Training_model_CNN_optimized.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# import librosa\n",
    "import torchaudio as ta\n",
    "ta.set_audio_backend(\"sox_io\")\n",
    "import torch\n",
    "from torch.utils.data import DataLoader\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.autograd.profiler as profiler\n",
    "import pytorch_lightning as pl\n",
    "\n",
    "import numpy as np\n",
    "import os\n",
    "\n",
    "import IPython.display as ipd\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "import math\n",
    "\n",
    "import glob\n",
    "\n",
    "from tqdm.auto import tqdm\n",
    "\n",
    "from python_files.Noise_Reduction_Datagen_paths_optimized import Signal_Synthesis_DataGen\n",
    "# from python_files.unet_basic import Model\n",
    "\n",
    "import warnings\n",
    "\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "import gc\n",
    "\n",
    "from random import shuffle\n",
    "\n",
    "# from numba import jit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from tqdm.auto import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# from fastai.text.all import *\n",
    "# from fastai.data.core import DataLoaders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import fastai\n",
    "\n",
    "# fastai.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8700"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.backends.cudnn.version()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'2.0.0+cu118'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.backends.cudnn.benchmark = True\n",
    "\n",
    "torch.__version__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.cuda.is_available()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'NVIDIA GeForce RTX 3090'"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.cuda.get_device_name()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def get_paths(dir_path):\n",
    "    paths = []\n",
    "    for root, dirs, files in os.walk(dir_path):\n",
    "        for name in files:\n",
    "            if name.endswith(\".wav\") or name.endswith(\".mp3\"):\n",
    "                paths.append(os.path.join(root, name))\n",
    "                \n",
    "    paths = np.asarray(paths)\n",
    "    return paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1427394"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(signal_paths_sh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# noise_paths_sh = get_paths(\"./dataset/Reduced_noise/\")\n",
    "noise_paths_sh = get_paths(\"/home/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/dataset/dataset/UrbanSound8K-Resampled/all_files\")\n",
    "shuffle(noise_paths_sh)\n",
    "\n",
    "# signal_paths_sh = get_paths(\"./dataset/Reduced_clean_signals/\")\n",
    "signal_paths_sh = get_paths(\"/home/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/dataset/dataset/cv-corpus-5.1-2020-06-22-Resampled/en\")\n",
    "shuffle(signal_paths_sh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1426394 /home/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/dataset/dataset/cv-corpus-5.1-2020-06-22-Resampled/en/clips/common_voice_en_18694067.mp3\n"
     ]
    }
   ],
   "source": [
    "noise_paths = noise_paths_sh[:-1000]\n",
    "signal_paths = signal_paths_sh[:-1000]\n",
    "signal_dir = \"\"#\"./dataset/cv-corpus-5.1-2020-06-22-Resampled/en/clips\"\n",
    "noise_save_path = \"\"#\"./dataset_loader_files/noise_paths_resampled_save.npy\"\n",
    "train = True\n",
    "n_fft=1024\n",
    "win_length=n_fft\n",
    "hop_len=n_fft//4\n",
    "create_specgram = False\n",
    "perform_stft = False\n",
    "default_sr = 16000\n",
    "sec = (16384/default_sr)*2\n",
    "augment=True\n",
    "device_datagen = \"cpu\"\n",
    "\n",
    "train_ds = Signal_Synthesis_DataGen(noise_paths, signal_paths, signal_dir, \\\n",
    "                 n_fft=n_fft, win_length=win_length, hop_len=hop_len, create_specgram=create_specgram, \\\n",
    "                 perform_stft=perform_stft, normalize=True, default_sr=default_sr, sec=sec, epsilon=1e-5, augment=False, device=device_datagen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000 /home/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/dataset/dataset/cv-corpus-5.1-2020-06-22-Resampled/en/clips/common_voice_en_595946.mp3\n"
     ]
    }
   ],
   "source": [
    "noise_paths = noise_paths_sh[-1000:]\n",
    "signal_paths = signal_paths_sh[-1000:]\n",
    "signal_dir = \"\"#\"./dataset/cv-corpus-5.1-2020-06-22-Resampled/en/clips\"\n",
    "noise_save_path = \"\"#\"./dataset_loader_files/noise_paths_resampled_save.npy\"\n",
    "train = False\n",
    "\n",
    "val_ds = Signal_Synthesis_DataGen(noise_paths, signal_paths, signal_dir,\\\n",
    "                 n_fft=n_fft, win_length=win_length, hop_len=hop_len, create_specgram=create_specgram, \\\n",
    "                 perform_stft=perform_stft, normalize=True, default_sr=default_sr, sec=sec, epsilon=1e-5, augment=False, device=device_datagen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([3])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "torch.randint(0, 4, size=[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "ename": "RuntimeError",
     "evalue": "Failed to load audio from /home/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/dataset/dataset/cv-corpus-5.1-2020-06-22-Resampled/en/clips/common_voice_en_18664637.mp3",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[17], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m sig \u001b[38;5;241m=\u001b[39m \u001b[43mval_ds\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[38;5;21;43m__getitem__\u001b[39;49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m)\u001b[49m[\u001b[38;5;241m0\u001b[39m]\n",
