Sync from GitHub (HuggingFaceIntegerations)

Makefile

@@ -0,0 +1,30 @@
+.PHONY: install test debug one-test format format-check lint all
+
+install:
+	python -m pip install --upgrade pip
+	python -m pip install -r requirements.txt
+
+test:
+	python -m pytest -vv \
+		--cov=app \
+		--cov-report=term-missing \
+		--cov-fail-under=80
+
+debug:
+	python -m pytest -vv -x
+
+# Run a single test by node id (file::test_name)
+# Usage: make one-test T=test_app.py::test_hi_to_en_uses_hi_en_pipe
+one-test:
+	python -m pytest -vv "$(T)"
+
+format:
+	black *.py
+
+format-check:
+	black --check *.py
+
+lint:
+	pylint --disable=R,C,no-member app.py test_app.py
+
+all: install lint format-check test

README.md

@@ -0,0 +1,39 @@
+---
+title: Hindi ↔ English + Hinglish Translator
+emoji: 🌐
+colorFrom: blue
+colorTo: purple
+sdk: gradio
+sdk_version: 6.6.0
+app_file: app.py
+license: mit
+short_description: Hindi↔English translator + Hinglish transliteration
+tags:
+  - translation
+  - machine-translation
+  - hindi
+  - hinglish
+  - indic-languages
+  - gradio
+models:
+  - Helsinki-NLP/opus-mt-hi-en
+  - Helsinki-NLP/opus-mt-en-hi
+---
+
+# Hindi ↔ English + Hinglish Translator
+
+A simple, fast multilingual translator demo built with Hugging Face Transformers and Gradio.
+
+## Features
+- **Hindi → English** — Neural machine translation (MarianMT)
+- **English → Hindi** — Neural machine translation (MarianMT)
+- **Hindi → Hinglish** — Roman transliteration using ITRANS scheme
+
+### Powered by
+- [Helsinki-NLP/opus-mt-hi-en](https://huggingface.co/Helsinki-NLP/opus-mt-hi-en)
+- [Helsinki-NLP/opus-mt-en-hi](https://huggingface.co/Helsinki-NLP/opus-mt-en-hi)
+- [indic-transliteration library](https://github.com/AnimeshSinha1309/indic-transliteration)
+
+Just type or paste text, select a task, adjust max length if needed, and click **Translate**!
+
+Made with ❤️ in India 🇮🇳  

app.py

@@ -0,0 +1,152 @@
+from functools import lru_cache
+from typing import Literal, Any
+
+import gradio as gr
+from transformers import pipeline
+from indic_transliteration import sanscript
+from indic_transliteration.sanscript import transliterate
+import os
+os.environ["TF_ENABLE_ONEDNN_OPTS"] = "0"  # silences oneDNN
+
+# Model identifiers (Helsinki-NLP MarianMT models)
+HI_EN_MODEL = "Helsinki-NLP/opus-mt-hi-en"
+EN_HI_MODEL = "Helsinki-NLP/opus-mt-en-hi"
+
+TaskType = Literal[
+    "Hindi → English",
+    "English → Hindi",
+    "Hindi → Hinglish (transliteration)",
+]
+
+
+@lru_cache(maxsize=None)
+def get_hi_en_pipe() -> pipeline:
+    """Lazy-load Hindi → English translation pipeline"""
+    return pipeline("translation", model=HI_EN_MODEL)
+
+
+@lru_cache(maxsize=None)
+def get_en_hi_pipe() -> pipeline:
+    """Lazy-load English → Hindi translation pipeline"""
+    return pipeline("translation", model=EN_HI_MODEL)
+
+
+def translate(
+    text: str | None,
+    task: TaskType,
+    max_length: int = 256,
+) -> str:
+    """
+    Main translation function used by the Gradio interface.
+    Handles empty input gracefully and supports three tasks.
+    """
+    text = (text or "").strip()
+    if not text:
+        return ""
+
+    try:
+        if task == "Hindi → English":
+            pipe = get_hi_en_pipe()
+            result = pipe(
+                text,
+                max_length=max_length,
+                num_beams=4,
+                early_stopping=True,
+            )
+            return result[0]["translation_text"]
+
+        if task == "English → Hindi":
+            pipe = get_en_hi_pipe()
+            result = pipe(
+                text,
+                max_length=max_length,
+                num_beams=4,
+                early_stopping=True,
+            )
+            return result[0]["translation_text"]
+
+        if task == "Hindi → Hinglish (transliteration)":
+            return transliterate(text, sanscript.DEVANAGARI, sanscript.ITRANS)
+
+        return "Unsupported task selected"
+
+    except (ValueError, RuntimeError, OSError) as e:
+        return f"Translation error: {str(e)}"
+    except Exception as e:  # pylint: disable=broad-exception-caught
+        return f"Unexpected error during translation: {str(e)}"
+
+
+def build_demo() -> gr.Blocks:
+    """Build and configure the Gradio interface"""
+    with gr.Blocks(title="Hindi ↔ English + Hinglish Translator") as demo:
+        gr.Markdown(
+            """
+            # Hindi ↔ English + Hinglish Translator
+            
+            - **Hindi → English** & **English → Hindi**: Neural machine translation  
+            - **Hindi → Hinglish**: Roman transliteration (Devanagari → ITRANS)
+            """
+        )
+
+        with gr.Row():
+            task = gr.Dropdown(
+                choices=[
+                    "Hindi → English",
+                    "English → Hindi",
+                    "Hindi → Hinglish (transliteration)",
+                ],
+                value="Hindi → English",
+                label="Task",
+                interactive=True,
+            )
+            max_len = gr.Slider(
+                32, 512, value=256, step=16,
+                label="Max output length",
+                info="Higher values allow longer translations (slower)",
+            )
+
+        input_text = gr.Textbox(
+            label="Input text",
+            lines=5,
+            placeholder="नमस्ते दुनिया! या Hello world...",
+        )
+
+        translate_btn = gr.Button("Translate", variant="primary")
+
+        output_text = gr.Textbox(
+            label="Output",
+            lines=5,
+            interactive=False,
+        )
+
+        translate_btn.click(
+            fn=translate,
+            inputs=[input_text, task, max_len],
+            outputs=output_text,
+        )
+
+        gr.Examples(
+            examples=[
+                ["नमस्ते! सब ठीक है?", "Hindi → English", 128],
+                ["How are you today?", "English → Hindi", 128],
+                ["नमस्ते भाई, क्या हाल है?", "Hindi → Hinglish (transliteration)", 128],
+                ["खुश रहो और मुस्कुराते रहो", "Hindi → Hinglish (transliteration)", 128],
+            ],
+            inputs=[input_text, task, max_len],
+            label="Quick examples",
+        )
+
+    return demo
+
+
+demo = build_demo()
+
+
+if __name__ == "__main__":
+    # pragma: no cover - UI launch not tested
+    demo.launch(
+        # share=False,
+        # inbrowser=False,
+        # server_name="0.0.0.0",
+        # server_port=7860,
+    )

