Max005/DeepfakeDetection
0
1{2 "cells": [3 {4 "cell_type": "markdown",5 "metadata": {},6 "source": [7 "## Import data"8 ]9 },10 {11 "cell_type": "code",12 "execution_count": 2,13 "metadata": {},14 "outputs": [15 {16 "name": "stdout",17 "output_type": "stream",18 "text": [19 "Collecting torchaudio\n",20 " Downloading torchaudio-2.5.1-cp312-cp312-win_amd64.whl.metadata (6.5 kB)\n",21 "Requirement already satisfied: torch==2.5.1 in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torchaudio) (2.5.1)\n",22 "Requirement already satisfied: filelock in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torch==2.5.1->torchaudio) (3.13.1)\n",23 "Requirement already satisfied: typing-extensions>=4.8.0 in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torch==2.5.1->torchaudio) (4.11.0)\n",24 "Requirement already satisfied: networkx in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torch==2.5.1->torchaudio) (3.3)\n",25 "Requirement already satisfied: jinja2 in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torch==2.5.1->torchaudio) (3.1.4)\n",26 "Requirement already satisfied: fsspec in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torch==2.5.1->torchaudio) (2024.6.1)\n",27 "Requirement already satisfied: setuptools in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torch==2.5.1->torchaudio) (75.1.0)\n",28 "Requirement already satisfied: sympy==1.13.1 in c:\\users\\asus\\anaconda3\\lib\\site-packages (from torch==2.5.1->torchaudio) (1.13.1)\n",29 "Requirement already satisfied: mpmath<1.4,>=1.1.0 in c:\\users\\asus\\anaconda3\\lib\\site-packages (from sympy==1.13.1->torch==2.5.1->torchaudio) (1.3.0)\n",30 "Requirement already satisfied: MarkupSafe>=2.0 in c:\\users\\asus\\anaconda3\\lib\\site-packages (from jinja2->torch==2.5.1->torchaudio) (2.1.3)\n",31 "Downloading torchaudio-2.5.1-cp312-cp312-win_amd64.whl (2.4 MB)\n",32 " ---------------------------------------- 0.0/2.4 MB ? eta -:--:--\n",33 " ---------------------------------------- 2.4/2.4 MB 11.6 MB/s eta 0:00:00\n",34 "Installing collected packages: torchaudio\n",35 "Successfully installed torchaudio-2.5.1\n",36 "Note: you may need to restart the kernel to use updated packages.\n"37 ]38 }39 ],40 "source": [41 "pip install torchaudio"42 ]43 },44 {45 "cell_type": "code",46 "execution_count": null,47 "metadata": {},48 "outputs": [49 {50 "ename": "ModuleNotFoundError",51 "evalue": "No module named 'datasets'",52 "output_type": "error",53 "traceback": [54 "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",55 "\u001b[1;31mModuleNotFoundError\u001b[0m Traceback (most recent call last)",56 "Cell \u001b[1;32mIn[3], line 5\u001b[0m\n\u001b[0;32m 3\u001b[0m \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01mos\u001b[39;00m\n\u001b[0;32m 4\u001b[0m \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01mtorchaudio\u001b[39;00m\n\u001b[1;32m----> 5\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01mdatasets\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m DatasetDict, load_dataset\n\u001b[0;32m 7\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mprepare_dataset\u001b[39m(directory):\n\u001b[0;32m 8\u001b[0m data \u001b[38;5;241m=\u001b[39m {\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mpath\u001b[39m\u001b[38;5;124m\"\u001b[39m: [], \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mlabel\u001b[39m\u001b[38;5;124m\"\u001b[39m: []}\n",57 "\u001b[1;31mModuleNotFoundError\u001b[0m: No module named 'datasets'"58 ]59 },60 {61 "ename": "",62 "evalue": "",63 "output_type": "error",64 "traceback": [65 "\u001b[1;31mThe Kernel crashed while executing code in the current cell or a previous cell. \n",66 "\u001b[1;31mPlease review the code in the cell(s) to identify a possible cause of the failure. \n",67 "\u001b[1;31mClick <a href='https://aka.ms/vscodeJupyterKernelCrash'>here</a> for more info. \n",68 "\u001b[1;31mView Jupyter <a href='command:jupyter.viewOutput'>log</a> for further details."69 ]70 }71 ],72 "source": [73 "## The data is not pushed to repo, only model and training logs etc are uploaded\n",74 "\n",75 "import os\n",76 "import torchaudio\n",77 "from datasets import DatasetDict, load_dataset\n",78 "\n",79 "def prepare_dataset(directory):\n",80 " data = {\"path\": [], \"label\": []}\n",81 " labels = {\"fake\": 0, \"real\": 1} # Map fake to 0 and real to 1\n",82 "\n",83 " for label, label_id in labels.items():\n",84 " folder_path = os.path.join(directory, label)\n",85 " for file in os.listdir(folder_path):\n",86 " if file.endswith(\".wav\"):\n",87 " data[\"path\"].append(os.path.join(folder_path, file))\n",88 " data[\"label\"].append(label_id)\n",89 " return data\n",90 "\n",91 "# Prepare train, validation, and test datasets\n",92 "train_data = prepare_dataset(r\"dataset\\for-norm\\for-norm\\testing\")\n",93 "val_data = prepare_dataset(r\"dataset\\for-norm\\for-norm\\testing\")\n",94 "test_data = prepare_dataset(r\"dataset\\for-norm\\for-norm\\testing\")\n"95 ]96 },97 {98 "cell_type": "code",99 "execution_count": 4,100 "metadata": {},101 "outputs": [],102 "source": [103 "from datasets import Dataset\n",104 "\n",105 "train_dataset = Dataset.from_dict(train_data)\n",106 "val_dataset = Dataset.from_dict(val_data)\n",107 "test_dataset = Dataset.from_dict(test_data)\n",108 "\n",109 "dataset = DatasetDict({\"train\": train_dataset, \"validation\": val_dataset, \"test\": test_dataset})\n"110 ]111 },112 {113 "cell_type": "markdown",114 "metadata": {},115 "source": [116 "## Import Model"117 ]118 },119 {120 "cell_type": "code",121 "execution_count": 5,122 "metadata": {},123 "outputs": [124 {125 "name": "stderr",126 "output_type": "stream",127 "text": [128 "c:\\Users\\60165\\AppData\\Local\\Programs\\Python\\Python39\\lib\\site-packages\\transformers\\configuration_utils.py:302: UserWarning: Passing `gradient_checkpointing` to a config initialization is deprecated and will be removed in v5 Transformers. Using `model.gradient_checkpointing_enable()` instead, or if you are using the `Trainer` API, pass `gradient_checkpointing=True` in your `TrainingArguments`.