Upload folder using huggingface_hub

README.md

@@ -1,13 +1,40 @@
 ---
 title: Sentiment Analyzer
-emoji: 🚀
+emoji: 🎭
 colorFrom: indigo
-colorTo: indigo
+colorTo: pink
 sdk: gradio
 sdk_version: 6.19.0
-python_version: '3.13'
 app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 🎭 Sentiment Analyzer
+
+A lightweight text sentiment classifier running DistilBERT (67M params). 
+Type any sentence and get instant positive/negative analysis.
+
+**Why this exists:** Small models can do real work. This runs on CPU in 
+milliseconds — no GPU needed, no queue, no waiting.
+
+**What I learned building this:**
+- Loading transformers pipelines for inference
+- Building clean Gradio interfaces
+- Deploying to Hugging Face Spaces
+
+## Usage
+
+```python
+from transformers import pipeline
+
+classifier = pipeline(
+    "sentiment-analysis",
+    model="distilbert-base-uncased-finetuned-sst-2-english"
+)
+classifier("This is great!")
+# [{'label': 'POSITIVE', 'score': 0.9998}]
+```
+
+## Links
+- **HF Profile:** [arinbalyan](https://huggingface.co/arinbalyan)
+- **Training Notebook:** Coming soon (Kaggle fine-tune on custom data)

app.py

@@ -0,0 +1,58 @@
+import gradio as gr
+from transformers import pipeline
+
+# ponytail: single pipeline call, no wrapping class needed
+classifier = pipeline(
+    "sentiment-analysis",
+    model="distilbert-base-uncased-finetuned-sst-2-english",
+)
+
+# ponytail: examples embedded, no external file
+EXAMPLES = [
+    "This movie was absolutely incredible! Best I've seen all year.",
+    "Waste of time. Terrible acting, predictable plot.",
+    "It was okay, nothing special but not bad either.",
+    "The cinematography was stunning but the story fell flat.",
+    "I've watched it three times already. Pure masterpiece.",
+]
+
+def analyze(text):
+    result = classifier(text[:512])[0]  # ponytail: hard truncate, no tokenizer logic
+    label = result["label"]
+    score = result["score"]
+    emoji = "😊" if label == "POSITIVE" else "😞"
+    return {emoji: score if label == "POSITIVE" else 1 - score}
+
+with gr.Blocks(theme=gr.themes.Soft(), title="Sentiment Analyzer") as demo:
+    gr.Markdown(
+        """
+    # 🎭 Sentiment Analyzer
+
+    Tiny model, big opinions. Type any text and see if it's positive or negative.
+
+    **Model:** DistilBERT fine-tuned on SST-2 (67M params, runs on CPU)
+    """
+    )
+    with gr.Row():
+        text_input = gr.Textbox(
+            label="Your text",
+            placeholder="What's on your mind?",
+            lines=3,
+        )
+    with gr.Row():
+        btn = gr.Button("Analyze", variant="primary", scale=0)
+    with gr.Row():
+        output = gr.Label(label="Sentiment", num_top_classes=2)
+
+    btn.click(fn=analyze, inputs=text_input, outputs=output)
+    gr.Examples(examples=EXAMPLES, inputs=text_input)
+
+    gr.Markdown(
+        """
+    ---
+    Built with [DistilBERT](https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english) | [arinbalyan](https://huggingface.co/arinbalyan)
+    """
+    )
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,3 @@
+# ponytail: pinned to what HF Spaces already has
+transformers>=4.30
+torch>=2.0