{"cells":[{"metadata":{},"cell_type":"markdown","source":"# Birdsong Pytorch Baseline: ResNeSt50-fast (Inference)","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"## About\n\nIn this notebook, I try ResNeSt, which is the one of state of the art in image recognition.  \n\nThis is a notebook for **_inference & submission_**. I shared training process as another one:  \nhttps://www.kaggle.com/ttahara/training-birdsong-baseline-resnest50-fast  \nIf you want to know experimental details, see it.\n\nMost of this notebook consists of [great baseline](https://www.kaggle.com/hidehisaarai1213/inference-pytorch-birdcall-resnet-baseline) shared by @hidehisaarai1213 .  \nThank you for sharing !","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"## Prepare","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"### import libraries","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"pip install ../input/resnest50-fast-package/resnest-0.0.6b20200701/resnest/","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"!pip install ../input/pydub0241/pydub-0.24.1-py2.py3-none-any.whl","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"import os\nimport gc\nimport time\nimport math\nimport shutil\nimport random\nimport warnings\nimport typing as tp\nfrom pathlib import Path\nfrom contextlib import contextmanager\nfrom pydub import AudioSegment\n\nimport yaml\nfrom joblib import delayed, Parallel\n\nimport cv2\nimport librosa\nimport audioread\nimport soundfile as sf\n\nimport numpy as np\nimport pandas as pd\n\nfrom fastprogress import progress_bar\nfrom sklearn.metrics import f1_score\nfrom sklearn.model_selection import StratifiedKFold\n\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.nn import Conv2d, Module, Linear, BatchNorm2d, ReLU\nfrom torch.nn.modules.utils import _pair\nimport torch.utils.data as data\n\npd.options.display.max_rows = 500\npd.options.display.max_columns = 500\nfrom sklearn.preprocessing import LabelEncoder\nimport resnest.torch as resnest_torch\nimport matplotlib.pyplot as plt","execution_count":null,"outputs":[]},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","collapsed":true,"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":false},"cell_type":"markdown","source":"### define utilities","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"def set_seed(seed: int = 42):\n    random.seed(seed)\n    np.random.seed(seed)\n    os.environ[\"PYTHONHASHSEED\"] = str(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed(seed)  # type: ignore\n#     torch.backends.cudnn.deterministic = True  # type: ignore\n#     torch.backends.cudnn.benchmark = True  # type: ignore\n    \n\n@contextmanager\ndef timer(name: str) -> None:\n    \"\"\"Timer Util\"\"\"\n    t0 = time.time()\n    print(\"[{}] start\".format(name))\n    yield\n    print(\"[{}] done in {:.0f} s\".format(name, time.time() - t0))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# logger = get_logger(\"main.log\")\nset_seed(1213)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### read data","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"ROOT = Path.cwd().parent\nINPUT_ROOT = ROOT / \"input\"\nRAW_DATA = INPUT_ROOT / \"birdsong-recognition\"\nTRAIN_AUDIO_DIR = RAW_DATA / \"train_audio\"\n# TRAIN_RESAMPLED_AUDIO_DIRS = [\n#   INPUT_ROOT / \"birdsong-resampled-train-audio-{:0>2}\".format(i)  for i in range(5)\n# ]\nTEST_AUDIO_DIR = RAW_DATA / \"test_audio\"","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train = pd.read_csv(RAW_DATA / \"train.csv\")","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"if not TEST_AUDIO_DIR.exists():\n    TEST_AUDIO_DIR = INPUT_ROOT / \"birdcall-check\" / \"test_audio\"\n    test = pd.read_csv(INPUT_ROOT / \"birdcall-check\" / \"test.csv\")\nelse:\n    test = pd.read_csv(RAW_DATA / \"test.csv\")","execution_count":null,"outputs":[]},{"metadata":{"_kg_hide-output":true,"trusted":true},"cell_type":"code","source":"train.head()","execution_count":null,"outputs":[]},{"metadata":{"_kg_hide-output":true,"trusted":true},"cell_type":"code","source":"test.