\n",129 " warnings.warn(\n",130 "Some weights of Wav2Vec2ForSequenceClassification were not initialized from the model checkpoint at facebook/wav2vec2-base and are newly initialized: ['classifier.bias', 'classifier.weight', 'projector.bias', 'projector.weight']\n",131 "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"132 ]133 }134 ],135 "source": [136 "from transformers import AutoProcessor\n",137 "from transformers import AutoModelForAudioClassification\n",138 "\n",139 "# Initialize processor\n",140 "model_name = \"facebook/wav2vec2-base\" # Replace with your model if different\n",141 "model = AutoModelForAudioClassification.from_pretrained(model_name, num_labels=2) # Adjust `num_labels` based on your dataset\n",142 "processor = AutoProcessor.from_pretrained(model_name)\n"143 ]144 },145 {146 "cell_type": "markdown",147 "metadata": {},148 "source": [149 "## Preprocess Data"150 ]151 },152 {153 "cell_type": "code",154 "execution_count": null,155 "metadata": {},156 "outputs": [157 {158 "data": {159 "application/vnd.jupyter.widget-view+json": {160 "model_id": "af2c5f31f0db43ee9975023b28e2c57c",161 "version_major": 2,162 "version_minor": 0163 },164 "text/plain": [165 "Map: 0%| | 0/4634 [00:00<?, ? examples/s]"166 ]167 },168 "metadata": {},169 "output_type": "display_data"170 },171 {172 "data": {173 "application/vnd.jupyter.widget-view+json": {174 "model_id": "051c3226b30145f3adbc708cf8afd6a5",175 "version_major": 2,176 "version_minor": 0177 },178 "text/plain": [179 "Map: 0%| | 0/4634 [00:00<?, ? examples/s]"180 ]181 },182 "metadata": {},183 "output_type": "display_data"184 },185 {186 "data": {187 "application/vnd.jupyter.widget-view+json": {188 "model_id": "86de10f37b7441e3a9c9f0ee88bf5149",189 "version_major": 2,190 "version_minor": 0191 },192 "text/plain": [193 "Map: 0%| | 0/4634 [00:00<?, ? examples/s]"194 ]195 },196 "metadata": {},197 "output_type": "display_data"198 },199 {200 "name": "stdout",201 "output_type": "stream",202 "text": [203 "tensor(0) <class 'torch.Tensor'>\n",204 "torch.int64\n"205 ]206 }207 ],208 "source": [209 "import torch\n",210 "\n",211 "\n",212 "def preprocess_function(batch):\n",213 " audio = torchaudio.load(batch[\"path\"])[0].squeeze().numpy()\n",214 " inputs = processor(\n",215 " audio,\n",216 " sampling_rate=16000,\n",217 " padding=True,\n",218 " truncation=True,\n",219 " max_length=32000, \n",220 " return_tensors=\"pt\"\n",221 " )\n",222 " batch[\"input_values\"] = inputs.input_values[0]\n",223 " # Ensure labels are converted to LongTensor\n",224 " batch[\"label\"] = torch.tensor(batch[\"label\"], dtype=torch.long) # Convert label to LongTensor\n",225 " return batch\n",226 "\n",227 "processed_dataset = dataset.map(preprocess_function, remove_columns=[\"path\"], batched=False)\n",228 "# Set format to torch tensors for compatibility with PyTorch\n",229 "processed_dataset.set_format(type=\"torch\", columns=[\"input_values\", \"label\"])\n",230 "\n",231 "# Double-check the label type again\n",232 "print(processed_dataset[\"train\"][0][\"label\"], type(processed_dataset[\"train\"][0][\"label\"]))\n",233 "print(processed_dataset[\"train\"][0][\"label\"].dtype) # Should print torch.int64\n"234 ]235 },236 {237 "cell_type": "markdown",238 "metadata": {},239 "source": [240 "## Map Training Labels"241 ]242 },243 {244 "cell_type": "code",245 "execution_count": 12,246 "metadata": {},247 "outputs": [248 {249 "name": "stdout",250 "output_type": "stream",251 "text": [252 "Labels: {0: 'Fake', 1: 'Real'}\n",253 "Labels: {0: 'Fake', 1: 'Real'}\n"254 ]255 }256 ],257 "source": [258 "# Ensure labels are in numerical format (e.g., 0, 1)\n",259 "id2label = {0: \"Fake\", 1: \"Real\"} # Define the mapping based on your dataset\n",260 "label2id = {v: k for k, v in id2label.items()} # Reverse mapping\n",261 "\n",262 "\n",263 "print(\"Labels:\", id2label)\n",264 "\n",265 "# Update the model's configuration with labels\n",266 "model.config.id2label = id2label\n",267 "model.config.label2id = label2id\n",268 "\n",269 "print(\"Labels:\", model.config.id2label) # Verify\n"270 ]271 },272 {273 "cell_type": "code",274 "execution_count": null,275 "metadata": {},276 "outputs": [],277 "source": [278 "from transformers import DataCollatorWithPadding\n",279 "\n",280 "# Use the processor's tokenizer for padding\n",281 "data_collator = DataCollatorWithPadding(tokenizer=processor, padding=True)\n"282 ]283 },284 {285 "cell_type": "markdown",286 "metadata": {},287 "source": [288 "## Initialize Training Arguments"289 ]290 },291 {292 "cell_type": "code",293 "execution_count": 8,294 "metadata": {},295 "outputs": [296 {297 "name": "stdout",298 "output_type": "stream",299 "text": [300 "TrainingArguments initialized successfully!\n"301 ]302 },303 {304 "name": "stderr",305 "output_type": "stream",306 "text": [307 "c:\\Users\\60165\\AppData\\Local\\Programs\\Python\\Python39\\lib\\site-packages\\transformers\\training_args.py:1545: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead\n",308 " warnings.warn(\n"309 ]310 }311 ],312 "source": [313 "from transformers import TrainingArguments\n",314 "\n",315 "training_args = TrainingArguments(\n",316 " output_dir=\"./results\",\n",317 " evaluation_strategy=\"epoch\",\n",318 " save_strategy=\"epoch\",\n",319 " learning_rate=5e-5,\n",320 " per_device_train_batch_size=8,\n",321 " per_device_eval_batch_size=8,\n",322 " num_train_epochs=3,\n",323 " weight_decay=0.01,\n",324 " logging_dir=\"./logs\",\n",325 " logging_steps=10,\n",326 " save_total_limit=2,\n",327 " fp16=True, \n",328 " push_to_hub=False,\n",329 ")\n",330 "print(\"TrainingArguments initialized successfully!\")\n"331 ]332 },333 {334 "cell_type": "code",335 "execution_count": 9,336 "metadata": {},337 "outputs": [],338 "source": [339 "from transformers import Trainer\n",340 "\n",341 "trainer = Trainer(\n",342 " model=model,\n",343 " args=training_args,\n",344 " train_dataset=processed_dataset[\"train\"],\n",345 " eval_dataset=processed_dataset[\"validation\"],\n",346 " tokenizer=processor, # Required for the data collator\n",347 " data_collator=data_collator,\n",348 ")\n"349 ]350 },351 {352 "cell_type": "markdown",353 "metadata": {},354 "source": [355 "## Start Training"356 ]357 },358 {359 "cell_type": "code",360 "execution_count": 33,361 "metadata": {},362 "outputs": [363 {364 "data": {365 "application/vnd.jupyter.widget-view+json": {366 "model_id": "53044eac53174227a4eb19becbeb0bbf",367 "version_major": 2,368 "version_minor": 0369 },370 "text/plain": [371 " 0%| | 0/1740 [00:00<?, ?it/s]"372 ]373 },374 "metadata": {},375 "output_type": "display_data"376 },377 {378 "name": "stdout",379 "output_type": "stream",380 "text": [381 "{'loss': 0.6607, 'grad_norm': 2.5936832427978516, 'learning_rate': 4.971264367816092e-05, 'epoch': 0.02}\n",382 "{'loss': 0.4545, 'grad_norm': 5.04218864440918, 