deepinfant_training.ipynb

{
    "cells": [
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "# DeepInfant: Infant Cry Classification\n",
       "\n",
       "This notebook demonstrates how to train and evaluate the DeepInfant model for infant cry classification using HuggingFace integration."
      ]
     },
     {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
       "import torch\n",
       "import matplotlib.pyplot as plt\n",
       "from train import DeepInfantModel, DeepInfantDataset\n",
       "from torch.utils.data import DataLoader\n",
       "import seaborn as sns\n",
       "from sklearn.metrics import confusion_matrix\n",
       "import numpy as np\n",
       "from train_huggingface import DeepInfantHFModel, DeepInfantHFDataset"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "## 1. HuggingFace Dataset Preparation"
      ]
     },
     {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
       "from datasets import load_dataset, Audio\n",
       "\n",
       "# Load dataset from HuggingFace (replace with your dataset)\n",
       "dataset = load_dataset(\"your_username/your_dataset\")\n",
       "dataset = dataset.cast_column(\"audio\", Audio(sampling_rate=16000))\n",
       "\n",
       "print(f\"Training samples: {len(dataset['train'])}\")\n",
       "print(f\"Validation samples: {len(dataset['validation'])}\")"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "## 2. Model Configuration"
      ]
     },
     {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
       "from transformers import TrainingArguments, Trainer, AutoConfig\n",
       "\n",
       "# Configure model\n",
       "config = AutoConfig.from_pretrained('bert-base-uncased')\n",
       "config.num_labels = 5  # Number of cry classifications\n",
       "model = DeepInfantHFModel(config)\n",
       "\n",
       "# Define training arguments\n",
       "training_args = TrainingArguments(\n",
       "    output_dir=\"./results\",\n",
       "    num_train_epochs=50,\n",
       "    per_device_train_batch_size=32,\n",
       "    per_device_eval_batch_size=32,\n",
       "    evaluation_strategy=\"epoch\",\n",
       "    save_strategy=\"epoch\",\n",
       "    load_deep_infant_at_end=True,\n",
       "    push_to_hub=True,\n",
       "    logging_dir='./logs',\n",
       ")"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "## 3. Training"
      ]
     },
     {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
       "# Prepare datasets\n",
       "train_dataset = DeepInfantHFDataset(dataset['train'])\n",
       "eval_dataset = DeepInfantHFDataset(dataset['validation'])\n",
       "\n",
       "# Initialize trainer\n",
       "trainer = Trainer(\n",
       "    model=model,\n",
       "    args=training_args,\n",
       "    train_dataset=train_dataset,\n",
       "    eval_dataset=eval_dataset,\n",
       ")\n",
       "\n",
       "# Train model\n",
       "trainer.train()"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "## 4. Evaluation"
      ]
     },
     {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
       "# Evaluate the model\n",
       "eval_results = trainer.evaluate()\n",
       "print(\"Evaluation Results:\")\n",
       "for key, value in eval_results.items():\n",
       "    print(f\"{key}: {value:.4f}\")"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "## 5. Push to HuggingFace Hub"
      ]
     },
     {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
       "# Push the model to HuggingFace Hub\n",
       "trainer.push_to_hub()"
      ]
     },
     {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
       "## 6. Test Predictions"
      ]
     },
     {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
       "from predict import InfantCryPredictor\n",
       "\n",
       "# Load the trained model\n",
       "predictor = InfantCryPredictor('results/checkpoint-deepinfant')\n",
       "\n",
       "# Test on a single file\n",
       "test_file = \"path/to/test/audio.wav\"\n",
       "label, confidence = predictor.predict(test_file)\n",
       "print(f\"Prediction: {label} (Confidence: {confidence:.2%})\")"
      ]
     }
    ],
    "metadata": {
     "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
     },
     "language_info": {
      "codemirror_mode": {
       "name": "ipython",
       "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.0"
     }
    },
    "nbformat": 4,
    "nbformat_minor": 4
   }