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"sub = pd.read_csv(\"../input/birdsong-recognition/sample_submission.csv\")\nsub.to_csv(\"submission.csv\", index=False)  # this will be overwritten if everything goes well","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### set parameters","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"TARGET_SR = 32000\nmodel_config = {\n    \"base_model_name\": \"resnest50_fast_1s1x64d\",\n    \"pretrained\": False,\n    \"num_classes\": 264,\n    \"trained_weights\": \"../input/inception/fold_5.bin\"\n}\n\nmelspectrogram_parameters = {\n    \"n_mels\": 128,\n    \"fmin\": 20,\n    \"fmax\": 16000\n}","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Definition","execution_count":null},{"metadata":{},"cell_type":"markdown","source":"### Dataset\n\nFor `site_3`, I decided to use the same procedure as I did for `site_1` and `site_2`, which is, crop 5 seconds out of the clip and provide prediction on that short clip.\nThe only difference is that I crop 5 seconds short clip from start to the end of the `site_3` clip and aggeregate predictions for each short clip after I did prediction for all those short clips.","execution_count":null},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"BIRD_CODE = {\n    'aldfly': 0, 'ameavo': 1, 'amebit': 2, 'amecro': 3, 'amegfi': 4,\n    'amekes': 5, 'amepip': 6, 'amered': 7, 'amerob': 8, 'amewig': 9,\n    'amewoo': 10, 'amtspa': 11, 'annhum': 12, 'astfly': 13, 'baisan': 14,\n    'baleag': 15, 'balori': 16, 'banswa': 17, 'barswa': 18, 'bawwar': 19,\n    'belkin1': 20, 'belspa2': 21, 'bewwre': 22, 'bkbcuc': 23, 'bkbmag1': 24,\n    'bkbwar': 25, 'bkcchi': 26, 'bkchum': 27, 'bkhgro': 28, 'bkpwar': 29,\n    'bktspa': 30, 'blkpho': 31, 'blugrb1': 32, 'blujay': 33, 'bnhcow': 34,\n    'boboli': 35, 'bongul': 36, 'brdowl': 37, 'brebla': 38, 'brespa': 39,\n    'brncre': 40, 'brnthr': 41, 'brthum': 42, 'brwhaw': 43, 'btbwar': 44,\n    'btnwar': 45, 'btywar': 46, 'buffle': 47, 'buggna': 48, 'buhvir': 49,\n    'bulori': 50, 'bushti': 51, 'buwtea': 52, 'buwwar': 53, 'cacwre': 54,\n    'calgul': 55, 'calqua': 56, 'camwar': 57, 'cangoo': 58, 'canwar': 59,\n    'canwre': 60, 'carwre': 61, 'casfin': 62, 'caster1': 63, 'casvir': 64,\n    'cedwax': 65, 'chispa': 66, 'chiswi': 67, 'chswar': 68, 'chukar': 69,\n    'clanut': 70, 'cliswa': 71, 'comgol': 72, 'comgra': 73, 'comloo': 74,\n    'commer': 75, 'comnig': 76, 'comrav': 77, 'comred': 78, 'comter': 79,\n    'comyel': 80, 'coohaw': 81, 'coshum': 82, 'cowscj1': 83, 'daejun': 84,\n    'doccor': 85, 'dowwoo': 86, 'dusfly': 87, 'eargre': 88, 'easblu': 89,\n    'easkin': 90, 'easmea': 91, 'easpho': 92, 'eastow': 93, 'eawpew': 94,\n    'eucdov': 95, 'eursta': 96, 'evegro': 97, 'fiespa': 98, 'fiscro': 99,\n    'foxspa': 100, 'gadwal': 101, 'gcrfin': 102, 'gnttow': 103, 'gnwtea': 104,\n    'gockin': 105, 'gocspa': 106, 'goleag': 107, 'grbher3': 108, 'grcfly': 109,\n    'greegr': 110, 'greroa': 111, 'greyel': 112, 'grhowl': 113, 'grnher': 114,\n    'grtgra': 115, 'grycat': 116, 'gryfly': 117, 'haiwoo': 118, 'hamfly': 119,\n    'hergul': 120, 'herthr': 121, 'hoomer': 122, 'hoowar': 123, 'horgre': 124,\n    'horlar': 125, 'houfin': 126, 'houspa': 127, 'houwre': 128, 