'learning_rate': 4.9425287356321845e-05, 'epoch': 0.03}\n",383 "{'loss': 0.1779, 'grad_norm': 0.8874927163124084, 'learning_rate': 4.913793103448276e-05, 'epoch': 0.05}\n",384 "{'loss': 0.0833, 'grad_norm': 0.40262681245803833, 'learning_rate': 4.885057471264368e-05, 'epoch': 0.07}\n",385 "{'loss': 0.0948, 'grad_norm': 0.5579108595848083, 'learning_rate': 4.85632183908046e-05, 'epoch': 0.09}\n",386 "{'loss': 0.0128, 'grad_norm': 0.1585635393857956, 'learning_rate': 4.827586206896552e-05, 'epoch': 0.1}\n",387 "{'loss': 0.0696, 'grad_norm': 0.12149885296821594, 'learning_rate': 4.798850574712644e-05, 'epoch': 0.12}\n",388 "{'loss': 0.0065, 'grad_norm': 0.09655608981847763, 'learning_rate': 4.770114942528736e-05, 'epoch': 0.14}\n",389 "{'loss': 0.0052, 'grad_norm': 0.08148041367530823, 'learning_rate': 4.741379310344828e-05, 'epoch': 0.16}\n",390 "{'loss': 0.0041, 'grad_norm': 0.07030971348285675, 'learning_rate': 4.7126436781609195e-05, 'epoch': 0.17}\n",391 "{'loss': 0.0034, 'grad_norm': 0.05784648284316063, 'learning_rate': 4.6839080459770116e-05, 'epoch': 0.19}\n",392 "{'loss': 0.0029, 'grad_norm': 0.04742524400353432, 'learning_rate': 4.655172413793104e-05, 'epoch': 0.21}\n",393 "{'loss': 0.0024, 'grad_norm': 0.04222293198108673, 'learning_rate': 4.626436781609196e-05, 'epoch': 0.22}\n",394 "{'loss': 0.0021, 'grad_norm': 0.039586298167705536, 'learning_rate': 4.597701149425287e-05, 'epoch': 0.24}\n",395 "{'loss': 0.0018, 'grad_norm': 0.034121569246053696, 'learning_rate': 4.5689655172413794e-05, 'epoch': 0.26}\n",396 "{'loss': 0.0016, 'grad_norm': 0.031423550099134445, 'learning_rate': 4.5402298850574716e-05, 'epoch': 0.28}\n",397 "{'loss': 0.0418, 'grad_norm': 0.02912888117134571, 'learning_rate': 4.511494252873563e-05, 'epoch': 0.29}\n",398 "{'loss': 0.0831, 'grad_norm': 0.027829233556985855, 'learning_rate': 4.482758620689655e-05, 'epoch': 0.31}\n",399 "{'loss': 0.2431, 'grad_norm': 0.10419639199972153, 'learning_rate': 4.454022988505747e-05, 'epoch': 0.33}\n",400 "{'loss': 0.0024, 'grad_norm': 0.046179670840501785, 'learning_rate': 4.4252873563218394e-05, 'epoch': 0.34}\n",401 "{'loss': 0.0729, 'grad_norm': 0.03943876922130585, 'learning_rate': 4.396551724137931e-05, 'epoch': 0.36}\n",402 "{'loss': 0.0806, 'grad_norm': 0.05223412811756134, 'learning_rate': 4.367816091954024e-05, 'epoch': 0.38}\n",403 "{'loss': 0.0023, 'grad_norm': 0.041366685181856155, 'learning_rate': 4.339080459770115e-05, 'epoch': 0.4}\n",404 "{'loss': 0.0019, 'grad_norm': 0.03711611405014992, 'learning_rate': 4.3103448275862066e-05, 'epoch': 0.41}\n",405 "{'loss': 0.0016, 'grad_norm': 0.030888166278600693, 'learning_rate': 4.2816091954022994e-05, 'epoch': 0.43}\n",406 "{'loss': 0.0015, 'grad_norm': 0.027964089065790176, 'learning_rate': 4.252873563218391e-05, 'epoch': 0.45}\n",407 "{'loss': 0.0013, 'grad_norm': 0.025312568992376328, 'learning_rate': 4.224137931034483e-05, 'epoch': 0.47}\n",408 "{'loss': 0.0012, 'grad_norm': 0.02383635751903057, 'learning_rate': 4.195402298850575e-05, 'epoch': 0.48}\n",409 "{'loss': 0.0011, 'grad_norm': 0.021753674373030663, 'learning_rate': 4.166666666666667e-05, 'epoch': 0.5}\n",410 "{'loss': 0.001, 'grad_norm': 0.019886162132024765, 'learning_rate': 4.1379310344827587e-05, 'epoch': 0.52}\n",411 "{'loss': 0.0009, 'grad_norm': 0.018876325339078903, 'learning_rate': 4.109195402298851e-05, 'epoch': 0.53}\n",412 "{'loss': 0.0008, 'grad_norm': 0.017580321058630943, 'learning_rate': 4.080459770114943e-05, 'epoch': 0.55}\n",413 "{'loss': 0.0008, 'grad_norm': 0.015544956550002098, 'learning_rate': 4.0517241379310344e-05, 'epoch': 0.57}\n",414 "{'loss': 0.0007, 'grad_norm': 0.015221121720969677, 'learning_rate': 4.0229885057471265e-05, 'epoch': 0.59}\n",415 "{'loss': 0.0007, 'grad_norm': 0.014960471540689468, 'learning_rate': 3.9942528735632186e-05, 'epoch': 0.6}\n",416 "{'loss': 0.0006, 'grad_norm': 0.013560828752815723, 'learning_rate': 3.965517241379311e-05, 'epoch': 0.62}\n",417 "{'loss': 0.0006, 'grad_norm': 0.013700570911169052, 'learning_rate': 3.936781609195402e-05, 'epoch': 0.64}\n",418 "{'loss': 0.0006, 'grad_norm': 0.011968374252319336, 'learning_rate': 3.908045977011495e-05, 'epoch': 0.66}\n",419 "{'loss': 0.0005, 'grad_norm': 0.011644795536994934, 'learning_rate': 3.8793103448275865e-05, 'epoch': 0.67}\n",420 "{'loss': 0.0005, 'grad_norm': 0.011345883831381798, 'learning_rate': 3.850574712643678e-05, 'epoch': 0.69}\n",421 "{'loss': 0.0005, 'grad_norm': 0.010393058881163597, 'learning_rate': 3.82183908045977e-05, 'epoch': 0.71}\n",422 "{'loss': 0.0005, 'grad_norm': 0.010386484675109386, 'learning_rate': 3.793103448275862e-05, 'epoch': 0.72}\n",423 "{'loss': 0.0004, 'grad_norm': 0.009744665585458279, 'learning_rate': 3.764367816091954e-05, 'epoch': 0.74}\n",424 "{'loss': 0.0004, 'grad_norm': 0.009590468369424343, 'learning_rate': 3.735632183908046e-05, 'epoch': 0.76}\n",425 "{'loss': 0.0004, 'grad_norm': 0.009154150262475014, 'learning_rate': 3.7068965517241385e-05, 'epoch': 0.78}\n",426 "{'loss': 0.0004, 'grad_norm': 0.008997919037938118, 'learning_rate': 3.67816091954023e-05, 'epoch': 0.79}\n",427 "{'loss': 0.0004, 'grad_norm': 0.008509515784680843, 'learning_rate': 3.649425287356322e-05, 'epoch': 0.81}\n",428 "{'loss': 0.0004, 'grad_norm': 0.008223678916692734, 'learning_rate': 3.620689655172414e-05, 'epoch': 0.83}\n",429 "{'loss': 0.0003, 'grad_norm': 0.00758435670286417, 'learning_rate': 3.591954022988506e-05, 'epoch': 0.84}\n",430 "{'loss': 0.0003, 'grad_norm': 0.0074744271114468575, 'learning_rate': 3.563218390804598e-05, 'epoch': 0.86}\n",431 "{'loss': 0.0003, 'grad_norm': 0.007454875390976667, 'learning_rate': 3.53448275862069e-05, 'epoch': 0.88}\n",432 "{'loss': 0.0003, 'grad_norm': 0.007157924585044384, 'learning_rate': 3.505747126436782e-05, 'epoch': 0.9}\n",433 "{'loss': 0.0003, 'grad_norm': 0.006946589332073927, 'learning_rate': 3.4770114942528735e-05, 'epoch': 0.91}\n",434 "{'loss': 0.0003, 'grad_norm': 0.0067284563556313515, 'learning_rate': 3.4482758620689657e-05, 'epoch': 0.93}\n",435 "{'loss': 0.0003, 'grad_norm': 0.00652291439473629, 'learning_rate': 3.419540229885058e-05, 'epoch': 0.95}\n",436 "{'loss': 0.0003, 'grad_norm': 0.006468599662184715, 'learning_rate': 3.390804597701149e-05, 'epoch': 0.97}\n",437 "{'loss': 0.0003, 'grad_norm': 0.0061700050719082355, 'learning_rate': 3.3620689655172414e-05, 'epoch': 0.98}\n",438 "{'loss': 0.0002, 'grad_norm': 0.005980886053293943, 