      "File \u001b[0;32m/mnt/nvme1n1p2/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/python_files/Noise_Reduction_Datagen_paths_optimized.py:183\u001b[0m, in \u001b[0;36mSignal_Synthesis_DataGen.__getitem__\u001b[0;34m(self, idx)\u001b[0m\n\u001b[1;32m    180\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m torch\u001b[38;5;241m.\u001b[39mis_tensor(idx):\n\u001b[1;32m    181\u001b[0m     idx \u001b[38;5;241m=\u001b[39m idx\u001b[38;5;241m.\u001b[39mtolist()\n\u001b[0;32m--> 183\u001b[0m signal_noise_add, signal \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mget_ids\u001b[49m\u001b[43m(\u001b[49m\u001b[43midx\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    185\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m signal_noise_add, signal\n",
      "File \u001b[0;32m/mnt/nvme1n1p2/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/python_files/Noise_Reduction_Datagen_paths_optimized.py:132\u001b[0m, in \u001b[0;36mSignal_Synthesis_DataGen.get_ids\u001b[0;34m(self, idx)\u001b[0m\n\u001b[1;32m    128\u001b[0m noise_id \u001b[38;5;241m=\u001b[39m idx \u001b[38;5;241m-\u001b[39m signal_id\u001b[38;5;241m*\u001b[39m\u001b[38;5;28mlen\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnoise_paths)\n\u001b[1;32m    130\u001b[0m signal_path, noise_path \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mconstruct_signal_path(signal_id), \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnoise_paths[noise_id]\n\u001b[0;32m--> 132\u001b[0m signal_noise_add, signal \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mdevelop_data\u001b[49m\u001b[43m(\u001b[49m\u001b[43msignal_path\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnoise_path\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    134\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnormalize:\n\u001b[1;32m    135\u001b[0m     signal_noise_add \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mnormalize_signal(signal_noise_add)\n",
      "File \u001b[0;32m/mnt/nvme1n1p2/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/python_files/Noise_Reduction_Datagen_paths_optimized.py:153\u001b[0m, in \u001b[0;36mSignal_Synthesis_DataGen.develop_data\u001b[0;34m(self, signal_path, noise_path)\u001b[0m\n\u001b[1;32m    151\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m nsr \u001b[38;5;241m!=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdefault_sr:\n\u001b[1;32m    152\u001b[0m     noise \u001b[38;5;241m=\u001b[39m ta\u001b[38;5;241m.\u001b[39mtransforms\u001b[38;5;241m.\u001b[39mResample(orig_freq\u001b[38;5;241m=\u001b[39mnsr, new_freq\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mdefault_sr)(noise)\n\u001b[0;32m--> 153\u001b[0m signal, ssr \u001b[38;5;241m=\u001b[39m \u001b[43mta\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mload\u001b[49m\u001b[43m(\u001b[49m\u001b[43msignal_path\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    154\u001b[0m signal \u001b[38;5;241m=\u001b[39m signal[\u001b[38;5;241m0\u001b[39m]\u001b[38;5;66;03m#.to(self.device)\u001b[39;00m\n\u001b[1;32m    155\u001b[0m signal \u001b[38;5;241m=\u001b[39m signal\u001b[38;5;66;03m#.type(torch.float32)\u001b[39;00m\n",
      "File \u001b[0;32m~/Data/miniconda3/envs/kaggle-env/lib/python3.8/site-packages/torchaudio/backend/sox_io_backend.py:256\u001b[0m, in \u001b[0;36mload\u001b[0;34m(filepath, frame_offset, num_frames, normalize, channels_first, format)\u001b[0m\n\u001b[1;32m    254\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m ret \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m    255\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m ret\n\u001b[0;32m--> 256\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43m_fallback_load\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfilepath\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mframe_offset\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnum_frames\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnormalize\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mchannels_first\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43mformat\u001b[39;49m\u001b[43m)\u001b[49m\n",
      "File \u001b[0;32m~/Data/miniconda3/envs/kaggle-env/lib/python3.8/site-packages/torchaudio/backend/sox_io_backend.py:30\u001b[0m, in \u001b[0;36m_fail_load\u001b[0;34m(filepath, frame_offset, num_frames, normalize, channels_first, format)\u001b[0m\n\u001b[1;32m     22\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_fail_load\u001b[39m(\n\u001b[1;32m     23\u001b[0m     filepath: \u001b[38;5;28mstr\u001b[39m,\n\u001b[1;32m     24\u001b[0m     frame_offset: \u001b[38;5;28mint\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;241m0\u001b[39m,\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m     28\u001b[0m     \u001b[38;5;28mformat\u001b[39m: Optional[\u001b[38;5;28mstr\u001b[39m] \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m     29\u001b[0m ) \u001b[38;5;241m-\u001b[39m\u001b[38;5;241m>\u001b[39m Tuple[torch\u001b[38;5;241m.\u001b[39mTensor, \u001b[38;5;28mint\u001b[39m]:\n\u001b[0;32m---> 30\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mRuntimeError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mFailed to load audio from \u001b[39m\u001b[38;5;132;01m{}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;241m.\u001b[39mformat(filepath))\n",
      "\u001b[0;31mRuntimeError\u001b[0m: Failed to load audio from /home/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/dataset/dataset/cv-corpus-5.1-2020-06-22-Resampled/en/clips/common_voice_en_18664637.mp3"
     ]
    }
   ],
   "source": [
    "sig = val_ds.__getitem__(1)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "                <audio  controls=\"controls\" >\n",
       "                    <source src=\"data:audio/wav;base64,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\" type=\"audio/wav\" />\n",
       "                    Your browser does not support the audio element.\n",
       "                </audio>\n",