'indbun': 129,\n    'juntit1': 130, 'killde': 131, 'labwoo': 132, 'larspa': 133, 'lazbun': 134,\n    'leabit': 135, 'leafly': 136, 'leasan': 137, 'lecthr': 138, 'lesgol': 139,\n    'lesnig': 140, 'lesyel': 141, 'lewwoo': 142, 'linspa': 143, 'lobcur': 144,\n    'lobdow': 145, 'logshr': 146, 'lotduc': 147, 'louwat': 148, 'macwar': 149,\n    'magwar': 150, 'mallar3': 151, 'marwre': 152, 'merlin': 153, 'moublu': 154,\n    'mouchi': 155, 'moudov': 156, 'norcar': 157, 'norfli': 158, 'norhar2': 159,\n    'normoc': 160, 'norpar': 161, 'norpin': 162, 'norsho': 163, 'norwat': 164,\n    'nrwswa': 165, 'nutwoo': 166, 'olsfly': 167, 'orcwar': 168, 'osprey': 169,\n    'ovenbi1': 170, 'palwar': 171, 'pasfly': 172, 'pecsan': 173, 'perfal': 174,\n    'phaino': 175, 'pibgre': 176, 'pilwoo': 177, 'pingro': 178, 'pinjay': 179,\n    'pinsis': 180, 'pinwar': 181, 'plsvir': 182, 'prawar': 183, 'purfin': 184,\n    'pygnut': 185, 'rebmer': 186, 'rebnut': 187, 'rebsap': 188, 'rebwoo': 189,\n    'redcro': 190, 'redhea': 191, 'reevir1': 192, 'renpha': 193, 'reshaw': 194,\n    'rethaw': 195, 'rewbla': 196, 'ribgul': 197, 'rinduc': 198, 'robgro': 199,\n    'rocpig': 200, 'rocwre': 201, 'rthhum': 202, 'ruckin': 203, 'rudduc': 204,\n    'rufgro': 205, 'rufhum': 206, 'rusbla': 207, 'sagspa1': 208, 'sagthr': 209,\n    'savspa': 210, 'saypho': 211, 'scatan': 212, 'scoori': 213, 'semplo': 214,\n    'semsan': 215, 'sheowl': 216, 'shshaw': 217, 'snobun': 218, 'snogoo': 219,\n    'solsan': 220, 'sonspa': 221, 'sora': 222, 'sposan': 223, 'spotow': 224,\n    'stejay': 225, 'swahaw': 226, 'swaspa': 227, 'swathr': 228, 'treswa': 229,\n    'truswa': 230, 'tuftit': 231, 'tunswa': 232, 'veery': 233, 'vesspa': 234,\n    'vigswa': 235, 'warvir': 236, 'wesblu': 237, 'wesgre': 238, 'weskin': 239,\n    'wesmea': 240, 'wessan': 241, 'westan': 242, 'wewpew': 243, 'whbnut': 244,\n    'whcspa': 245, 'whfibi': 246, 'whtspa': 247, 'whtswi': 248, 'wilfly': 249,\n    'wilsni1': 250, 'wiltur': 251, 'winwre3': 252, 'wlswar': 253, 'wooduc': 254,\n    'wooscj2': 255, 'woothr': 256, 'y00475': 257, 'yebfly': 258, 'yebsap': 259,\n    'yehbla': 260, 'yelwar': 261, 'yerwar': 262, 'yetvir': 263\n}\n\nINV_BIRD_CODE = {v: k for k, v in BIRD_CODE.items()}","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"class TestDataset(data.Dataset):\n    def __init__(self, df: pd.DataFrame, clip: np.ndarray,\n                 img_size=224, melspectrogram_parameters={}):\n        self.df = df\n        self.clip = clip\n        self.img_size = img_size\n        self.melspectrogram_parameters = melspectrogram_parameters\n        \n    def __len__(self):\n        return len(self.df)\n    \n    def __getitem__(self, idx: int):\n        SR = 32000\n        sample = self.df.loc[idx, :]\n        site = sample.site\n        row_id = sample.row_id\n        \n        if site == \"site_3\":\n            y = self.clip.astype(np.float32)\n            len_y = len(y)\n            start = 0\n            end = SR * 5\n            images = []\n            while len_y > start:\n                y_batch = y[start:end].astype(np.float32)\n                if len(y_batch) != (SR * 5):\n                    break\n                start = end\n                end = end + SR * 5\n                \n                melspec = librosa.feature.melspectrogram(y_batch,\n                                                         sr=SR,\n                                                         **self.melspectrogram_parameters)\n                melspec = librosa.power_to_db(melspec).astype(np.float32)\n                image=melspec\n                delta = librosa.feature.delta(image)\n                accelerate = librosa.feature.delta(image, order=2)\n                image = np.stack([image, delta, accelerate], axis=0)\n                image = image.astype(np.float32) / 100.0\n                images.append(image)\n                \n            images = np.asarray(images)\n            return images, row_id, site\n        else:\n            end_seconds = int(sample.seconds)\n            start_seconds = int(end_seconds - 5)\n            \n            start_index = SR * start_seconds\n            end_index = SR * end_seconds\n            \n            y = self.clip[start_index:end_index].astype(np.float32)\n\n            melspec = librosa.feature.melspectrogram(y, sr=SR, **self.melspectrogram_parameters)\n            melspec = librosa.power_to_db(melspec).astype(np.float32)\n\n            image=melspec\n            delta = librosa.feature.delta(image)\n            accelerate = librosa.feature.delta(image, order=2)\n            image = np.stack([image, delta, accelerate], axis=0)\n            image = image.astype(np.float32) / 100.0\n\n            return image, row_id, site","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### model\n\n* I forked this code from authors' original implementation. [GitHub](https://github.com/zhanghang1989/ResNeSt)","execution_count":null},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_model1(args: tp.Dict):\n\n    model = getattr(resnest_torch,'resnest50_fast_1s1x64d')(pretrained=False)\n    del model.fc\n    model.fc = nn.Sequential(\n        nn.Linear(2048, 1024),\n        nn.ReLU(),\n        nn.Dropout(p=0.2),\n        nn.Linear(1024, 1024),\n        nn.ReLU(),\n        nn.Dropout(p=0.2),\n        nn.Linear(1024, args[\"num_classes\"]),\n    )\n    \n    state_dict = torch.load(args[\"trained_weights\"])\n    model.load_state_dict(state_dict)\n    device = torch.device(\"cuda\")\n    model.to(device)\n    model.eval()\n    \n    return model","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"class get_model(nn.Module):\n \n    def __init__(self, num_classes=264, aux_logits=True, transform_input=False):\n        super(get_model, self).__init__()\n        self.aux_logits = aux_logits\n        self.transform_input = transform_input\n        self.Conv2d_1a_3x3 = BasicConv2d(3, 32, kernel_size=3, stride=2)\n        self.Conv2d_2a_3x3 = BasicConv2d(32, 32, kernel_size=3)\n        self.Conv2d_2b_3x3 = BasicConv2d(32, 64, kernel_size=3, padding=1)\n        self.Conv2d_3b_1x1 = BasicConv2d(64, 80, kernel_size=1)\n        self.Conv2d_4a_3x3 = BasicConv2d(80, 192, kernel_size=3)\n        self.Mixed_5b = InceptionA(192, pool_features=32)\n        self.Mixed_5c = InceptionA(256, pool_features=64)\n        self.Mixed_5d = InceptionA(288, pool_features=64)\n        self.Mixed_6a = InceptionB(288)\n        self.Mixed_6b = InceptionC(768, channels_7x7=128)\n        self.Mixed_6c = InceptionC(768, channels_7x7=160)\n        self.Mixed_6d = InceptionC(768, channels_7x7=160)\n        self.Mixed_6e = InceptionC(768, channels_7x7=192)\n        if aux_logits:\n            self.AuxLogits = InceptionAux(768, num_classes)\n        self.Mixed_7a = InceptionD(768)\n        self.Mixed_7b = InceptionE(1280)\n        self.Mixed_7c = InceptionE(2048)\n        self.fc = nn.Linear(2048, num_classes)\n \n    def forward(self, x):\n        if self.transform_input:\n            x_ch0 = torch.unsqueeze(x[:, 0], 1) * (0.229 / 0.5) + (0.485 - 0.5) / 0.5\n            x_ch1 = torch.unsqueeze(x[:, 1], 1) * (0.224 / 0.5) + (0.456 - 0.5) / 0.5\n            x_ch2 = torch.unsqueeze(x[:, 2], 1) * (0.225 / 0.5) + (0.406 - 0.5) / 0.5\n            x = torch.cat((x_ch0, x_ch1, x_ch2), 1)\n        # 299 