    \"datasets/pascal_voc\",\n",
     "    download=True,\n",
     "    transform=transform,\n",
-    "    target_transform=lambda t: 0,\n",
+    "    target_transform=target_transform,\n",
     ")\n",
     "# or create a dataset from a folder containing images or videos:\n",
     "# dataset = LightlyDataset(\"path/to/folder\")"
@@ -154,7 +174,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "11",
+   "id": "13",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -170,7 +190,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "12",
+   "id": "14",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -185,7 +205,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "13",
+   "id": "15",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -195,7 +215,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "14",
+   "id": "16",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -205,7 +225,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "15",
+   "id": "17",
    "metadata": {},
    "outputs": [],
    "source": [

diff --git a/examples/notebooks/pytorch/ijepa.ipynb b/examples/notebooks/pytorch/ijepa.ipynb
@@ -128,21 +128,41 @@
     "transform = IJEPATransform()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "8",
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
+   "source": [
+    "we ignore object detection annotations by setting target_transform to return 0\n",
+    "or create a dataset from a folder containing images or videos:\n",
+    "dataset = LightlyDataset(\"path/to/folder\")"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "8",
+   "id": "9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def target_transform(t):\n",
+    "    return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "10",
    "metadata": {},
    "outputs": [],
    "source": [
-    "# we ignore object detection annotations by setting target_transform to return 0\n",
-    "# or create a dataset from a folder containing images or videos:\n",
-    "# dataset = LightlyDataset(\"path/to/folder\")\n",
     "dataset = torchvision.datasets.VOCDetection(\n",
     "    \"datasets/pascal_voc\",\n",
     "    download=True,\n",
     "    transform=transform,\n",
-    "    target_transform=lambda t: 0,\n",
+    "    target_transform=target_transform,\n",
     ")\n",
     "data_loader = torch.utils.data.DataLoader(\n",
     "    dataset, collate_fn=collator, batch_size=10, persistent_workers=False\n",
@@ -152,7 +172,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "9",
+   "id": "11",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -169,7 +189,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "10",
+   "id": "12",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -181,7 +201,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "11",
+   "id": "13",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -194,7 +214,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "12",
+   "id": "14",
    "metadata": {},
    "outputs": [],
    "source": [

diff --git a/examples/notebooks/pytorch/mae.ipynb b/examples/notebooks/pytorch/mae.ipynb
@@ -141,16 +141,38 @@
    "cell_type": "code",
    "execution_count": null,
    "id": "7",
-   "metadata": {},
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
    "outputs": [],
    "source": [
     "transform = MAETransform()\n",
-    "# we ignore object detection annotations by setting target_transform to return 0\n",
+    "# we ignore object detection annotations by setting target_transform to return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def target_transform(t):\n",
+    "    return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "dataset = torchvision.datasets.VOCDetection(\n",
     "    \"datasets/pascal_voc\",\n",
     "    download=True,\n",
     "    transform=transform,\n",
-    "    target_transform=lambda t: 0,\n",
+    "    target_transform=target_transform,\n",
     ")\n",
     "# or create a dataset from a folder containing images or videos:\n",
     "# dataset = LightlyDataset(\"path/to/folder\")"
@@ -159,7 +181,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "8",
+   "id": "10",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -175,7 +197,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "9",
+   "id": "11",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -186,7 +208,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "10",
+   "id": "12",
    "metadata": {},
    "outputs": [],
    "source": [

diff --git a/examples/notebooks/pytorch/msn.ipynb b/examples/notebooks/pytorch/msn.ipynb
@@ -134,16 +134,38 @@
    "cell_type": "code",
    "execution_count": null,
    "id": "8",
-   "metadata": {},
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
    "outputs": [],
    "source": [
     "transform = MSNTransform()\n",
-    "# we ignore object detection annotations by setting target_transform to return 0\n",
+    "# we ignore object detection annotations by setting target_transform to return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def target_transform(t):\n",
+    "    return 0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "10",
+   "metadata": {},
+   "outputs": [],
+   "source": [
     "dataset = torchvision.datasets.VOCDetection(\n",
     "    \"datasets/pascal_voc\",\n",
     "    download=True,\n",
     "    transform=transform,\n",
-    "    target_transform=lambda t: 0,\n",
+    "    target_transform=target_transform,\n",
     ")\n",
     "# or create a dataset from a folder containing images or videos:\n",
     "# dataset = LightlyDataset(\"path/to/folder\")"
@@ -152,7 +174,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "9",
+   "id": "11",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -168,7 +190,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "10",
+   "id": "12",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -178,7 +200,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "11",
+   "id": "13",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -193,7 +215,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "12",
+   "id": "14",
    "metadata": {},
    "outputs": [],
    "source": [