refactor(samples): load named JSON presets from files

app.py

@@ -4,205 +4,26 @@ import gradio as gr
 
 from model import extract, generate_template
 
+templates_directory = Path(__file__).parent / "data" / "templates"
 structured_json_templates = {
-    "Basic receipt": """{
-  "merchant": "verbatim-string",
-  "purchase_date": "date",
-  "purchase_time": "time",
-  "total": "number",
-  "currency": "currency",
-  "item_count": "integer",
-  "paid": "boolean",
-  "payment_method": ["cash", "credit-card", "debit-card", "other"],
-  "items": ["verbatim-string"]
-}""",
-    "Advanced receipt": """{
-  "merchant": {
-    "name": "verbatim-string",
-    "address": "verbatim-string",
-    "phone": "phone-number"
-  },
-  "purchase_date": "date",
-  "purchase_time": "time",
-  "store_number": "verbatim-string",
-  "operator_number": "verbatim-string",
-  "terminal_number": "verbatim-string",
-  "transaction_number": "verbatim-string",
-  "items": [
-    {
-      "description": "verbatim-string",
-      "sku": "verbatim-string",
-      "quantity": "number",
-      "unit": "unit-code",
-      "unit_price": "number",
-      "discount": "number",
-      "line_total": "number",
-      "tax_code": "verbatim-string"
-    }
-  ],
-  "item_count": "integer",
-  "subtotal": "number",
-  "tax": "number",
-  "total": "number",
-  "currency": "currency",
-  "payment_method": ["cash", "credit-card", "debit-card", "other"],
-  "card_last_four": "verbatim-string",
-  "amount_tendered": "number",
-  "change_due": "number"
-}""",
-    "Basic bank statement": """{
-  "financial_institution": "verbatim-string",
-  "account_holder": "verbatim-string",
-  "account_number": "verbatim-string",
-  "statement_period": {
-    "start_date": "date",
-    "end_date": "date"
-  },
-  "opening_balance": "number",
-  "closing_balance": "number",
-  "currency": "currency",
-  "transaction_count": "integer",
-  "transactions": [
-    {
-      "date": "date",
-      "description": "verbatim-string",
-      "transaction_type": ["debit", "credit", "other"],
-      "amount": "number",
-      "balance": "number"
-    }
-  ]
-}""",
-    "Advanced bank statement": """{
-  "financial_institution": {
-    "name": "verbatim-string",
-    "address": "verbatim-string",
-    "phone": "phone-number",
-    "website": "url"
-  },
-  "account_holder": {
-    "name": "verbatim-string",
-    "address": "verbatim-string"
-  },
-  "account": {
-    "type": "verbatim-string",
-    "number": "verbatim-string",
-    "branch_number": "verbatim-string",
-    "transit_number": "verbatim-string",
-    "routing_number": "verbatim-string",
-    "iban": "iban",
-    "bic": "bic"
-  },
-  "statement_period": {
-    "start_date": "date",
-    "end_date": "date"
-  },
-  "summary": {
-    "opening_balance": "number",
-    "total_credits": "number",
-    "total_debits": "number",
-    "total_fees": "number",
-    "total_interest": "number",
-    "closing_balance": "number",
-    "currency": "currency"
-  },
-  "transaction_count": "integer",
-  "transactions": [
-    {
-      "transaction_date": "date",
-      "posting_date": "date",
-      "description": "verbatim-string",
-      "reference_number": "verbatim-string",
-      "category": [
-        "deposit",
-        "withdrawal",
-        "transfer",
-        "payment",
-        "purchase",
-        "fee",
-        "interest",
-        "refund",
-        "other"
-      ],
-      "debit": "number",
-      "credit": "number",
-      "balance": "number",
-      "currency": "currency"
-    }
-  ]
-}""",
-    "Invoice with line items": """{
-  "document_type": "verbatim-string",
-  "vendor": {
-    "name": "verbatim-string",
-    "address": "verbatim-string",
-    "email": "email-address",
-    "phone": "phone-number"
-  },
-  "invoice_number": "verbatim-string",
-  "invoice_date": "verbatim-string",
-  "line_items": [
-    {
-      "description": "verbatim-string",
-      "category": ["product", "service", "shipping", "tax", "other"],
-      "quantity": "integer",
-      "unit_price": "number",
-      "unit": "unit-code",
-      "amount": "number"
-    }
-  ],
-  "subtotal": "number",
-  "tax": "number",
-  "total": "number",
-  "currency": "currency"
-}""",
-    "Advanced global record": """{
-  "summary": "string",
-  "reference_text": "verbatim-string",
-  "sequence_number": "integer",
-  "amount": "number",
-  "effective_date": "date",
-  "effective_time": "time",
-  "signed_at": "date-time",
-  "term": "duration",
-  "active": "boolean",
-  "country": "country",
-  "currency": "currency",
-  "language": "language",
-  "language_tag": "language-tag",
-  "script": "script",
-  "website": "url",
-  "email": "email-address",
-  "phone": "phone-number",
-  "iban": "iban",
-  "bic": "bic",
-  "measurement_unit": "unit-code",
-  "regions": {
-    "us": "region:US",
-    "fr": "region:FR",
-    "ie": "region:IE",
-    "gb": "region:GB",
-    "it": "region:IT",
-    "es": "region:ES",
-    "de": "region:DE",
-    "pt": "region:PT",
-    "ca": "region:CA",
-    "mx": "region:MX",
-    "br": "region:BR",
-    "au": "region:AU",
-    "jp": "region:JP",
-    "kr": "region:KR"
-  },
-  "contacts": [
-    {
-      "name": "verbatim-string",
-      "role": "string",
-      "email": "email-address"
-    }
-  ],
-  "status": ["draft", "active", "expired", "unknown"],
-  "applicable_labels": [["legal", "financial", "technical", "personal"]],
-  "keywords": ["string"]
-}""",
+    "Basic receipt": (templates_directory / "basic-receipt.json").read_text(
+        encoding="utf-8"
+    ),
+    "Invoice with line items": (
+        templates_directory / "invoice-with-line-items.json"
+    ).read_text(encoding="utf-8"),
+    "Basic bank statement": (
+        templates_directory / "basic-bank-statement.json"
+    ).read_text(encoding="utf-8"),
+    "Advanced bank statement": (
+        templates_directory / "advanced-bank-statement.json"
+    ).read_text(encoding="utf-8"),
+    "Model task catalog": (
+        templates_directory / "model-task-catalog.json"
+    ).read_text(encoding="utf-8"),
+    "Todo list": (templates_directory / "todo-list.json").read_text(
+        encoding="utf-8"
+    ),
 }
 