x 299 x 3\n        x = self.Conv2d_1a_3x3(x)\n        # 149 x 149 x 32\n        x = self.Conv2d_2a_3x3(x)\n        # 147 x 147 x 32\n        x = self.Conv2d_2b_3x3(x)\n        # 147 x 147 x 64\n        x = F.max_pool2d(x, kernel_size=3, stride=2)\n        # 73 x 73 x 64\n        x = self.Conv2d_3b_1x1(x)\n        # 73 x 73 x 80\n        x = self.Conv2d_4a_3x3(x)\n        # 71 x 71 x 192\n        x = F.max_pool2d(x, kernel_size=3, stride=2)\n        # 35 x 35 x 192\n        x = self.Mixed_5b(x)\n        # 35 x 35 x 256\n        x = self.Mixed_5c(x)\n        # 35 x 35 x 288\n        x = self.Mixed_5d(x)\n        # 35 x 35 x 288\n        x = self.Mixed_6a(x)\n        # 17 x 17 x 768\n        x = self.Mixed_6b(x)\n        # 17 x 17 x 768\n        x = self.Mixed_6c(x)\n        # 17 x 17 x 768\n        x = self.Mixed_6d(x)\n        # 17 x 17 x 768\n        x = self.Mixed_6e(x)\n        # 17 x 17 x 768\n        if self.training and self.aux_logits:\n            aux = self.AuxLogits(x)\n        # 17 x 17 x 768\n        x = self.Mixed_7a(x)\n        # 8 x 8 x 1280\n        x = self.Mixed_7b(x)\n        # 8 x 8 x 2048\n        x = self.Mixed_7c(x)\n        # 8 x 8 x 2048\n        x = F.avg_pool2d(x, kernel_size=(2,8))\n        # 1 x 1 x 2048\n        x = F.dropout(x, training=self.training)\n        # 1 x 1 x 2048\n        x = x.view(x.size(0), -1)\n        # 2048\n        x = self.fc(x)\n        # 1000 (num_classes)\n        if self.training and self.aux_logits:\n            return (x+aux)/2\n        return x\n \n \nclass InceptionA(nn.Module):\n \n    def __init__(self, in_channels, pool_features):\n        super(InceptionA, self).__init__()\n        self.branch1x1 = BasicConv2d(in_channels, 64, kernel_size=1)\n \n        self.branch5x5_1 = BasicConv2d(in_channels, 48, kernel_size=1)\n        self.branch5x5_2 = BasicConv2d(48, 64, kernel_size=5, padding=2)\n \n        self.branch3x3dbl_1 = BasicConv2d(in_channels, 64, kernel_size=1)\n        self.branch3x3dbl_2 = BasicConv2d(64, 96, kernel_size=3, padding=1)\n        self.branch3x3dbl_3 = BasicConv2d(96, 96, kernel_size=3, padding=1)\n \n        self.branch_pool = BasicConv2d(in_channels, pool_features, kernel_size=1)\n \n    def forward(self, x):\n        branch1x1 = self.branch1x1(x)\n \n        branch5x5 = self.branch5x5_1(x)\n        branch5x5 = self.branch5x5_2(branch5x5)\n \n        branch3x3dbl = self.branch3x3dbl_1(x)\n        branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)\n        branch3x3dbl = self.branch3x3dbl_3(branch3x3dbl)\n \n        branch_pool = F.avg_pool2d(x, kernel_size=3, stride=1, padding=1)\n        branch_pool = self.branch_pool(branch_pool)\n \n        outputs = [branch1x1, branch5x5, branch3x3dbl, branch_pool]\n        return torch.cat(outputs, 1)\n \n \nclass InceptionB(nn.Module):\n \n    def __init__(self, in_channels):\n        super(InceptionB, self).__init__()\n        self.branch3x3 = BasicConv2d(in_channels, 384, kernel_size=3, stride=2)\n \n        self.branch3x3dbl_1 = BasicConv2d(in_channels, 64, kernel_size=1)\n        self.branch3x3dbl_2 = BasicConv2d(64, 96, kernel_size=3, padding=1)\n        self.branch3x3dbl_3 = BasicConv2d(96, 96, kernel_size=3, stride=2)\n \n    def forward(self, x):\n        branch3x3 = self.branch3x3(x)\n \n        branch3x3dbl = self.branch3x3dbl_1(x)\n        branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)\n        branch3x3dbl = self.branch3x3dbl_3(branch3x3dbl)\n \n        branch_pool = F.max_pool2d(x, kernel_size=3, stride=2)\n \n        outputs = [branch3x3, branch3x3dbl, branch_pool]\n        return torch.cat(outputs, 1)\n \n \nclass InceptionC(nn.Module):\n \n    def __init__(self, in_channels, channels_7x7):\n        super(InceptionC, self).