'learning_rate': 3.3333333333333335e-05, 'epoch': 1.0}\n"439 ]440 },441 {442 "data": {443 "application/vnd.jupyter.widget-view+json": {444 "model_id": "9fe6f453d3dc4644b7e7adf483e8064b",445 "version_major": 2,446 "version_minor": 0447 },448 "text/plain": [449 " 0%| | 0/580 [00:00<?, ?it/s]"450 ]451 },452 "metadata": {},453 "output_type": "display_data"454 },455 {456 "name": "stdout",457 "output_type": "stream",458 "text": [459 "{'eval_loss': 0.00017967642634175718, 'eval_runtime': 566.6055, 'eval_samples_per_second': 8.179, 'eval_steps_per_second': 1.024, 'epoch': 1.0}\n",460 "{'loss': 0.0002, 'grad_norm': 0.005443067755550146, 'learning_rate': 3.3045977011494256e-05, 'epoch': 1.02}\n",461 "{'loss': 0.0002, 'grad_norm': 0.005919346585869789, 'learning_rate': 3.275862068965517e-05, 'epoch': 1.03}\n",462 "{'loss': 0.0002, 'grad_norm': 0.00538916140794754, 'learning_rate': 3.24712643678161e-05, 'epoch': 1.05}\n",463 "{'loss': 0.0002, 'grad_norm': 0.00514333276078105, 'learning_rate': 3.218390804597701e-05, 'epoch': 1.07}\n",464 "{'loss': 0.0002, 'grad_norm': 0.005011783912777901, 'learning_rate': 3.1896551724137935e-05, 'epoch': 1.09}\n",465 "{'loss': 0.0002, 'grad_norm': 0.005112846381962299, 'learning_rate': 3.160919540229885e-05, 'epoch': 1.1}\n",466 "{'loss': 0.0002, 'grad_norm': 0.004895139951258898, 'learning_rate': 3.132183908045977e-05, 'epoch': 1.12}\n",467 "{'loss': 0.0002, 'grad_norm': 0.004565018694847822, 'learning_rate': 3.103448275862069e-05, 'epoch': 1.14}\n",468 "{'loss': 0.0002, 'grad_norm': 0.00477330107241869, 'learning_rate': 3.0747126436781606e-05, 'epoch': 1.16}\n",469 "{'loss': 0.0002, 'grad_norm': 0.004563583992421627, 'learning_rate': 3.045977011494253e-05, 'epoch': 1.17}\n",470 "{'loss': 0.0002, 'grad_norm': 0.004568100906908512, 'learning_rate': 3.017241379310345e-05, 'epoch': 1.19}\n",471 "{'loss': 0.0002, 'grad_norm': 0.0046887630596756935, 'learning_rate': 2.988505747126437e-05, 'epoch': 1.21}\n",472 "{'loss': 0.0002, 'grad_norm': 0.004261981230229139, 'learning_rate': 2.9597701149425288e-05, 'epoch': 1.22}\n",473 "{'loss': 0.0002, 'grad_norm': 0.004290647804737091, 'learning_rate': 2.9310344827586206e-05, 'epoch': 1.24}\n",474 "{'loss': 0.0002, 'grad_norm': 0.004188802093267441, 'learning_rate': 2.9022988505747127e-05, 'epoch': 1.26}\n",475 "{'loss': 0.0002, 'grad_norm': 0.003917949739843607, 'learning_rate': 2.8735632183908045e-05, 'epoch': 1.28}\n",476 "{'loss': 0.0002, 'grad_norm': 0.003940439783036709, 'learning_rate': 2.844827586206897e-05, 'epoch': 1.29}\n",477 "{'loss': 0.0002, 'grad_norm': 0.004128696396946907, 'learning_rate': 2.8160919540229884e-05, 'epoch': 1.31}\n",478 "{'loss': 0.0002, 'grad_norm': 0.004070794675499201, 'learning_rate': 2.787356321839081e-05, 'epoch': 1.33}\n",479 "{'loss': 0.0002, 'grad_norm': 0.0037206218112260103, 'learning_rate': 2.7586206896551727e-05, 'epoch': 1.34}\n",480 "{'loss': 0.0001, 'grad_norm': 0.00400934973731637, 'learning_rate': 2.7298850574712648e-05, 'epoch': 1.36}\n",481 "{'loss': 0.0001, 'grad_norm': 0.0037558332551270723, 'learning_rate': 2.7011494252873566e-05, 'epoch': 1.38}\n",482 "{'loss': 0.0001, 'grad_norm': 0.0035916264168918133, 'learning_rate': 2.672413793103448e-05, 'epoch': 1.4}\n",483 "{'loss': 0.0001, 'grad_norm': 0.003707454539835453, 'learning_rate': 2.6436781609195405e-05, 'epoch': 1.41}\n",484 "{'loss': 0.0001, 'grad_norm': 0.0034801277797669172, 'learning_rate': 2.6149425287356323e-05, 'epoch': 1.43}\n",485 "{'loss': 0.0001, 'grad_norm': 0.003501839004456997, 'learning_rate': 2.5862068965517244e-05, 'epoch': 1.45}\n",486 "{'loss': 0.0001, 'grad_norm': 0.003458078484982252, 'learning_rate': 2.5574712643678162e-05, 'epoch': 1.47}\n",487 "{'loss': 0.0001, 'grad_norm': 0.0031666585709899664, 'learning_rate': 2.5287356321839083e-05, 'epoch': 1.48}\n",488 "{'loss': 0.0001, 'grad_norm': 0.0033736126497387886, 'learning_rate': 2.5e-05, 'epoch': 1.5}\n",489 "{'loss': 0.0001, 'grad_norm': 0.003289664164185524, 'learning_rate': 2.4712643678160922e-05, 'epoch': 1.52}\n",490 "{'loss': 0.0001, 'grad_norm': 0.0031710772309452295, 'learning_rate': 2.442528735632184e-05, 'epoch': 1.53}\n",491 "{'loss': 0.0001, 'grad_norm': 0.0029777924064546824, 'learning_rate': 2.413793103448276e-05, 'epoch': 1.55}\n",492 "{'loss': 0.0001, 'grad_norm': 0.003144340356811881, 'learning_rate': 2.385057471264368e-05, 'epoch': 1.57}\n",493 "{'loss': 0.0001, 'grad_norm': 0.0029524925630539656, 'learning_rate': 2.3563218390804597e-05, 'epoch': 1.59}\n",494 "{'loss': 0.0001, 'grad_norm': 0.002912016585469246, 'learning_rate': 2.327586206896552e-05, 'epoch': 1.6}\n",495 "{'loss': 0.0001, 'grad_norm': 0.002869528019800782, 'learning_rate': 2.2988505747126437e-05, 'epoch': 1.62}\n",496 "{'loss': 0.0001, 'grad_norm': 0.002966447500512004, 'learning_rate': 2.2701149425287358e-05, 'epoch': 1.64}\n",497 "{'loss': 0.0001, 'grad_norm': 0.0027959959115833044, 'learning_rate': 2.2413793103448276e-05, 'epoch': 1.66}\n",498 "{'loss': 0.0001, 'grad_norm': 0.0030403095297515392, 'learning_rate': 2.2126436781609197e-05, 'epoch': 1.67}\n",499 "{'loss': 0.0001, 'grad_norm': 0.0026587999891489744, 'learning_rate': 2.183908045977012e-05, 'epoch': 1.69}\n",500 "{'loss': 0.0001, 'grad_norm': 0.002689346671104431, 'learning_rate': 2.1551724137931033e-05, 'epoch': 1.71}\n",501 "{'loss': 0.0001, 'grad_norm': 0.002710141707211733, 'learning_rate': 2.1264367816091954e-05, 'epoch': 1.72}\n",502 "{'loss': 0.0001, 'grad_norm': 0.002674366347491741, 'learning_rate': 2.0977011494252875e-05, 'epoch': 1.74}\n",503 "{'loss': 0.0001, 'grad_norm': 0.0026578502729535103, 'learning_rate': 2.0689655172413793e-05, 'epoch': 1.76}\n",504 "{'loss': 0.0001, 'grad_norm': 0.00243232655338943, 'learning_rate': 2.0402298850574715e-05, 'epoch': 1.78}\n",505 "{'loss': 0.0001, 'grad_norm': 0.0025773164816200733, 'learning_rate': 2.0114942528735632e-05, 'epoch': 1.79}\n",506 "{'loss': 0.0001, 'grad_norm': 0.0024439615663141012, 'learning_rate': 1.9827586206896554e-05, 'epoch': 1.81}\n",507 "{'loss': 0.0001, 'grad_norm': 0.0024733347818255424, 'learning_rate': 1.9540229885057475e-05, 'epoch': 1.83}\n",508 "{'loss': 0.0001, 'grad_norm': 0.002439699834212661, 'learning_rate': 1.925287356321839e-05, 'epoch': 1.84}\n",509 "{'loss': 0.0001, 'grad_norm': 0.0025980097707360983, 'learning_rate': 1.896551724137931e-05, 'epoch': 1.86}\n",510 "{'loss': 0.0001, 'grad_norm': 0.002387199318036437, 'learning_rate': 