 default_structured_json_template = "Invoice with line items"
@@ -258,8 +79,8 @@ with gr.Blocks(title="Image Data Extractor") as demo:
                 ("Invoice with line items", "invoice-with-items.png"),
                 ("Advanced bank statement", "BankStatementChequing.png"),
                 ("Basic bank statement", "bank statement blog image.webp"),
-                ("Advanced global record", "tags.png"),
-                ("Advanced global record", "task-list.png"),
+                ("Model task catalog", "tags.png"),
+                ("Todo list", "task-list.png"),
             ]
         ],
         inputs=[template_preset, image, text, enable_thinking],

data/templates/advanced-bank-statement.json

@@ -0,0 +1,58 @@
+{
+  "financial_institution": {
+    "name": "verbatim-string",
+    "address": "verbatim-string",
+    "phone": "phone-number",
+    "website": "url"
+  },
+  "account_holder": {
+    "name": "verbatim-string",
+    "address": "verbatim-string"
+  },
+  "account": {
+    "type": "verbatim-string",
+    "number": "verbatim-string",
+    "branch_number": "verbatim-string",
+    "transit_number": "verbatim-string",
+    "routing_number": "verbatim-string",
+    "iban": "iban",
+    "bic": "bic"
+  },
+  "statement_period": {
+    "start_date": "date",
+    "end_date": "date"
+  },
+  "summary": {
+    "opening_balance": "number",
+    "total_credits": "number",
+    "total_debits": "number",
+    "total_fees": "number",
+    "total_interest": "number",
+    "closing_balance": "number",
+    "currency": "currency"
+  },
+  "transaction_count": "integer",
+  "transactions": [
+    {
+      "transaction_date": "date",
+      "posting_date": "date",
+      "description": "verbatim-string",
+      "reference_number": "verbatim-string",
+      "category": [
+        "deposit",
+        "withdrawal",
+        "transfer",
+        "payment",
+        "purchase",
+        "fee",
+        "interest",
+        "refund",
+        "other"
+      ],
+      "debit": "number",
+      "credit": "number",
+      "balance": "number",
+      "currency": "currency"
+    }
+  ]
+}