__init__()\n        self.branch1x1 = BasicConv2d(in_channels, 192, kernel_size=1)\n \n        c7 = channels_7x7\n        self.branch7x7_1 = BasicConv2d(in_channels, c7, kernel_size=1)\n        self.branch7x7_2 = BasicConv2d(c7, c7, kernel_size=(1, 7), padding=(0, 3))\n        self.branch7x7_3 = BasicConv2d(c7, 192, kernel_size=(7, 1), padding=(3, 0))\n \n        self.branch7x7dbl_1 = BasicConv2d(in_channels, c7, kernel_size=1)\n        self.branch7x7dbl_2 = BasicConv2d(c7, c7, kernel_size=(7, 1), padding=(3, 0))\n        self.branch7x7dbl_3 = BasicConv2d(c7, c7, kernel_size=(1, 7), padding=(0, 3))\n        self.branch7x7dbl_4 = BasicConv2d(c7, c7, kernel_size=(7, 1), padding=(3, 0))\n        self.branch7x7dbl_5 = BasicConv2d(c7, 192, kernel_size=(1, 7), padding=(0, 3))\n \n        self.branch_pool = BasicConv2d(in_channels, 192, kernel_size=1)\n \n    def forward(self, x):\n        branch1x1 = self.branch1x1(x)\n \n        branch7x7 = self.branch7x7_1(x)\n        branch7x7 = self.branch7x7_2(branch7x7)\n        branch7x7 = self.branch7x7_3(branch7x7)\n \n        branch7x7dbl = self.branch7x7dbl_1(x)\n        branch7x7dbl = self.branch7x7dbl_2(branch7x7dbl)\n        branch7x7dbl = self.branch7x7dbl_3(branch7x7dbl)\n        branch7x7dbl = self.branch7x7dbl_4(branch7x7dbl)\n        branch7x7dbl = self.branch7x7dbl_5(branch7x7dbl)\n \n        branch_pool = F.avg_pool2d(x, kernel_size=3, stride=1, padding=1)\n        branch_pool = self.branch_pool(branch_pool)\n \n        outputs = [branch1x1, branch7x7, branch7x7dbl, branch_pool]\n        return torch.cat(outputs, 1)\n \n \nclass InceptionD(nn.Module):\n \n    def __init__(self, in_channels):\n        super(InceptionD, self).__init__()\n        self.branch3x3_1 = BasicConv2d(in_channels, 192, kernel_size=1)\n        self.branch3x3_2 = BasicConv2d(192, 320, kernel_size=3, stride=2)\n \n        self.branch7x7x3_1 = BasicConv2d(in_channels, 192, kernel_size=1)\n        self.branch7x7x3_2 = BasicConv2d(192, 192, kernel_size=(1, 7), padding=(0, 3))\n        self.branch7x7x3_3 = BasicConv2d(192, 192, kernel_size=(7, 1), padding=(3, 0))\n        self.branch7x7x3_4 = BasicConv2d(192, 192, kernel_size=3, stride=2)\n \n    def forward(self, x):\n        branch3x3 = self.branch3x3_1(x)\n        branch3x3 = self.branch3x3_2(branch3x3)\n \n        branch7x7x3 = self.branch7x7x3_1(x)\n        branch7x7x3 = self.branch7x7x3_2(branch7x7x3)\n        branch7x7x3 = self.branch7x7x3_3(branch7x7x3)\n        branch7x7x3 = self.branch7x7x3_4(branch7x7x3)\n \n        branch_pool = F.max_pool2d(x, kernel_size=3, stride=2)\n        outputs = [branch3x3, branch7x7x3, branch_pool]\n        return torch.cat(outputs, 1)\n \n \nclass InceptionE(nn.Module):\n \n    def __init__(self, in_channels):\n        super(InceptionE, self).__init__()\n        self.branch1x1 = BasicConv2d(in_channels, 320, kernel_size=1)\n \n        self.branch3x3_1 = BasicConv2d(in_channels, 384, kernel_size=1)\n        self.branch3x3_2a = BasicConv2d(384, 384, kernel_size=(1, 3), padding=(0, 1))\n        self.branch3x3_2b = BasicConv2d(384, 384, kernel_size=(3, 1), padding=(1, 0))\n \n        self.branch3x3dbl_1 = BasicConv2d(in_channels, 448, kernel_size=1)\n        self.branch3x3dbl_2 = BasicConv2d(448, 384, kernel_size=3, padding=1)\n        self.branch3x3dbl_3a = BasicConv2d(384, 384, kernel_size=(1, 3), padding=(0, 1))\n        self.branch3x3dbl_3b = BasicConv2d(384, 384, kernel_size=(3, 1), padding=(1, 0))\n \n        self.branch_pool = BasicConv2d(in_channels, 192, kernel_size=1)\n \n    def