       "              "
      ],
      "text/plain": [
       "<IPython.lib.display.Audio object>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = sig.squeeze().t().to(\"cpu\").numpy()\n",
    "ipd.Audio(x, rate=default_sr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "BATCH_SIZE = 104\n",
    "shuffle = True\n",
    "num_workers = 0\n",
    "pin_memory = False\n",
    "\n",
    "# data_loader = DataLoader(signal_synthesis_dataset, batch_size=BATCH_SIZE, shuffle=shuffle, num_workers=num_workers)\n",
    "# data_loader = DataLoader(train_ds, batch_size=BATCH_SIZE, shuffle=shuffle, num_workers=num_workers, pin_memory=pin_memory)\n",
    "\n",
    "dls = DataLoaders.from_dsets(train_ds, val_ds, bs=BATCH_SIZE, num_workers=num_workers, pin_memory=pin_memory).to(\"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n",
      "tensor(1.0000) tensor(1.0000e-05)\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<timed exec>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/fastai/data/load.py\u001b[0m in \u001b[0;36m__iter__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     99\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbefore_iter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    100\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__idxs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_idxs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;31m# called in context of main process (not workers/subprocesses)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 101\u001b[0;31m         \u001b[0;32mfor\u001b[0m \u001b[0mb\u001b[0m \u001b[0;32min\u001b[0m \u001b[0m_loaders\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfake_l\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnum_workers\u001b[0m\u001b[0;34m==\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfake_l\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    102\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdevice\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mto_device\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdevice\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    103\u001b[0m             \u001b[0;32myield\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mafter_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/torch/utils/data/dataloader.py\u001b[0m in \u001b[0;36m__next__\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    519\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sampler_iter\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    520\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_reset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 521\u001b[0;31m             \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_next_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    522\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_num_yielded\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    523\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_dataset_kind\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0m_DatasetKind\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mIterable\u001b[0m \u001b[0;32mand\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/torch/utils/data/dataloader.py\u001b[0m in \u001b[0;36m_next_data\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    559\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0m_next_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    560\u001b[0m         \u001b[0mindex\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_next_index\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# may raise StopIteration\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 561\u001b[0;31m         \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_dataset_fetcher\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfetch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;31m# may raise StopIteration\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    562\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_pin_memory\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    563\u001b[0m             \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_utils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpin_memory\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpin_memory\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py\u001b[0m in \u001b[0;36mfetch\u001b[0;34m(self, possibly_batched_index)\u001b[0m\n\u001b[1;32m     32\u001b[0m                 \u001b[0;32mraise\u001b[0m \u001b[0mStopIteration\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     33\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 34\u001b[0;31m             \u001b[0mdata\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnext\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset_iter\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     35\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcollate_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     36\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/fastai/data/load.py\u001b[0m in \u001b[0;36mcreate_batches\u001b[0;34m(self, samps)\u001b[0m\n\u001b[1;32m    108\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mit\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0miter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    109\u001b[0m         \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfilter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;32mlambda\u001b[0m \u001b[0mo\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0mo\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdo_item\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msamps\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 110\u001b[0;31m         \u001b[0;32myield\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mmap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdo_batch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mchunkify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    111\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    112\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mnew\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdataset\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcls\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/fastcore/basics.py\u001b[0m in \u001b[0;36mchunked\u001b[0;34m(it, chunk_sz, drop_last, n_chunks)\u001b[0m\n\u001b[1;32m    211\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mit\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mIterator\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mit\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0miter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mit\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    212\u001b[0m     \u001b[0;32mwhile\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 213\u001b[0;31m         \u001b[0mres\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mitertools\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mislice\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mit\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mchunk_sz\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    214\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mres\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m==\u001b[0m\u001b[0mchunk_sz\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mdrop_last\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32myield\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    215\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m<\u001b[0m\u001b[0mchunk_sz\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/fastai/data/load.py\u001b[0m in \u001b[0;36mdo_item\u001b[0;34m(self, s)\u001b[0m\n\u001b[1;32m    123\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mprebatched\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbs\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    124\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mdo_item\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 125\u001b[0;31m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mafter_item\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate_item\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    126\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mSkipItemException\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    127\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mchunkify\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mb\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprebatched\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0mchunked\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdrop_last\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/fastai/data/load.py\u001b[0m in \u001b[0;36mcreate_item\u001b[0;34m(self, s)\u001b[0m\n\u001b[1;32m    129\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mrandomize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrng\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrandom\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mRandom\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrng\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrandint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m**\u001b[0m\u001b[0;36m32\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    130\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mretain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;32mreturn\u001b[0m \u001b[0mretain_types\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mis_listy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 131\u001b[0;31m     \u001b[0;32mdef\u001b[0m \u001b[0mcreate_item\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;32mreturn\u001b[0m \u001b[0mnext\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mit\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    132\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mcreate_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mfa_collate\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mfa_convert\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprebatched\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    133\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mdo_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mretain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbefore_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/python_files/Noise_Reduction_Datagen_paths_optimized.py\u001b[0m in \u001b[0;36m__getitem__\u001b[0;34m(self, idx)\u001b[0m\n\u001b[1;32m    181\u001b[0m             \u001b[0midx\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0midx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtolist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    182\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 183\u001b[0;31m         \u001b[0msignal_noise_add\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msignal\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_ids\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0midx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    184\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    185\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0msignal_noise_add\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msignal\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/python_files/Noise_Reduction_Datagen_paths_optimized.py\u001b[0m in \u001b[0;36mget_ids\u001b[0;34m(self, idx)\u001b[0m\n\u001b[1;32m    130\u001b[0m         \u001b[0msignal_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnoise_path\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconstruct_signal_path\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msignal_id\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnoise_paths\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mnoise_id\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    131\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 132\u001b[0;31m         \u001b[0msignal_noise_add\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msignal\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdevelop_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msignal_path\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnoise_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    133\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    134\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnormalize\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/git-repos/Noise-Reduction-Deep-Learning/python_files/Noise_Reduction_Datagen_paths_optimized.py\u001b[0m in \u001b[0;36mdevelop_data\u001b[0;34m(self, signal_path, noise_path)\u001b[0m\n\u001b[1;32m    151\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mnsr\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdefault_sr\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    152\u001b[0m             \u001b[0mnoise\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mta\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransforms\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mResample\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0morig_freq\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnsr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnew_freq\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdefault_sr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnoise\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 153\u001b[0;31m         \u001b[0msignal\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mssr\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mta\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msignal_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    154\u001b[0m         \u001b[0msignal\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msignal\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;31m#.to(self.device)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    155\u001b[0m         \u001b[0msignal\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msignal\u001b[0m\u001b[0;31m#.type(torch.float32)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/mnt/nvme0n1p2/vamsik1211/Data/pytorch-local/lib/python3.7/site-packages/torchaudio/backend/sox_backend.py\u001b[0m in \u001b[0;36mload\u001b[0;34m(filepath, out, normalization, channels_first, num_frames, offset, signalinfo, encodinginfo, filetype)\u001b[0m\n\u001b[1;32m     54\u001b[0m         \u001b[0msignalinfo\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     55\u001b[0m         \u001b[0mencodinginfo\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 56\u001b[0;31m         \u001b[0mfiletype\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     57\u001b[0m     )\n\u001b[1;32m     58\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "%%time\n",
    "for data in dls.train:\n",
    "    print(data[0].max(), data[0].min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Unsqueezer(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Unsqueezer, self).__init__()\n",
    "        \n",
    "    def forward(self, x):\n",
    "        return x.usqueeze(dim=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "dls.after_batch = Unsqueezer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %%time\n",
    "# for data in dls.train:\n",
    "#     print(data[0].max(), data[0].min())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "80142855"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_ds.__len__()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Negative_SNR_Loss(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Negative_SNR_Loss, self).__init__()\n",
    "    \n",
    "    def forward(self, sig_pred, sig_true):\n",
    "        \n",
    "        sig_true_sq = torch.square(sig_true)\n",
    "        sig_pred_sq = torch.square(sig_true - sig_pred)\n",
    "\n",
    "        sig_true_mean = torch.mean(sig_true_sq)\n",
    "        sig_pred_mean = torch.mean(sig_pred_sq)\n",
    "\n",
    "        snr = sig_true_mean / sig_pred_mean + 1e-7\n",
    "        loss = -1*torch.log10(snr)\n",
    "\n",
    "        return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Mod_MSELoss(nn.Module):\n",
    "    def __init__(self, mul_factor):\n",
    "        super(Mod_MSELoss, self).__init__()\n",
    "        self.loss_fn = nn.MSELoss(reduction=\"mean\")\n",
    "        self.mul_factor = mul_factor\n",
    "        \n",
    "        \n",
    "    def forward(self, sig_pred, sig_true):\n",
    "        loss = self.loss_fn(sig_pred, sig_true)\n",
    "        loss = self.mul_factor*loss\n",
    "        return loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# class InstantLayerNormalization(nn.Module):\n",
    "#     def __init__(self, in_shape, out_shape):\n",
    "#         self.in_shape = in_shape\n",
    "#         self.out_shape = out_shape\n",
    "        \n",
    "#         self.epsilon = 1e-7\n",
    "#         self.gamma = None\n",
    "#         self.beta = None\n",
    "        \n",
    "#         super(InstantLayerNormalization, self).__init__()\n",
    "        \n",
    "#         self.gamma = torch.ones(out_shape)\n",
    "#         self.gamma = nn.Parameter(self.gamma)\n",
    "        \n",
    "#         self.beta = torch.zeros(out_shape)\n",
    "#         self.beta = nn.Parameter(self.beta)\n",
    "        \n",
    "#     def forward(self, inps):\n",
    "#         mean = inps.mean(-1, keepdim=True)\n",
    "#         variance = torch.mean(torch.square(inps - mean), dim=-1, keepdim=True)\n",