1.867816091954023e-05, 'epoch': 1.88}\n",511 "{'loss': 0.0001, 'grad_norm': 0.0023106117732822895, 'learning_rate': 1.839080459770115e-05, 'epoch': 1.9}\n",512 "{'loss': 0.0001, 'grad_norm': 0.0023344189394265413, 'learning_rate': 1.810344827586207e-05, 'epoch': 1.91}\n",513 "{'loss': 0.0001, 'grad_norm': 0.0023740960750728846, 'learning_rate': 1.781609195402299e-05, 'epoch': 1.93}\n",514 "{'loss': 0.0001, 'grad_norm': 0.002346088644117117, 'learning_rate': 1.752873563218391e-05, 'epoch': 1.95}\n",515 "{'loss': 0.0001, 'grad_norm': 0.002391340211033821, 'learning_rate': 1.7241379310344828e-05, 'epoch': 1.97}\n",516 "{'loss': 0.0001, 'grad_norm': 0.002250733319669962, 'learning_rate': 1.6954022988505746e-05, 'epoch': 1.98}\n",517 "{'loss': 0.0001, 'grad_norm': 0.002164299599826336, 'learning_rate': 1.6666666666666667e-05, 'epoch': 2.0}\n"518 ]519 },520 {521 "data": {522 "application/vnd.jupyter.widget-view+json": {523 "model_id": "ed9239dd5b4346509d243cbc378bcf44",524 "version_major": 2,525 "version_minor": 0526 },527 "text/plain": [528 " 0%| | 0/580 [00:00<?, ?it/s]"529 ]530 },531 "metadata": {},532 "output_type": "display_data"533 },534 {535 "name": "stdout",536 "output_type": "stream",537 "text": [538 "{'eval_loss': 6.0841484810225666e-05, 'eval_runtime': 581.8939, 'eval_samples_per_second': 7.964, 'eval_steps_per_second': 0.997, 'epoch': 2.0}\n",539 "{'loss': 0.0001, 'grad_norm': 0.0022026619408279657, 'learning_rate': 1.6379310344827585e-05, 'epoch': 2.02}\n",540 "{'loss': 0.0001, 'grad_norm': 0.002242449903860688, 'learning_rate': 1.6091954022988507e-05, 'epoch': 2.03}\n",541 "{'loss': 0.0001, 'grad_norm': 0.002464097458869219, 'learning_rate': 1.5804597701149425e-05, 'epoch': 2.05}\n",542 "{'loss': 0.0001, 'grad_norm': 0.0022003022022545338, 'learning_rate': 1.5517241379310346e-05, 'epoch': 2.07}\n",543 "{'loss': 0.0001, 'grad_norm': 0.0021785416174679995, 'learning_rate': 1.5229885057471265e-05, 'epoch': 2.09}\n",544 "{'loss': 0.0001, 'grad_norm': 0.0021638190373778343, 'learning_rate': 1.4942528735632185e-05, 'epoch': 2.1}\n",545 "{'loss': 0.0001, 'grad_norm': 0.0022439502645283937, 'learning_rate': 1.4655172413793103e-05, 'epoch': 2.12}\n",546 "{'loss': 0.0001, 'grad_norm': 0.0020717435982078314, 'learning_rate': 1.4367816091954022e-05, 'epoch': 2.14}\n",547 "{'loss': 0.0001, 'grad_norm': 0.0020531516056507826, 'learning_rate': 1.4080459770114942e-05, 'epoch': 2.16}\n",548 "{'loss': 0.0001, 'grad_norm': 0.0019899189937859774, 'learning_rate': 1.3793103448275863e-05, 'epoch': 2.17}\n",549 "{'loss': 0.0001, 'grad_norm': 0.0020303332712501287, 'learning_rate': 1.3505747126436783e-05, 'epoch': 2.19}\n",550 "{'loss': 0.0001, 'grad_norm': 0.0021102086175233126, 'learning_rate': 1.3218390804597702e-05, 'epoch': 2.21}\n",551 "{'loss': 0.0001, 'grad_norm': 0.0019932142458856106, 'learning_rate': 1.2931034482758622e-05, 'epoch': 2.22}\n",552 "{'loss': 0.0001, 'grad_norm': 0.002080111298710108, 'learning_rate': 1.2643678160919542e-05, 'epoch': 2.24}\n",553 "{'loss': 0.0001, 'grad_norm': 0.0020179597195237875, 'learning_rate': 1.2356321839080461e-05, 'epoch': 2.26}\n",554 "{'loss': 0.0001, 'grad_norm': 0.0019549003336578608, 'learning_rate': 1.206896551724138e-05, 'epoch': 2.28}\n",555 "{'loss': 0.0001, 'grad_norm': 0.0020865327678620815, 'learning_rate': 1.1781609195402299e-05, 'epoch': 2.29}\n",556 "{'loss': 0.0001, 'grad_norm': 0.0018828624160960317, 'learning_rate': 1.1494252873563218e-05, 'epoch': 2.31}\n",557 "{'loss': 0.0001, 'grad_norm': 0.0018662698566913605, 'learning_rate': 1.1206896551724138e-05, 'epoch': 2.33}\n",558 "{'loss': 0.0001, 'grad_norm': 0.001857285387814045, 'learning_rate': 1.091954022988506e-05, 'epoch': 2.34}\n",559 "{'loss': 0.0001, 'grad_norm': 0.001844724640250206, 'learning_rate': 1.0632183908045977e-05, 'epoch': 2.36}\n",560 "{'loss': 0.0001, 'grad_norm': 0.0017886353889480233, 'learning_rate': 1.0344827586206897e-05, 'epoch': 2.38}\n",561 "{'loss': 0.0001, 'grad_norm': 0.0019668207969516516, 'learning_rate': 1.0057471264367816e-05, 'epoch': 2.4}\n",562 "{'loss': 0.0001, 'grad_norm': 0.0018605877412483096, 'learning_rate': 9.770114942528738e-06, 'epoch': 2.41}\n",563 "{'loss': 0.0001, 'grad_norm': 0.0018027386395260692, 'learning_rate': 9.482758620689655e-06, 'epoch': 2.43}\n",564 "{'loss': 0.0001, 'grad_norm': 0.0018370413454249501, 'learning_rate': 9.195402298850575e-06, 'epoch': 2.45}\n",565 "{'loss': 0.0001, 'grad_norm': 0.0019249517936259508, 'learning_rate': 8.908045977011495e-06, 'epoch': 2.47}\n",566 "{'loss': 0.0001, 'grad_norm': 0.0019102703081443906, 'learning_rate': 8.620689655172414e-06, 'epoch': 2.48}\n",567 "{'loss': 0.0001, 'grad_norm': 0.0019130830187350512, 'learning_rate': 8.333333333333334e-06, 'epoch': 2.5}\n",568 "{'loss': 0.0001, 'grad_norm': 0.0019449306419119239, 'learning_rate': 8.045977011494253e-06, 'epoch': 2.52}\n",569 "{'loss': 0.0001, 'grad_norm': 0.001796119031496346, 'learning_rate': 7.758620689655173e-06, 'epoch': 2.53}\n",570 "{'loss': 0.0001, 'grad_norm': 0.0017440468072891235, 'learning_rate': 7.4712643678160925e-06, 'epoch': 2.55}\n",571 "{'loss': 0.0001, 'grad_norm': 0.0017786856042221189, 'learning_rate': 7.183908045977011e-06, 'epoch': 2.57}\n",572 "{'loss': 0.0001, 'grad_norm': 0.0018597355810925364, 'learning_rate': 6.896551724137932e-06, 'epoch': 2.59}\n",573 "{'loss': 0.0001, 'grad_norm': 0.0017648187931627035, 'learning_rate': 6.609195402298851e-06, 'epoch': 2.6}\n",574 "{'loss': 0.0001, 'grad_norm': 0.0017601278377696872, 'learning_rate': 6.321839080459771e-06, 'epoch': 2.62}\n",575 "{'loss': 0.0001, 'grad_norm': 0.0017502185655757785, 'learning_rate': 6.03448275862069e-06, 'epoch': 2.64}\n",576 "{'loss': 0.0001, 'grad_norm': 0.001753892400301993, 'learning_rate': 5.747126436781609e-06, 'epoch': 2.66}\n",577 "{'loss': 0.0001, 'grad_norm': 0.0016946644755080342, 'learning_rate': 5.45977011494253e-06, 'epoch': 2.67}\n",578 "{'loss': 0.0001, 'grad_norm': 0.001783599378541112, 'learning_rate': 5.172413793103448e-06, 'epoch': 2.69}\n",579 "{'loss': 0.0001, 'grad_norm': 0.0017759180627763271, 'learning_rate': 4.885057471264369e-06, 'epoch': 2.71}\n",580 "{'loss': 0.0001, 'grad_norm': 0.0017218819120898843, 'learning_rate': 4.5977011494252875e-06, 'epoch': 2.72}\n",581 "{'loss': 0.0001, 'grad_norm': 0.0016811942914500833, 'learning_rate': 4.310344827586207e-06, 'epoch': 2.74}\n",582 "{'loss': 0.0001, 'grad_norm': 0.0017582618165761232, 