forward(self, x):\n        branch1x1 = self.branch1x1(x)\n \n        branch3x3 = self.branch3x3_1(x)\n        branch3x3 = [\n            self.branch3x3_2a(branch3x3),\n            self.branch3x3_2b(branch3x3),\n        ]\n        branch3x3 = torch.cat(branch3x3, 1)\n \n        branch3x3dbl = self.branch3x3dbl_1(x)\n        branch3x3dbl = self.branch3x3dbl_2(branch3x3dbl)\n        branch3x3dbl = [\n            self.branch3x3dbl_3a(branch3x3dbl),\n            self.branch3x3dbl_3b(branch3x3dbl),\n        ]\n        branch3x3dbl = torch.cat(branch3x3dbl, 1)\n \n        branch_pool = F.avg_pool2d(x, kernel_size=3, stride=1, padding=1)\n        branch_pool = self.branch_pool(branch_pool)\n \n        outputs = [branch1x1, branch3x3, branch3x3dbl, branch_pool]\n        return torch.cat(outputs, 1)\n \n \nclass InceptionAux(nn.Module):\n \n    def __init__(self, in_channels, num_classes):\n        super(InceptionAux, self).__init__()\n        self.conv0 = BasicConv2d(in_channels, 128, kernel_size=1)\n        self.conv1 = BasicConv2d(128, 768, kernel_size=(1,5))\n        self.conv1.stddev = 0.01\n        self.fc = nn.Linear(768, num_classes)\n        self.fc.stddev = 0.001\n \n    def forward(self, x):\n        # 17 x 17 x 768\n        x = F.avg_pool2d(x, kernel_size=5, stride=3)\n        # 5 x 5 x 768\n        x = self.conv0(x)\n        # 5 x 5 x 128\n        x = self.conv1(x)\n        # 1 x 1 x 768\n        x = x.view(x.size(0), -1)\n        # 768\n        x = self.fc(x)\n        # 1000\n        return x\n \n \nclass BasicConv2d(nn.Module):\n \n    def __init__(self, in_channels, out_channels, **kwargs):\n        super(BasicConv2d, self).__init__()\n        self.conv = nn.Conv2d(in_channels, out_channels, bias=False, **kwargs)\n        self.bn = nn.BatchNorm2d(out_channels, eps=0.001)\n \n    def forward(self, x):\n        x = self.conv(x)\n        x = self.bn(x)\n        return F.relu(x, inplace=True)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Prediction loop","execution_count":null},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"def prediction_for_clip(test_df: pd.DataFrame, \n                        clip: np.ndarray, \n                        model, \n                        mel_params: dict, \n                        threshold=0.5,\n                        maxpreds=3, # New param --> @kkiller\n                       ):\n    \n    \"\"\"\n    Original code:  @hidehisaarai1213\n    First refacto : @ttahara\n    Second refacto: @kkiller\n    \"\"\"\n\n    dataset = TestDataset(df=test_df, \n                          clip=clip,\n                          img_size=224,\n                          melspectrogram_parameters=mel_params)\n    loader = data.DataLoader(dataset, batch_size=1, shuffle=False)\n    device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n    \n    model.eval()\n    prediction_dict = {}\n    for image, row_id, site in progress_bar(loader):\n        site = site[0]\n        row_id = row_id[0]\n        if site in {\"site_1\", \"site_2\"}:\n            image = image.to(device)\n\n            with torch.no_grad():\n                prediction = F.sigmoid(model(image))\n                proba = prediction.detach().cpu().numpy().reshape(-1)\n\n            events = proba >= threshold\n            labels = np.argsort(-proba)[:events.sum()].tolist()\n\n        else:\n            # to avoid prediction on large batch\n            image = image.squeeze(0)\n            batch_size = 16\n            whole_size = image.size(0)\n            if whole_size % batch_size == 0:\n                n_iter = whole_size // batch_size\n            else:\n                n_iter = whole_size // batch_size + 1\n                \n            probas = []\n            for batch_i in range(n_iter):\n                batch = image[batch_i * batch_size:(batch_i + 1) * batch_size]\n                if batch.ndim == 3:\n                    batch = batch.unsqueeze(0)\n\n                batch = batch.to(device)\n                with torch.no_grad():\n                    prediction = F.sigmoid(model(batch))\n                    proba = prediction.detach().cpu().numpy()\n                    probas.append(proba)\n                \n            probas = np.vstack(probas)\n            probas = probas.max(0)\n            events = (probas>=threshold)\n            labels = np.argsort(-probas)[:events.sum()].tolist()\n            \n        if len(labels) == 0:\n            prediction_dict[row_id] = \"nocall\"\n        else:\n            labels_str_list = list(map(lambda x: INV_BIRD_CODE[x], labels))\n            label_string = \" \".join(labels_str_list[:maxpreds])\n            prediction_dict[row_id] = label_string\n    return prediction_dict","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"def prediction(test_df: pd.DataFrame,\n               test_audio: Path,\n               model_config: dict,\n               mel_params: dict,\n               target_sr: int,\n               threshold=0.5):\n    model = get_model()\n    state_dict = torch.load('../input/inceptio/fold_[3, 4]_epoch_36.bin')\n    model.load_state_dict(state_dict)\n    device = torch.device(\"cuda\")\n    model.to(device)\n    model.eval()\n    \n    unique_audio_id = test_df.audio_id.unique()\n\n    warnings.filterwarnings(\"ignore\")\n    prediction_dfs = []\n    for audio_id in unique_audio_id:\n        with timer(f\"Loading {audio_id}\"):\n            sound = AudioSegment.from_file(test_audio / (audio_id + \".mp3\"),format='mp3')\n            sound = sound.set_frame_rate(target_sr)\n            clip = np.array(sound.get_array_of_samples(), dtype=np.float32)          \n            \n        test_df_for_audio_id = test_df.query(\n            f\"audio_id == '{audio_id}'\").reset_index(drop=True)\n        with timer(f\"Prediction on {audio_id}\"):\n            prediction_dict = prediction_for_clip(test_df_for_audio_id,\n                                                  clip=clip,\n                                                  model=model,\n                                                  mel_params=mel_params,\n                                                  threshold=threshold)\n        row_id = list(prediction_dict.keys())\n        birds = list(prediction_dict.values())\n        prediction_df = pd.DataFrame({\n            \"row_id\": row_id,\n            \"birds\": birds\n        })\n        prediction_dfs.append(prediction_df)\n    \n    prediction_df = pd.concat(prediction_dfs, axis=0, sort=False).reset_index(drop=True)\n    return prediction_df","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Prediction","execution_count":null},{"metadata":{"trusted":true,"_kg_hide-output":true},"cell_type":"code","source":"submission = prediction(test_df=test,\n                        test_audio=TEST_AUDIO_DIR,\n                        model_config=model_config,\n                        mel_params=melspectrogram_parameters,\n                        target_sr=TARGET_SR,\n                        threshold=0.1)\nsubmission.to_csv(\"submission.csv\", index=False)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true,"_kg_hide-output":true},"cell_type":"code","source":"submission","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"markdown","source":"## EOF","execution_count":null}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}