    "#         std = torch.sqrt(variance + self.epsilon)\n",
    "        \n",
    "#         outs = (inps - mean) / std\n",
    "#         print(outs.shape, self.gamma.shape)\n",
    "#         outs = outs * self.gamma\n",
    "#         outs = outs + self.beta\n",
    "        \n",
    "#         return outs\n",
    "    \n",
    "# class Multiply():\n",
    "#     def __init__(self):\n",
    "#         super(Multiply, self).__init__()\n",
    "    \n",
    "#     def forward(self, ten1, ten2):\n",
    "#         mul_out = torch.mul(ten1, ten2)\n",
    "#         return mul_out\n",
    "\n",
    "# class NoiseReducer(pl.LightningModule):\n",
    "#     def __init__(self, default_sr, n_fft, win_length, hop_len, sec, dropout=0.5, batch_first=True, stride=2, normalized=False, bidir=False):\n",
    "        \n",
    "#         self.default_sr = default_sr\n",
    "#         self.n_fft = n_fft\n",
    "#         self.win_length = win_length\n",
    "#         self.hop_len = hop_len\n",
    "#         self.sec = sec\n",
    "#         self.normalized = normalized\n",
    "        \n",
    "#         self.conv_filters = 512\n",
    "        \n",
    "#         # Universal LSTM Units\n",
    "#         self.batch_first = True\n",
    "#         self.dropout = 0.25\n",
    "#         self.bidir = bidir\n",
    "#         self.lstm_prelu_ins = int(np.floor(((default_sr*sec) - (win_length - 1) - 1)/ hop_len + 5))\n",
    "        \n",
    "#         # LSTM 1 UNITS\n",
    "#         self.rnn1_dims = n_fft // 2 + 1\n",
    "#         self.hidden_size_1 = 256\n",
    "#         self.num_layers = 2\n",
    "       \n",
    "        \n",
    "#         # LSTM 2 UNITS\n",
    "#         self.rnn2_dims = self.conv_filters\n",
    "#         self.hidden_size_2 = self.hidden_size_1\n",
    "        \n",
    "#         # Conv1d Layer Units\n",
    "#         self.conv1_in = 1\n",
    "#         self.conv1_out = self.conv_filters\n",
    "        \n",
    "        \n",
    "        \n",
    "#         # InstanceNorm Layer Units\n",
    "#         self.instance1_in = self.rnn1_dims\n",
    "#         self.instance2_in = self.conv1_out\n",
    "        \n",
    "#         # Dense1 Layer Units\n",
    "#         self.dense1_in = self.hidden_size_1\n",
    "#         self.dense1_out = self.rnn1_dims #int(np.floor(((default_sr*sec) - (win_length - 1) - 1)/ hop_len + 5))#3))\n",
    "        \n",
    "#         # Dense2 Layer Units\n",
    "#         self.dense2_in = self.hidden_size_2\n",
    "#         self.dense2_out = self.conv1_out\n",
    "        \n",
    "#         # Dense3 Layer Units\n",
    "#         self.dense3_in = self.hidden_size_1\n",
    "#         self.dense3_out = self.rnn1_dims\n",
    "        \n",
    "#         # Conv2d Layer Units\n",
    "#         self.conv2_in = self.dense2_out\n",
    "#         self.conv2_out = self.conv_filters\n",
    "        \n",
    "\n",
    "#         super(NoiseReducer, self).__init__()\n",
    "        \n",
    "#         self.loss_fn = Mod_MSELoss(mul_factor=1000)\n",
    "        \n",
    "#         self.lstm1 = nn.LSTM(input_size=self.rnn1_dims, hidden_size=self.hidden_size_1, num_layers=self.num_layers, batch_first=self.batch_first, dropout=self.dropout, bidirectional=self.bidir)\n",
    "#         self.lstm3 = nn.LSTM(input_size=self.rnn1_dims, hidden_size=self.hidden_size_1, num_layers=self.num_layers, batch_first=self.batch_first, dropout=self.dropout, bidirectional=self.bidir)\n",
    "# #         self.lstm1_1 = nn.LSTMCell(input_size=self.rnn1_dims, hidden_size==self.hidden_size_1, )\n",
    "        \n",
    "#         print(self.rnn2_dims)\n",
    "#         self.lstm2 = nn.LSTM(input_size=self.rnn2_dims, hidden_size=self.hidden_size_2, num_layers=self.num_layers*2, batch_first=self.batch_first, dropout=self.dropout, bidirectional=self.bidir)\n",
    "        \n",
    "        \n",
    "#         self.instancenorm1 = nn.InstanceNorm1d(self.rnn1_dims)\n",
    "#         self.instancenorm2 = nn.InstanceNorm1d(self.rnn2_dims)\n",
    "#         self.instancenorm3 = nn.InstanceNorm1d(self.rnn1_dims)\n",
    "        \n",
    "#         self.dense1 = nn.Linear(self.dense1_in, self.dense1_out)\n",
    "#         self.dense2 = nn.Linear(self.dense2_in, self.dense2_out)\n",
    "#         self.dense3 = nn.Linear(self.dense3_in, self.dense3_out)\n",
    "        \n",
    "#         self.conv1 = nn.Conv1d(self.conv1_in, self.conv1_out, kernel_size=3, stride=1, padding=1)\n",
    "#         self.conv2 = nn.Conv1d(self.conv2_in, self.conv2_out, kernel_size=3, stride=1, padding=1)\n",
    "        \n",
    "#         self.prelu_conv1 = nn.PReLU(self.conv1_out)\n",
    "#         self.prelu_conv2 = nn.PReLU(self.conv2_out)\n",
    "        \n",
    "#         self.prelu_lstm1 = nn.PReLU(self.lstm_prelu_ins)\n",
    "#         self.prelu_lstm3 = nn.PReLU(self.lstm_prelu_ins)\n",
    "#         self.prelu_lstm2 = nn.PReLU(self.hidden_size_2)\n",
    "        \n",
    "#     @torch.jit.export\n",
    "#     def stft_layer(self, sig):\n",
    "        \n",
    "#         sig_stft = torch.stft(sig, n_fft=self.n_fft, hop_length=self.hop_len, win_length=self.win_length)\n",
    "        \n",
    "#         sig_cplx = torch.view_as_complex(sig_stft)\n",
    "#         mag = sig_cplx.abs().permute(0, 2, 1)\n",
    "#         angle = sig_cplx.angle().permute(0, 2, 1)\n",
    "\n",
    "# #         mag = sig_stft[:,:,:,0].permute(0, 2, 1)\n",
    "# #         angle = sig_stft[:,:,:,1].permute(0, 2, 1)\n",
    "        \n",
    "#         return [mag, angle]\n",
    "    \n",
    "#     @torch.jit.export\n",
    "#     def istft_layer(self, mag, angle):\n",
    "#         mag = mag.permute(0, 2, 1)\n",
    "#         angle = angle.permute(0, 2, 1)\n",
    "#         mag = torch.unsqueeze(mag, dim=-1)\n",
    "#         angle = torch.unsqueeze(angle, dim=-1)\n",
