Deploy space_sentiment from monorepo

README.md

@@ -1,13 +1,23 @@
 ---
-title: Sentiment
-emoji: 😻
-colorFrom: pink
-colorTo: blue
+title: ONNX Sentiment Demo
+emoji: 💬
+colorFrom: yellow
+colorTo: red
 sdk: gradio
-sdk_version: 6.15.2
-python_version: '3.13'
+sdk_version: 5.23.1
+python_version: "3.12"
 app_file: app.py
 pinned: false
+license: mit
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# ONNX Sentiment Demo
+
+Binary sentiment classifier trained **from scratch** in PyTorch (bag-of-words + linear layer), exported to ONNX, served with **ONNX Runtime**.
+
+Built from `spaces/space_sentiment/` in the monorepo. Retrain and re-export with:
+
+```bash
+python train.py
+python export_onnx.py
+```

app.py

@@ -0,0 +1,49 @@
+"""Gradio demo: bag-of-words sentiment via ONNX Runtime."""
+
+import json
+from pathlib import Path
+
+import gradio as gr
+import numpy as np
+import onnxruntime as ort
+
+from preprocess import text_to_bow
+
+VOCAB_PATH = Path(__file__).resolve().parent / "vocab.json"
+vocab_data = json.loads(VOCAB_PATH.read_text())
+VOCAB: list[str] = vocab_data["vocab"]
+LABELS: list[str] = vocab_data["labels"]
+
+session = ort.InferenceSession("sentiment.onnx")
+
+
+def predict(text: str) -> dict[str, float]:
+    if not text.strip():
+        return {label: 0.0 for label in LABELS}
+
+    bow = text_to_bow(text, VOCAB)[np.newaxis, :]
+    logits = session.run(None, {"bow": bow})[0][0]
+    exp_logits = np.exp(logits - logits.max())
+    probs = exp_logits / exp_logits.sum()
+
+    return {label: float(prob) for label, prob in zip(LABELS, probs)}
+
+
+demo = gr.Interface(
+    fn=predict,
+    inputs=gr.Textbox(label="Text", placeholder="I love this product"),
+    outputs=gr.Label(num_top_classes=2, label="Sentiment"),
+    title="ONNX Sentiment Demo",
+    description=(
+        "Tiny bag-of-words classifier trained from scratch in PyTorch "
+        "(no pretrained model). Exported to ONNX for inference."
+    ),
+    examples=[
+        ["I love this"],
+        ["This is awful"],
+        ["Amazing product"],
+        ["So disappointed"],
+    ],
+)
+
+demo.launch()

preprocess.py

@@ -0,0 +1,43 @@
+"""Bag-of-words preprocessing for tiny sentiment classifier."""
+
+from __future__ import annotations
+
+import json
+import re
+from pathlib import Path
+
+import numpy as np
+
+VOCAB_PATH = Path(__file__).resolve().parent / "vocab.json"
+
+TOKEN_PATTERN = re.compile(r"[a-z]+")
+
+
+def tokenize(text: str) -> list[str]:
+    return TOKEN_PATTERN.findall(text.lower())
+
+
+def build_vocab(texts: list[str]) -> list[str]:
+    words: set[str] = set()
+    for text in texts:
+        words.update(tokenize(text))
+    return sorted(words)
+
+
+def text_to_bow(text: str, vocab: list[str]) -> np.ndarray:
+    word_to_idx = {word: idx for idx, word in enumerate(vocab)}
+    bow = np.zeros(len(vocab), dtype=np.float32)
+    for token in tokenize(text):
+        idx = word_to_idx.get(token)
+        if idx is not None:
+            bow[idx] = 1.0
+    return bow
+
+
+def save_vocab(vocab: list[str], labels: list[str]) -> None:
+    VOCAB_PATH.write_text(json.dumps({"vocab": vocab, "labels": labels}, indent=2))
+
+
+def load_vocab() -> tuple[list[str], list[str]]:
+    data = json.loads(VOCAB_PATH.read_text())
+    return data["vocab"], data["labels"]

requirements.txt

@@ -0,0 +1,3 @@
+# Runtime deps for the HF Space (gradio is provided by the Space SDK).
+onnxruntime>=1.18.0
+numpy>=1.26.0

sentiment.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03b106867260d989f12eab99a18bfd962b77745f65041b7a042f16648dec2108
+size 571

vocab.json

@@ -0,0 +1,29 @@
+{
+  "vocab": [
+    "amazing",
+    "awful",
+    "best",
+    "day",
+    "disappointed",
+    "ever",
+    "experience",
+    "great",
+    "happy",
+    "hate",
+    "i",
+    "is",
+    "it",
+    "love",
+    "product",
+    "purchase",
+    "so",
+    "terrible",
+    "this",
+    "with",
+    "worst"
+  ],
+  "labels": [
+    "negative",
+    "positive"
+  ]
+}