Merge pull request #12 from run-llama/feat/examples

Improve the basic and manual examples
run-llama · Jul 25, 2024 · 011f9d5 · 011f9d5
2 parents f335e26 + 07857b2
commit 011f9d5
Show file tree

Hide file tree

Showing 3 changed files with 120 additions and 125 deletions.
diff --git a/examples/demo_basic.ipynb b/examples/demo_basic.ipynb
@@ -4,7 +4,29 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# LlamaExtract Usage"
+    "# Infer a schema to extract data from files"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "In this notebook, we will demonstrate how to infer a schema from a set of files and using it to extract structured data from invoice PDF files.\n",
+    "\n",
+    "The steps are:\n",
+    "1. Infer a schema from the invoices files.\n",
+    "2. Extract structured data (i.e. JSONs) from invoice PDF files\n",
+    "\n",
+    "Additional Resources:\n",
+    "- `LlamaExtract`: https://docs.cloud.llamaindex.ai/"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Setup\n",
+    "Install `llama-extract` client library:"
    ]
   },
   {
@@ -16,6 +38,13 @@
     "%pip install llama-extract"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Apply `nest_asyncio` and bring your own LlamaCloud API key:"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -32,6 +61,14 @@
     "os.environ[\"LLAMA_CLOUD_API_KEY\"] = \"llx-...\""
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Infer the schema\n",
+    "First, let's infer the schema using the invoice files with `LlamaExtract`."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -47,6 +84,13 @@
     ")"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Preview the inferred schema:"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -64,6 +108,14 @@
     "print(extraction_schema.data_schema)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Extract structured data\n",
+    "Now with the schema, we can extract structured data (i.e. JSON) from the our invoices files."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -84,6 +136,13 @@
     ")"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Preview the extracted data:"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,

diff --git a/examples/demo_existent_schema.ipynb b/examples/demo_existent_schema.ipynb
diff --git a/examples/demo_manual.ipynb b/examples/demo_manual.ipynb
@@ -4,7 +4,29 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "# Create a schema with your own schema to extract data from files"
+    "# Manually create a schema to extract data from files"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "In this notebook, we will demonstrate how to manually create a schema and using it to extract structured data from invoice PDF files.\n",
+    "\n",
+    "The steps are:\n",
+    "1. Create a schema using a valid JSON schema object.\n",
+    "2. Extract structured data (i.e. JSONs) from invoice PDF files\n",
+    "\n",
+    "Additional Resources:\n",
+    "- `LlamaExtract`: https://docs.cloud.llamaindex.ai/"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Setup\n",
+    "Install `llama-extract` client library:"
    ]
   },
   {
@@ -16,6 +38,13 @@
     "%pip install llama-extract"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Apply `nest_asyncio` and bring your own LlamaCloud API key:"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -32,6 +61,14 @@
     "os.environ[\"LLAMA_CLOUD_API_KEY\"] = \"llx-...\""
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Create the schema\n",
+    "First, let's create the schema using a valid JSON schema object with `LlamaExtract`."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -54,6 +91,13 @@
     "extraction_schema = extractor.create_schema(\"Test Schema\", data_schema)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Let's preview the created schema:"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -71,6 +115,14 @@
     "print(extraction_schema)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Extract structured data\n",
+    "Now with the schema, we can extract structured data (i.e. JSON) from the our invoices files."
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -91,6 +143,13 @@
     ")"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Preview the extracted data:"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,