    "#         pre_stft = torch.cat((mag, angle), dim=-1)\n",
    "#         stft_sig = torch.istft(pre_stft, n_fft=self.n_fft, win_length=self.win_length, hop_length=self.hop_len)\n",
    "        \n",
    "#         return stft_sig\n",
    "    \n",
    "#     @torch.cuda.amp.autocast()\n",
    "#     def forward(self, inp_tensor):\n",
    "        \n",
    "#         mag, angle = self.stft_layer(inp_tensor)\n",
    "#         mag_norm = self.instancenorm1(mag)\n",
    "#         angle_norm = self.instancenorm3(angle)\n",
    "        \n",
    "#         x_mag, hidden_states_mag = self.lstm1(mag_norm)\n",
    "#         x_angle, hidden_states_angle = self.lstm3(angle_norm)\n",
    "\n",
    "#         mask_mag = F.relu(self.dense1(x_mag))\n",
    "#         estimated_mag = torch.mul(mag, mask_mag)\n",
    "        \n",
    "#         mask_angle = F.relu(self.dense3(x_angle))\n",
    "#         estimated_angle = torch.mul(angle, mask_angle)\n",
    "        \n",
    "#         signal = self.istft_layer(estimated_mag, estimated_angle)\n",
    "#         signal = signal.unsqueeze(dim=1)\n",
    "\n",
    "#         feature_rep = self.conv1(signal)\n",
    "#         feature_rep = self.prelu_conv2(feature_rep)\n",
    "        \n",
    "#         feature_norm = self.instancenorm2(feature_rep)\n",
    "#         feature_norm = feature_norm.permute(0, 2, 1)\n",
    "#         x, hidden_states = self.lstm2(feature_norm)\n",
    "#         mask = self.dense2(x)\n",
    "#         feature_mask = F.relu(mask)\n",
    "#         feature_mask = feature_mask.permute(0, 2, 1)\n",
    "\n",
    "#         estimate_feat = torch.mul(feature_rep, feature_mask)\n",
    "        \n",
    "#         estimate_frames = (self.conv2(estimate_feat))\n",
    "#         estimate_frames = self.prelu_conv2(estimate_frames)\n",
    "#         estimate_sig = torch.sum(estimate_frames, dim=1)\n",
    "        \n",
    "#         return estimate_sig\n",
    "    \n",
    "#     def training_step(self, batch, batch_idx):\n",
    "#         x, y = batch\n",
    "#         y_pred = self(x.unsqueeze(dim=1).to(device))\n",
    "#         loss = self.loss_fn(y_pred, y.unsqueeze(dim=1))\n",
    "\n",
    "#         return loss\n",
    "#     def validation_step(self, batch, batch_idx):\n",
    "#         x, y = batch\n",
    "#         y_pred = self(x.to(device))\n",
    "        \n",
    "#         loss = self.loss_fn(y_pred, y)\n",
    "#         self.log(\"val_loss\", loss, enable_graph=True, on_epoch=True, prog_bar=True)\n",
    "    \n",
    "#     def configure_optimizers(self):\n",
    "#         return torch.optim.AdamW(self.parameters(), lr=0.001, weight_decay=5e-5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DownSamplingLayer(nn.Module):\n",
    "    def __init__(self, channel_in, channel_out, dilation=1, kernel_size=15, stride=1, padding=7):\n",
    "        super(DownSamplingLayer, self).__init__()\n",
    "        self.main = nn.Sequential(\n",
    "            nn.Conv1d(channel_in, channel_out, kernel_size=kernel_size,\n",
    "                      stride=stride, padding=padding, dilation=dilation),\n",
    "            nn.BatchNorm1d(channel_out),\n",
    "            nn.LeakyReLU(negative_slope=0.1)\n",
    "        )\n",
    "\n",
    "    def forward(self, ipt):\n",
    "        return self.main(ipt)\n",
    "\n",
    "class UpSamplingLayer(nn.Module):\n",
    "    def __init__(self, channel_in, channel_out, kernel_size=5, stride=1, padding=2):\n",
    "        super(UpSamplingLayer, self).__init__()\n",
    "        self.main = nn.Sequential(\n",
    "            nn.Conv1d(channel_in, channel_out, kernel_size=kernel_size,\n",
    "                      stride=stride, padding=padding),\n",
    "            nn.BatchNorm1d(channel_out),\n",
    "            nn.LeakyReLU(negative_slope=0.1, inplace=True),\n",
    "        )\n",
    "\n",
    "    def forward(self, ipt):\n",
    "        return self.main(ipt)\n",
    "\n",
    "class Model(nn.Module):\n",
    "    def __init__(self, n_layers=12, channels_interval=24):\n",
    "        super(Model, self).__init__()\n",
    "        self.n_layers = n_layers\n",
    "        self.channels_interval = channels_interval\n",
    "        encoder_in_channels_list = [1] + [i * self.channels_interval for i in range(1, self.n_layers)]\n",
    "        encoder_out_channels_list = [i * self.channels_interval for i in range(1, self.n_layers + 1)]\n",
    "        range_list = [i for i in range(self.n_layers)]\n",
    "        #          1    => 2    => 3    => 4    => 5    => 6   => 7   => 8   => 9  => 10 => 11 =>12\n",
    "        # 16384 => 8192 => 4096 => 2048 => 1024 => 512 => 256 => 128 => 64 => 32 => 16 =>  8 => 4\n",
    "        self.encoder = nn.ModuleList()\n",
    "        for i in range(self.n_layers):\n",
    "            self.encoder.append(\n",
    "                DownSamplingLayer(\n",
    "                    channel_in=encoder_in_channels_list[i],\n",
    "                    channel_out=encoder_out_channels_list[i]\n",
    "                )\n",
    "            )\n",
    "\n",
    "        self.middle = nn.Sequential(\n",
    "            nn.Conv1d(self.n_layers * self.channels_interval, self.n_layers * self.channels_interval, 15, stride=1,\n",
    "                      padding=7),\n",
    "            nn.BatchNorm1d(self.n_layers * self.channels_interval),\n",
    "            nn.LeakyReLU(negative_slope=0.1, inplace=True)\n",
    "        )\n",
    "\n",
    "        decoder_in_channels_list = [(2 * i + 1) * self.channels_interval for i in range(1, self.n_layers)] + [\n",
    "            2 * self.n_layers * self.channels_interval]\n",
    "        decoder_in_channels_list = decoder_in_channels_list[::-1]\n",
    "        decoder_out_channels_list = encoder_out_channels_list[::-1]\n",
    "        self.decoder = nn.ModuleList()\n",