'learning_rate': 4.022988505747127e-06, 'epoch': 2.76}\n",583 "{'loss': 0.0001, 'grad_norm': 0.001848816522397101, 'learning_rate': 3.7356321839080462e-06, 'epoch': 2.78}\n",584 "{'loss': 0.0001, 'grad_norm': 0.0017523870337754488, 'learning_rate': 3.448275862068966e-06, 'epoch': 2.79}\n",585 "{'loss': 0.0001, 'grad_norm': 0.001707645715214312, 'learning_rate': 3.1609195402298854e-06, 'epoch': 2.81}\n",586 "{'loss': 0.0001, 'grad_norm': 0.0017925987485796213, 'learning_rate': 2.8735632183908046e-06, 'epoch': 2.83}\n",587 "{'loss': 0.0001, 'grad_norm': 0.001785592525266111, 'learning_rate': 2.586206896551724e-06, 'epoch': 2.84}\n",588 "{'loss': 0.0001, 'grad_norm': 0.0017369745764881372, 'learning_rate': 2.2988505747126437e-06, 'epoch': 2.86}\n",589 "{'loss': 0.0001, 'grad_norm': 0.0017363271908834577, 'learning_rate': 2.0114942528735633e-06, 'epoch': 2.88}\n",590 "{'loss': 0.0001, 'grad_norm': 0.0017762900097295642, 'learning_rate': 1.724137931034483e-06, 'epoch': 2.9}\n",591 "{'loss': 0.0001, 'grad_norm': 0.0017800360219553113, 'learning_rate': 1.4367816091954023e-06, 'epoch': 2.91}\n",592 "{'loss': 0.0001, 'grad_norm': 0.0016894094878807664, 'learning_rate': 1.1494252873563219e-06, 'epoch': 2.93}\n",593 "{'loss': 0.0001, 'grad_norm': 0.0016883889911696315, 'learning_rate': 8.620689655172415e-07, 'epoch': 2.95}\n",594 "{'loss': 0.0001, 'grad_norm': 0.0017332383431494236, 'learning_rate': 5.747126436781609e-07, 'epoch': 2.97}\n",595 "{'loss': 0.0001, 'grad_norm': 0.0018206291133537889, 'learning_rate': 2.8735632183908047e-07, 'epoch': 2.98}\n",596 "{'loss': 0.0001, 'grad_norm': 0.0017392894951626658, 'learning_rate': 0.0, 'epoch': 3.0}\n"597 ]598 },599 {600 "data": {601 "application/vnd.jupyter.widget-view+json": {602 "model_id": "de595459d7b642babd74e21ce354d064",603 "version_major": 2,604 "version_minor": 0605 },606 "text/plain": [607 " 0%| | 0/580 [00:00<?, ?it/s]"608 ]609 },610 "metadata": {},611 "output_type": "display_data"612 },613 {614 "name": "stdout",615 "output_type": "stream",616 "text": [617 "{'eval_loss': 4.472154250834137e-05, 'eval_runtime': 541.083, 'eval_samples_per_second': 8.564, 'eval_steps_per_second': 1.072, 'epoch': 3.0}\n",618 "{'train_runtime': 10729.5411, 'train_samples_per_second': 1.296, 'train_steps_per_second': 0.162, 'train_loss': 0.01232232698998762, 'epoch': 3.0}\n"619 ]620 },621 {622 "data": {623 "text/plain": [624 "TrainOutput(global_step=1740, training_loss=0.01232232698998762, metrics={'train_runtime': 10729.5411, 'train_samples_per_second': 1.296, 'train_steps_per_second': 0.162, 'total_flos': 2.5228134820702045e+17, 'train_loss': 0.01232232698998762, 'epoch': 3.0})"625 ]626 },627 "execution_count": 33,628 "metadata": {},629 "output_type": "execute_result"630 }631 ],632 "source": [633 "trainer.train()\n"634 ]635 },636 {637 "cell_type": "markdown",638 "metadata": {},639 "source": [640 "## Save The Model"641 ]642 },643 {644 "cell_type": "code",645 "execution_count": 34,646 "metadata": {},647 "outputs": [648 {649 "data": {650 "text/plain": [651 "[]"652 ]653 },654 "execution_count": 34,655 "metadata": {},656 "output_type": "execute_result"657 }658 ],659 "source": [660 "# Save the trained model and processor\n",661 "trainer.save_model(\"./trained_model\") # Saves the model to the specified directory\n",662 "processor.save_pretrained(\"./trained_model\") # Saves the processor as well\n"663 ]664 },665 {666 "cell_type": "code",667 "execution_count": 35,668 "metadata": {},669 "outputs": [670 {671 "data": {672 "application/vnd.jupyter.widget-view+json": {673 "model_id": "ae19dee506db46f4b39d402e43d16276",674 "version_major": 2,675 "version_minor": 0676 },677 "text/plain": [678 " 0%| | 0/580 [00:00<?, ?it/s]"679 ]680 },681 "metadata": {},682 "output_type": "display_data"683 },684 {685 "name": "stdout",686 "output_type": "stream",687 "text": [688 "Evaluation Metrics:\n",689 "eval_loss: 4.472154250834137e-05\n",690 "eval_runtime: 527.4045\n",691 "eval_samples_per_second: 8.786\n",692 "eval_steps_per_second: 1.1\n",693 "epoch: 3.0\n"694 ]695 }696 ],697 "source": [698 "# Evaluate the model on the validation dataset\n",699 "evaluation_metrics = trainer.evaluate()\n",700 "\n",701 "# Print the evaluation metrics\n",702 "print(\"Evaluation Metrics:\")\n",703 "for metric, value in evaluation_metrics.items():\n",704 " print(f\"{metric}: {value}\")\n"705 ]706 },707 {708 "cell_type": "code",709 "execution_count": 10,710 "metadata": {},711 "outputs": [],712 "source": [713 "from transformers import AutoProcessor, AutoModelForAudioClassification\n",714 "\n",715 "# Load the trained model and processor\n",716 "model_path = \"./trained_model\" # Path to your saved model\n",717 "model = AutoModelForAudioClassification.from_pretrained(model_path)\n",718 "processor = AutoProcessor.from_pretrained(model_path)\n"719 ]720 },721 {722 "cell_type": "markdown",723 "metadata": {},724 "source": [725 "## Single Audio Testing"726 ]727 },728 {729 "cell_type": "code",730 "execution_count": 34,731 "metadata": {},732 "outputs": [],733 "source": [734 "def prepare_audio(file_path, sampling_rate=16000, duration=10):\n",735 " \"\"\"\n",736 " Prepares audio by loading, resampling, and returning it in manageable chunks.\n",737 " \n",738 " Parameters:\n",739 " - file_path: Path to the audio file.\n",740 " - sampling_rate: Target sampling rate for the audio.\n",741 " - duration: Duration in seconds for each chunk.\n",742 " \n",743 " Returns:\n",744 " - A list of audio chunks, each as a numpy array.