requirements.txt

@@ -0,0 +1,10 @@
+gradio==6.6.0
+transformers==4.45
+tf-keras==2.20.1
+torch==2.10.0
+sentencepiece==0.2.1
+indic-transliteration
+pytest==9.0.2
+pytest-cov==7.0.0
+black==26.1.0
+pylint==4.0.4

test_app.py

@@ -0,0 +1,69 @@
+import app
+import gradio as gr
+
+
+class DummyPipe:
+    def __init__(self, out_text: str):
+        self.out_text = out_text
+
+    def __call__(self, text, **kwargs):  # accept all kwargs → matches real pipeline
+        return [{"translation_text": self.out_text}]
+
+
+def test_translate_empty_returns_empty():
+    assert app.translate("", "Hindi → English") == ""
+    assert app.translate("   ", "English → Hindi") == ""
+    assert app.translate(None, "Hindi → Hinglish (transliteration)") == ""
+
+
+def test_hi_to_en_uses_hi_en_pipe(monkeypatch):
+    app.get_hi_en_pipe.cache_clear()
+    monkeypatch.setattr(app, "get_hi_en_pipe", lambda: DummyPipe("hello"))
+    assert app.translate("नमस्ते", "Hindi → English") == "hello"
+
+
+def test_en_to_hi_uses_en_hi_pipe(monkeypatch):
+    app.get_en_hi_pipe.cache_clear()
+    monkeypatch.setattr(app, "get_en_hi_pipe", lambda: DummyPipe("नमस्ते"))
+    assert app.translate("hello", "English → Hindi") == "नमस्ते"
+
+
+def test_hi_to_hinglish_transliteration_nonempty():
+    out = app.translate("नमस्ते भाई कैसे हो?", "Hindi → Hinglish (transliteration)")
+    assert isinstance(out, str)
+    assert len(out) > 0
+
+def test_build_demo_returns_blocks():
+    demo = app.build_demo()
+    assert isinstance(demo, gr.Blocks)
+    assert demo.title == "Hindi ↔ English + Hinglish Translator"  # or whatever your title is
+
+def get_all_component_types(block):
+    types = set()
+    def recurse(b):
+        if hasattr(b, 'children'):
+            for child in b.children or []:
+                types.add(type(child).__name__)
+                recurse(child)
+    recurse(block)
+    return types
+
+
+def test_build_demo_has_expected_components():
+    demo = app.build_demo()
+    component_types = get_all_component_types(demo)
+    
+    assert "Dropdown" in component_types
+    assert "Slider" in component_types
+    assert "Textbox" in component_types
+    assert "Button" in component_types
+    
+def test_translate_catches_translation_error(monkeypatch):
+    def broken_pipe(*args, **kwargs):
+        raise RuntimeError("mock model crash")
+
+    monkeypatch.setattr(app, "get_hi_en_pipe", lambda: broken_pipe)
+    
+    result = app.translate("नमस्ते", "Hindi → English")
+    assert "error" in result.lower()
+    assert "mock model crash" in result   # or just check prefix