data/templates/basic-bank-statement.json

@@ -0,0 +1,22 @@
+{
+  "financial_institution": "verbatim-string",
+  "account_holder": "verbatim-string",
+  "account_number": "verbatim-string",
+  "statement_period": {
+    "start_date": "date",
+    "end_date": "date"
+  },
+  "opening_balance": "number",
+  "closing_balance": "number",
+  "currency": "currency",
+  "transaction_count": "integer",
+  "transactions": [
+    {
+      "date": "date",
+      "description": "verbatim-string",
+      "transaction_type": ["debit", "credit", "other"],
+      "amount": "number",
+      "balance": "number"
+    }
+  ]
+}

data/templates/basic-receipt.json

@@ -0,0 +1,11 @@
+{
+  "merchant": "verbatim-string",
+  "purchase_date": "date",
+  "purchase_time": "time",
+  "total": "number",
+  "currency": "currency",
+  "item_count": "integer",
+  "paid": "boolean",
+  "payment_method": ["cash", "credit-card", "debit-card", "other"],
+  "items": ["verbatim-string"]
+}

data/templates/generated-bank-statement-template.json

@@ -1,40 +0,0 @@
-{
-  "account_holder": {
-    "address": {
-      "street_address": "string",
-      "postal_code": "string",
-      "city": "verbatim-string",
-      "province": "region:CA"
-    },
-    "name": "verbatim-string"
-  },
-  "statement_period": {
-    "end_date": "date",
-    "start_date": "date"
-  },
-  "account_number": "string",
-  "bank_details": {
-    "name": "verbatim-string",
-    "address": {
-      "street_address": "verbatim-string",
-      "postal_code": "verbatim-string",
-      "city": "verbatim-string",
-      "province": "region:CA"
-    },
-    "phone_number": "phone-number"
-  },
-  "transactions": [
-    {
-      "description": "string",
-      "date": "date",
-      "withdrawals": "number",
-      "balance": "number",
-      "ref": "verbatim-string",
-      "deposits": "number"
-    }
-  ],
-  "totals": {
-    "withdrawals": "number",
-    "deposits": "number"
-  }
-}

data/templates/invoice-with-line-items.json

@@ -0,0 +1,25 @@
+{
+  "document_type": "verbatim-string",
+  "vendor": {
+    "name": "verbatim-string",
+    "address": "verbatim-string",
+    "email": "email-address",
+    "phone": "phone-number"
+  },
+  "invoice_number": "verbatim-string",
+  "invoice_date": "verbatim-string",
+  "line_items": [
+    {
+      "description": "verbatim-string",
+      "category": ["product", "service", "shipping", "tax", "other"],
+      "quantity": "integer",
+      "unit_price": "number",
+      "unit": "unit-code",
+      "amount": "number"
+    }
+  ],
+  "subtotal": "number",
+  "tax": "number",
+  "total": "number",
+  "currency": "currency"
+}

data/templates/model-task-catalog.json

@@ -0,0 +1,52 @@
+{
+  "Image Generation": [
+    [
+      "Text to Image",
+      "Image to Text",
+      "Image-to-Image",
+      "Image-to-Video",
+      "Unconditional Image Generation",
+      "Video Classification",
+      "Text-to-Video",
+      "Zero-Shot Image Classification",
+      "Mask Generation",
+      "Zero-Shot Object Detection",
+      "Text-to-3D",
+      "Image-to-3D",
+      "Image Feature Extraction",
+      "Keypoint Detection",
+      "Video-to-Video"
+    ]
+  ],
+  "Natural Language Processing": [
+    [
+      "Text Classification",
+      "Token Classification",
+      "Table Question Answering",
+      "Question Answering",
+      "Zero-Shot Classification",
+      "Translation",
+      "Summarization",
+      "Feature Extraction",
+      "Text Generation",
+      "Fill-Mask",
+      "Sentence Similarity",
+      "Text Ranking"
+    ]
+  ],
+  "Audio": [
+    [
+      "Text-to-Speech",
+      "Text-to-Audio",
+      "Automatic Speech Recognition",
+      "Audio-to-Audio",
+      "Audio Classification",
+      "Voice Activity Detection"
+    ]
+  ],
+  "Tabular": [
+    ["Tabular Classification", "Tabular Regression", "Time Series Forecasting"]
+  ],
+  "Reinforcement Learning": [["Reinforcement Learning", "Robotics"]],
+  "Other": [["Graph Machine Learning"]]
+}

data/templates/todo-list.json

@@ -0,0 +1,10 @@
+{
+  "todos": [
+    {
+      "status": ["completed", "pending"],
+      "description": "string",
+      "related_file": "verbatim-string",
+      "related_cell": "verbatim-string"
+    }
+  ]
+}