\n",745 " \"\"\"\n",746 " # Load and resample the audio file\n",747 " waveform, original_sampling_rate = torchaudio.load(file_path)\n",748 " \n",749 " # Convert stereo to mono if necessary\n",750 " if waveform.shape[0] > 1: # More than 1 channel\n",751 " waveform = torch.mean(waveform, dim=0, keepdim=True)\n",752 " \n",753 " # Resample if needed\n",754 " if original_sampling_rate != sampling_rate:\n",755 " resampler = torchaudio.transforms.Resample(orig_freq=original_sampling_rate, new_freq=sampling_rate)\n",756 " waveform = resampler(waveform)\n",757 " \n",758 " # Calculate chunk size in samples\n",759 " chunk_size = sampling_rate * duration\n",760 " audio_chunks = []\n",761 "\n",762 " # Split the audio into chunks\n",763 " for start in range(0, waveform.shape[1], chunk_size):\n",764 " chunk = waveform[:, start:start + chunk_size]\n",765 " \n",766 " # Pad the last chunk if it's shorter than the chunk size\n",767 " if chunk.shape[1] < chunk_size:\n",768 " padding = chunk_size - chunk.shape[1]\n",769 " chunk = torch.nn.functional.pad(chunk, (0, padding))\n",770 " \n",771 " audio_chunks.append(chunk.squeeze().numpy())\n",772 " \n",773 " return audio_chunks\n"774 ]775 },776 {777 "cell_type": "code",778 "execution_count": 35,779 "metadata": {},780 "outputs": [781 {782 "name": "stdout",783 "output_type": "stream",784 "text": [785 "Chunk shape: (160000,)\n",786 "Logits for chunk 1: tensor([[ 4.6742, -5.1778]])\n",787 "Chunk shape: (160000,)\n",788 "Logits for chunk 2: tensor([[ 4.7219, -5.2332]])\n",789 "Chunk shape: (160000,)\n",790 "Logits for chunk 3: tensor([[ 4.7545, -5.2641]])\n",791 "Chunk shape: (160000,)\n",792 "Logits for chunk 4: tensor([[ 4.6714, -5.1740]])\n",793 "Chunk shape: (160000,)\n",794 "Logits for chunk 5: tensor([[ 4.7660, -5.2743]])\n",795 "Chunk shape: (160000,)\n",796 "Logits for chunk 6: tensor([[ 4.7724, -5.2836]])\n",797 "Chunk shape: (160000,)\n",798 "Logits for chunk 7: tensor([[ 4.7268, -5.2362]])\n",799 "Chunk shape: (160000,)\n",800 "Logits for chunk 8: tensor([[ 4.6898, -5.1898]])\n",801 "Chunk shape: (160000,)\n",802 "Logits for chunk 9: tensor([[ 4.6646, -5.1708]])\n",803 "Chunk shape: (160000,)\n",804 "Logits for chunk 10: tensor([[ 4.5948, -5.0867]])\n",805 "Chunk shape: (160000,)\n",806 "Logits for chunk 11: tensor([[ 4.7512, -5.2579]])\n",807 "Chunk shape: (160000,)\n",808 "Logits for chunk 12: tensor([[-4.5599, 5.0363]])\n",809 "Chunk shape: (160000,)\n",810 "Logits for chunk 13: tensor([[-0.4980, 0.5546]])\n",811 "Chunk shape: (160000,)\n",812 "Logits for chunk 14: tensor([[ 4.7295, -5.2358]])\n",813 "Chunk shape: (160000,)\n",814 "Logits for chunk 15: tensor([[ 4.7426, -5.2534]])\n",815 "Chunk shape: (160000,)\n",816 "Logits for chunk 16: tensor([[ 1.9405, -2.1493]])\n",817 "Chunk shape: (160000,)\n",818 "Logits for chunk 17: tensor([[ 4.7168, -5.2235]])\n",819 "Chunk shape: (160000,)\n",820 "Logits for chunk 18: tensor([[ 4.6801, -5.1907]])\n",821 "Chunk shape: (160000,)\n",822 "Logits for chunk 19: tensor([[ 4.7454, -5.2568]])\n",823 "Chunk shape: (160000,)\n",824 "Logits for chunk 20: tensor([[ 4.7642, -5.2723]])\n",825 "Chunk shape: (160000,)\n",826 "Logits for chunk 21: tensor([[ 4.7868, -5.2969]])\n",827 "Chunk shape: (160000,)\n",828 "Logits for chunk 22: tensor([[ 4.7600, -5.2690]])\n",829 "Chunk shape: (160000,)\n",830 "Logits for chunk 23: tensor([[ 4.7337, -5.2411]])\n",831 "Chunk shape: (160000,)\n",832 "Logits for chunk 24: tensor([[ 4.7835, -5.2943]])\n",833 "Chunk shape: (160000,)\n",834 "Logits for chunk 25: tensor([[ 4.7572, -5.2647]])\n",835 "Chunk shape: (160000,)\n",836 "Logits for chunk 26: tensor([[ 4.7485, -5.2581]])\n",837 "Chunk shape: (160000,)\n",838 "Logits for chunk 27: tensor([[ 4.6874, -5.2023]])\n",839 "Chunk shape: (160000,)\n",840 "Logits for chunk 28: tensor([[ 4.6877, -5.1922]])\n",841 "Chunk shape: (160000,)\n",842 "Logits for chunk 29: tensor([[ 4.7474, -5.2561]])\n",843 "Chunk shape: (160000,)\n",844 "Logits for chunk 30: tensor([[-4.3064, 4.7629]])\n",845 "Chunk shape: (160000,)\n",846 "Logits for chunk 31: tensor([[-3.8067, 4.2312]])\n",847 "Chunk shape: (160000,)\n",848 "Logits for chunk 32: tensor([[ 4.7217, -5.2325]])\n",849 "Chunk shape: (160000,)\n",850 "Logits for chunk 33: tensor([[ 4.7798, -5.2913]])\n",851 "Chunk shape: (160000,)\n",852 "Logits for chunk 34: tensor([[ 4.7214, -5.2355]])\n",853 "Chunk shape: (160000,)\n",854 "Logits for chunk 35: tensor([[ 4.7116, -5.2192]])\n",855 "Chunk shape: (160000,)\n",856 "Logits for chunk 36: tensor([[ 4.6687, -5.1812]])\n",857 "Chunk shape: (160000,)\n",858 "Logits for chunk 37: tensor([[-0.8128, 0.9402]])\n",859 "Chunk shape: (160000,)\n",860 "Logits for chunk 38: tensor([[ 4.7259, -5.2333]])\n",861 "Chunk shape: (160000,)\n",862 "Logits for chunk 39: tensor([[ 4.5698, -5.0731]])\n",863 "Chunk shape: (160000,)\n",864 "Logits for chunk 40: tensor([[ 4.7467, -5.2544]])\n",865 "Chunk shape: (160000,)\n",866 "Logits for chunk 41: tensor([[ 4.7781, -5.2884]])\n",867 "Chunk shape: (160000,)\n",868 "Logits for chunk 42: tensor([[ 4.7243, -5.2365]])\n",869 "Chunk shape: (160000,)\n",870 "Logits for chunk 43: tensor([[ 3.9325, -4.3570]])\n",871 "Chunk shape: (160000,)\n",872 "Logits for chunk 44: tensor([[-3.8786, 4.3105]])\n",873 "Chunk shape: (160000,)\n",874 "Logits for chunk 45: tensor([[ 3.3633, -3.6958]])\n",875 "Chunk shape: (160000,)\n",876 "Logits for chunk 46: tensor([[ 4.7127, -5.2213]])\n",877 "Chunk shape: (160000,)\n",878 "Logits for chunk 47: tensor([[ 0.0519, -0.0359]])\n",879 "Chunk shape: (160000,)\n",880 "Logits for chunk 48: tensor([[ 4.7457, -5.2535]])\n",881 "Chunk shape: (160000,)\n",882 "Logits for chunk 49: tensor([[ 3.4856, -3.8528]])\n",883 "Chunk shape: (160000,)\n",884 "Logits for chunk 50: tensor([[ 4.6485, -5.1538]])\n",885 "Chunk shape: (160000,)\n",886 "Logits for chunk 51: tensor([[ 4.6274, -5.1355]])\n",887 "Chunk shape: (160000,)\n",888 "Logits for chunk 52: tensor([[ 4.6852, -5.1872]])\n",889 "Chunk shape: (160000,)\n",890 "Logits for chunk 53: tensor([[ 4.7341, -5.2452]])\n",891 "Chunk shape: (160000,)\n",892 "Logits for chunk 54: tensor([[-4.5378, 5.0152]])\n",893 "Chunk shape: (160000,)\n",894 "Logits for chunk 55: tensor([[ 4.6822, -5.1887]])\n",895 "Chunk shape: (160000,)\n",896 "Logits for chunk 56: tensor([[ 4.7186, -5.2252]])\n",897 "Chunk shape: (160000,)\n",898 "Logits for chunk 57: tensor([[ 4.7688, -5.2787]])\n",899 "Chunk shape: (160000,)\n",900 "Logits for chunk 58: tensor([[ 4.7285, -5.2342]])\n",901 "Chunk shape: (160000,)\n",902 "Logits for chunk 59: tensor([[ 4.7447, -5.2550]])\n",903 "Chunk shape: (160000,)\n",904 "Logits for chunk 60: tensor([[ 4.5292, -5.0253]])\n",905 "Predicted Class: Fake\n"906 ]907 }908 ],909 "source": [910 "def predict_audio(file_path):\n",911 " \"\"\"\n",912 " Predicts the class of an audio file by aggregating predictions from chunks.\n",913 " \n",914 " Args:\n",915 " file_path (str): Path to the audio file.\n",916 "\n",917 " Returns:\n",918 " str: Predicted class label.