    "        for i in range(self.n_layers):\n",
    "            self.decoder.append(\n",
    "                UpSamplingLayer(\n",
    "                    channel_in=decoder_in_channels_list[i],\n",
    "                    channel_out=decoder_out_channels_list[i]\n",
    "                )\n",
    "            )\n",
    "\n",
    "        self.out = nn.Sequential(\n",
    "            nn.Conv1d(1 + self.channels_interval, 1, kernel_size=1, stride=1),\n",
    "            nn.Sigmoid()\n",
    "        )\n",
    "        \n",
    "    def forward(self, input):\n",
    "        tmp = []\n",
    "        o = input\n",
    "        for i, _ in enumerate(self.encoder.children()):\n",
    "            o = self.encoder[i](o)\n",
    "            tmp.append(o)\n",
    "            o = o[:, :, ::2]\n",
    "\n",
    "        o = self.middle(o)\n",
    "\n",
    "        for i, _ in enumerate(self.decoder.children()):\n",
    "            o = F.interpolate(o, scale_factor=2, mode=\"linear\", align_corners=True)\n",
    "            o = torch.cat([o, tmp[self.n_layers - i -1]], dim=1)\n",
    "            o = self.decoder[i](o)\n",
    "\n",
    "        o = torch.cat([o, input], dim=1)\n",
    "        o = self.out(o)\n",
    "        return o\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using Primary model\n"
     ]
    }
   ],
   "source": [
    "use_scripted_model = False\n",
    "w_decay = 1e-4\n",
    "\n",
    "if not use_scripted_model:\n",
    "    print(\"Using Primary model\")\n",
    "    model = Model(12, 24)#NoiseReducer(default_sr=default_sr, n_fft=n_fft, win_length=win_length, hop_len=hop_len, sec=sec).to(device)\n",
    "    model#.to(device)\n",
    "else:\n",
    "    print(\"Using Scripted Model\")\n",
    "    model = torch.jit.script(Model())\n",
    "\n",
    "optimizer = torch.optim.AdamW(model.parameters(), lr=0.0001, weight_decay=0)\n",
    "criterion = Mod_MSELoss(mul_factor=1000)\n",
    "n_epochs=100\n",
    "\n",
    "\n",
    "model.train()\n",
    "scaler = torch.cuda.amp.GradScaler()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = Learner(dls, model, loss_func=criterion, model_dir=\"./Model_saves/\", \\\n",
    "                cbs=[ShowGraphCallback(), SaveModelCallback(monitor='valid_loss', fname='bestmodel', every_epoch=True)]).to_fp16()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# learn.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# learn.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.lr = 0.33113112449646#0.19054607152938843"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# learn.load(\"cnn_model_best_1_78\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div>\n",
       "        <style>\n",
       "            /* Turns off some styling */\n",
       "            progress {\n",
       "                /* gets rid of default border in Firefox and Opera. */\n",
       "                border: none;\n",
       "                /* Needs to be in here for Safari polyfill so background images work as expected. */\n",
       "                background-size: auto;\n",
       "            }\n",
       "            .progress-bar-interrupted, .progress-bar-interrupted::-webkit-progress-bar {\n",
       "                background: #F44336;\n",
       "            }\n",
       "        </style>\n",
       "      <progress value='0' class='' max='20' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
       "      0.00% [0/20 00:00<00:00]\n",
       "    </div>\n",
       "    \n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table><p>\n",
       "\n",
       "    <div>\n",
       "        <style>\n",
       "            /* Turns off some styling */\n",
       "            progress {\n",
       "                /* gets rid of default border in Firefox and Opera. */\n",
       "                border: none;\n",
       "                /* Needs to be in here for Safari polyfill so background images work as expected. */\n",
       "                background-size: auto;\n",
       "            }\n",
       "            .progress-bar-interrupted, .progress-bar-interrupted::-webkit-progress-bar {\n",
       "                background: #F44336;\n",
       "            }\n",
       "        </style>\n",
       "      <progress value='4163' class='' max='770604' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
       "      0.54% [4163/770604 1:57:10<359:33:29 1.3296]\n",
       "    </div>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit_one_cycle(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.save(\"learner_cnn_save\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_state_dict(torch.load(\"./Model_saves/Pytorch_model_2_save_LSTM_512_filters.pt\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "optimizer.closure()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.save(model.state_dict(), \"./Model_saves/Pytorch_model_2_save_LSTM_512_filters.pt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%time\n",
    "noise_add_sig, main_sig = val_ds.__getitem__(2345)\n",
    "noise_add_sig = torch.unsqueeze(noise_add_sig, dim=0).to(device)\n",
    "main_sig = torch.unsqueeze(main_sig, dim=0).to(device)\n",
    "\n",
    "with torch.no_grad():\n",
    "    outs = learn.model(noise_add_sig).squeeze()\n",
    "\n",
    "outs.shape\n",
    "\n",
    "x = outs.t().to(\"cpu\").numpy()\n",
    "ipd.Audio(x, rate=default_sr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "outs.squeeze().t().to(\"cpu\").numpy().shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = noise_add_sig.squeeze().t().to(\"cpu\").numpy()\n",
    "ipd.Audio(x, rate=default_sr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = main_sig.squeeze().t().to(\"cpu\").numpy()\n",
    "ipd.Audio(x, rate=default_sr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "noise_add_sig.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}