added docker config and optimized model for HF Spaces

.gitattributes

@@ -4,3 +4,4 @@
 *.onnx filter=lfs diff=lfs merge=lfs -text
 *.pth filter=lfs diff=lfs merge=lfs -text
 *.pt filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -5,13 +5,17 @@ __pycache__/
 *.pyd
 .venv/
 
-# Model Files (DO NOT COMMIT)
+# Model Files (DO NOT COMMIT heavy weights)
 models/hub/
 models/wav2vec2-finetuned/
 models/*.pth
 models/*.pt
 models/*.safetensors
 models/*.bin
+
+# Allow ONLY the optimized engine for deployment
+!models/onnx_quantized/
+
 models/
 # Derived Data (re-generate with scripts)
 data/embeddings/

Dockerfile

@@ -0,0 +1,48 @@
+# Use a modern Python base image
+FROM python:3.12-slim
+
+# Set environment variables
+ENV PYTHONUNBUFFERED=1 \
+    PYTHONDONTWRITEBYTECODE=1 \
+    PORT=8501
+
+WORKDIR /app
+
+# Install system dependencies (FFMPEG is critical for librosa)
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    git \
+    ffmpeg \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install uv for fast dependency management
+COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/
+
+# Copy configuration files
+COPY pyproject.toml uv.lock ./
+
+# Install dependencies into the system site-packages (simpler for Docker)
+RUN uv pip install --system -r pyproject.toml
+
+# Copy project files
+COPY src/ ./src/
+COPY models/ ./models/
+
+# Create a non-root user for security (Hugging Face recommendation)
+RUN useradd -m -u 1000 user
+USER user
+ENV HOME=/home/user \
+    PATH=/home/user/.local/bin:$PATH
+
+WORKDIR $HOME/app
+COPY --chown=user . $HOME/app
+
+# Expose the HF default port
+EXPOSE 8501
+
+# Healthcheck
+HEALTHCHECK CMD curl --fail http://localhost:7860/_stcore/health
+
+# Launch the standalone app
+ENTRYPOINT ["streamlit", "run", "src/ui/app_standalone.py", "--server.port=8501", "--server.address=0.0.0.0"]

README.md

@@ -1,3 +1,19 @@
+---
+title: VigilAudio
+emoji: 🎙️
+colorFrom: red
+colorTo: gray
+sdk: docker
+app_port: 8501
+tags:
+  - streamlit
+  - onnx
+  - audio-classification
+pinned: false
+short_description: Emotional audio classification
+license: mit
+---
+
 # VigilAudio: AI-Powered Audio Moderation Engine
 
 **A production-ready audio emotion classification system built for content moderation.**

models/onnx_quantized/config.json

@@ -0,0 +1,127 @@
+{
+  "activation_dropout": 0.1,
+  "adapter_attn_dim": null,
+  "adapter_kernel_size": 3,
+  "adapter_stride": 2,
+  "add_adapter": false,
+  "apply_spec_augment": true,
+  "architectures": [
+    "Wav2Vec2ForSequenceClassification"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token_id": 1,
+  "classifier_proj_size": 256,
+  "codevector_dim": 256,
+  "contrastive_logits_temperature": 0.1,
+  "conv_bias": false,
+  "conv_dim": [
+    512,
+    512,
+    512,
+    512,
+    512,
+    512,
+    512
+  ],
+  "conv_kernel": [
+    10,
+    3,
+    3,
+    3,
+    3,
+    2,
+    2
+  ],
+  "conv_stride": [
+    5,
+    2,
+    2,
+    2,
+    2,
+    2,
+    2
+  ],
+  "ctc_loss_reduction": "sum",
+  "ctc_zero_infinity": false,
+  "diversity_loss_weight": 0.1,
+  "do_stable_layer_norm": false,
+  "dtype": "float32",
+  "eos_token_id": 2,
+  "feat_extract_activation": "gelu",
+  "feat_extract_dropout": 0.0,
+  "feat_extract_norm": "group",
+  "feat_proj_dropout": 0.1,
+  "feat_quantizer_dropout": 0.0,
+  "final_dropout": 0.1,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout": 0.1,
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "id2label": {
+    "0": "angry",
+    "1": "disgusted",
+    "2": "fearful",
+    "3": "happy",
+    "4": "neutral",
+    "5": "sad",
+    "6": "suprised"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "label2id": {
+    "angry": 0,
+    "disgusted": 1,
+    "fearful": 2,
+    "happy": 3,
+    "neutral": 4,
+    "sad": 5,
+    "suprised": 6
+  },
+  "layer_norm_eps": 1e-05,
+  "layerdrop": 0.1,
+  "mask_feature_length": 10,
+  "mask_feature_min_masks": 0,
+  "mask_feature_prob": 0.0,
+  "mask_time_length": 10,
+  "mask_time_min_masks": 2,
+  "mask_time_prob": 0.05,
+  "model_type": "wav2vec2",
+  "num_adapter_layers": 3,
+  "num_attention_heads": 12,
+  "num_codevector_groups": 2,
+  "num_codevectors_per_group": 320,
+  "num_conv_pos_embedding_groups": 16,
+  "num_conv_pos_embeddings": 128,
+  "num_feat_extract_layers": 7,
+  "num_hidden_layers": 12,
+  "num_negatives": 100,
+  "output_hidden_size": 768,
+  "pad_token_id": 0,
+  "proj_codevector_dim": 256,
+  "tdnn_dilation": [
+    1,
+    2,
+    3,
+    1,
+    1
+  ],
+  "tdnn_dim": [
+    512,
+    512,
+    512,
+    512,
+    1500
+  ],
+  "tdnn_kernel": [
+    5,
+    3,
+    3,
+    1,
+    1
+  ],
+  "transformers_version": "4.57.3",
+  "use_weighted_layer_sum": false,
+  "vocab_size": 32,
+  "xvector_output_dim": 512
+}

models/onnx_quantized/model_quantized.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1a7fbc3bca7e8210afacc795c3913e3293117e083d50979b59df1850c965ee77
+size 122141352

models/onnx_quantized/ort_config.json

@@ -0,0 +1,32 @@
+{
+  "one_external_file": true,
+  "opset": null,
+  "optimization": {},
+  "quantization": {
+    "activations_dtype": "QUInt8",
+    "activations_symmetric": false,
+    "format": "QOperator",
+    "is_static": false,
+    "mode": "IntegerOps",
+    "nodes_to_exclude": [],
+    "nodes_to_quantize": [],
+    "operators_to_quantize": [
+      "MatMul",
+      "Attention",
+      "LSTM",
+      "Gather",
+      "Transpose",
+      "EmbedLayerNormalization"
+    ],
+    "per_channel": false,
+    "qdq_add_pair_to_weight": false,
+    "qdq_dedicated_pair": false,
+    "qdq_op_type_per_channel_support_to_axis": {
+      "MatMul": 1
+    },
+    "reduce_range": false,
+    "weights_dtype": "QInt8",
+    "weights_symmetric": true
+  },
+  "use_external_data_format": false
+}

models/onnx_quantized/preprocessor_config.json

@@ -0,0 +1,9 @@
+{
+  "do_normalize": true,
+  "feature_extractor_type": "Wav2Vec2FeatureExtractor",
+  "feature_size": 1,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": false,
+  "sampling_rate": 16000
+}