\n",919 " \"\"\"\n",920 " # Prepare audio chunks\n",921 " audio_chunks = prepare_audio(file_path)\n",922 " predictions = []\n",923 "\n",924 " for i, chunk in enumerate(audio_chunks):\n",925 " # Prepare input for the model\n",926 " print(f\"Chunk shape: {chunk.shape}\")\n",927 " inputs = processor(\n",928 " chunk, sampling_rate=16000, return_tensors=\"pt\", padding=True\n",929 " )\n",930 " \n",931 " # Perform inference\n",932 " with torch.no_grad():\n",933 " outputs = model(**inputs)\n",934 " logits = outputs.logits\n",935 " print(f\"Logits for chunk {i + 1}: {logits}\") # Print the logits\n",936 " predicted_class = torch.argmax(logits, dim=1).item()\n",937 " predictions.append(predicted_class)\n",938 " \n",939 " # Aggregate predictions (e.g., majority voting)\n",940 " aggregated_prediction = max(set(predictions), key=predictions.count)\n",941 " \n",942 " # Convert class ID to label\n",943 " return model.config.id2label[aggregated_prediction]\n",944 "\n",945 "# Example: Test a single audio file\n",946 "file_path = r\"D:\\Year 3 Sem 2\\Godamlah\\Deepfake\\deepfake model ver3\\data\\KAGGLE\\AUDIO\\FAKE\\biden-to-linus.wav\" # Replace with your audio file path\n",947 "predicted_class = predict_audio(file_path)\n",948 "print(f\"Predicted Class: {predicted_class}\")\n"949 ]950 },951 {952 "cell_type": "markdown",953 "metadata": {},954 "source": [955 "## Batch Testing"956 ]957 },958 {959 "cell_type": "code",960 "execution_count": 36,961 "metadata": {},962 "outputs": [963 {964 "name": "stdout",965 "output_type": "stream",966 "text": [967 "Chunk shape: (160000,)\n",968 "Logits for chunk 1: tensor([[-3.3933, 3.7590]])\n",969 "Chunk shape: (160000,)\n",970 "Logits for chunk 1: tensor([[-3.3933, 3.7590]])\n",971 "Chunk shape: (160000,)\n",972 "Logits for chunk 1: tensor([[-1.5531, 1.7190]])\n",973 "Chunk shape: (160000,)\n",974 "Logits for chunk 1: tensor([[-1.5917, 1.7620]])\n",975 "Chunk shape: (160000,)\n",976 "Logits for chunk 1: tensor([[ 4.7569, -5.2631]])\n",977 "Chunk shape: (160000,)\n",978 "Logits for chunk 1: tensor([[ 4.7569, -5.2630]])\n",979 "Chunk shape: (160000,)\n",980 "Logits for chunk 1: tensor([[-4.5033, 4.9768]])\n",981 "Chunk shape: (160000,)\n",982 "Logits for chunk 1: tensor([[-4.5029, 4.9765]])\n",983 "Chunk shape: (160000,)\n",984 "Logits for chunk 1: tensor([[ 4.7639, -5.2653]])\n",985 "Chunk shape: (160000,)\n",986 "Logits for chunk 1: tensor([[ 4.7639, -5.2653]])\n",987 "{'file': 'human voice 1 to mr beast.mp3', 'predicted_class': 'Real'}\n",988 "{'file': 'human voice 1 to mr beast.wav', 'predicted_class': 'Real'}\n",989 "{'file': 'human voice 1.mp3', 'predicted_class': 'Real'}\n",990 "{'file': 'human voice 1.wav', 'predicted_class': 'Real'}\n",991 "{'file': 'human voice 2 to Jett.mp3', 'predicted_class': 'Fake'}\n",992 "{'file': 'human voice 2 to Jett.wav', 'predicted_class': 'Fake'}\n",993 "{'file': 'human voice 2.mp3', 'predicted_class': 'Real'}\n",994 "{'file': 'human voice 2.wav', 'predicted_class': 'Real'}\n",995 "{'file': 'text to audio jett.mp3', 'predicted_class': 'Fake'}\n",996 "{'file': 'text to audio jett.wav', 'predicted_class': 'Fake'}\n"997 ]998 }999 ],1000 "source": [1001 "import os\n",1002 "\n",1003 "def batch_predict(test_folder, limit=10):\n",1004 " \"\"\"\n",1005 " Batch processes audio files for predictions.\n",1006 "\n",1007 " Args:\n",1008 " test_folder (str): Path to the folder containing audio files.\n",1009 " limit (int): Maximum number of files to process. Set to None for all files.\n",1010 "\n",1011 " Returns:\n",1012 " list: A list of dictionaries containing file names and predicted classes.\n",1013 " \"\"\"\n",1014 " results = []\n",1015 " files = os.listdir(test_folder)\n",1016 "\n",1017 " # Limit the number of files processed if a limit is provided\n",1018 " if limit is not None:\n",1019 " files = files[:limit]\n",1020 "\n",1021 " # Process each file in the folder\n",1022 " for file_name in files:\n",1023 " file_path = os.path.join(test_folder, file_name)\n",1024 " try:\n",1025 " predicted_class = predict_audio(file_path) # Use the predict_audio function\n",1026 " results.append({\"file\": file_name, \"predicted_class\": predicted_class})\n",1027 " except Exception as e:\n",1028 " print(f\"Error processing {file_name}: {e}\")\n",1029 " \n",1030 " return results\n",1031 "\n",1032 "# Specify the folder path and limit\n",1033 "test_folder = r\"D:\\Year 3 Sem 2\\Godamlah\\Deepfake\\deepfake model ver3\\data\\real life test audio\" # Replace with your test folder path\n",1034 "results = batch_predict(test_folder, limit=10)\n",1035 "\n",1036 "# Print results\n",1037 "for result in results:\n",1038 " print(result)\n"1039 ]1040 },1041 {1042 "cell_type": "code",1043 "execution_count": 14,1044 "metadata": {},1045 "outputs": [1046 {1047 "data": {1048 "application/vnd.jupyter.widget-view+json": {1049 "model_id": "479899631f95453e9f82355f7511cff3",1050 "version_major": 2,1051 "version_minor": 01052 },1053 "text/plain": [1054 " 0%| | 0/580 [00:00<?, ?it/s]"1055 ]1056 },1057 "metadata": {},1058 "output_type": "display_data"1059 },1060 {1061 "name": "stdout",1062 "output_type": "stream",1063 "text": [1064 "{'eval_loss': 4.472154250834137e-05, 'eval_model_preparation_time': 0.0032, 'eval_accuracy': 1.0, 'eval_runtime': 1168.3696, 'eval_samples_per_second': 3.966, 'eval_steps_per_second': 0.496}\n"1065 ]1066 }1067 ],1068 "source": [1069 "import evaluate\n",1070 "\n",1071 "# Load the accuracy metric\n",1072 "accuracy_metric = evaluate.load(\"accuracy\")\n",1073 "\n",1074 "def compute_metrics(eval_pred):\n",1075 " logits, labels = eval_pred\n",1076 " predictions = logits.argmax(axis=-1) # Get the predicted class\n",1077 " accuracy = accuracy_metric.compute(predictions=predictions, references=labels)\n",1078 " return {\"accuracy\": accuracy[\"accuracy\"]}\n",1079 "\n",1080 "from transformers import Trainer\n",1081 "\n",1082 "trainer = Trainer(\n",1083 " model=model,\n",1084 " args=training_args,\n",1085 " train_dataset=processed_dataset[\"train\"],\n",1086 " eval_dataset=processed_dataset[\"validation\"],\n",1087 " tokenizer=processor, # Required for padding\n",1088 " data_collator=data_collator,\n",1089 " compute_metrics=compute_metrics, # Add this line\n",1090 ")\n",1091 "\n",1092 "# Evaluate the model\n",1093 "metrics = trainer.evaluate()\n",1094 "print(metrics)\n"1095 ]1096 }1097 ],1098 "metadata": {1099 "kernelspec": {1100 "display_name": "base",1101 "language": "python",1102 "name": "python3"1103 },1104 "language_info": {1105 "codemirror_mode": {1106 "name": "ipython",1107 "version": 31108 },1109 "file_extension": ".py",1110 "mimetype": "text/x-python",1111 "name": "python",1112 "nbconvert_exporter": "python",1113 "pygments_lexer": "ipython3",1114 "version": "3.12.7"1115 }1116 },1117 "nbformat": 4,1118 "nbformat_minor": 21119}1120