{"cells":[{"metadata":{},"cell_type":"markdown","source":"All cridets [@hidehisaarai1213](https://www.kaggle.com/hidehisaarai1213)\n\nThis notebook based on this [Introduction to Sound Event Detection](https://www.kaggle.com/hidehisaarai1213/introduction-to-sound-event-detection)"},{"metadata":{},"cell_type":"markdown","source":"### Install packages"},{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true,"_kg_hide-input":true,"_kg_hide-output":true},"cell_type":"code","source":"!pip -q install --upgrade pip\n!pip -q install timm\n!pip -q install torchlibrosa\n!pip -q install audiomentations","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### import packages"},{"metadata":{"_kg_hide-input":true,"trusted":true},"cell_type":"code","source":"import os, glob, random, time\nimport numpy as np, pandas as pd\nimport matplotlib.pyplot as plt\nimport librosa, librosa.display\nimport soundfile as sf\n\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\n\nfrom tqdm import tqdm\nfrom functools import partial\nfrom sklearn import metrics\nfrom sklearn.model_selection import StratifiedKFold\nfrom transformers import get_linear_schedule_with_warmup\nfrom torchlibrosa.stft import Spectrogram, LogmelFilterBank\nfrom torchlibrosa.augmentation import SpecAugmentation\n\nimport timm\nfrom timm.models.efficientnet import tf_efficientnet_b0_ns","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### About Sound Event Detection(SED)\n\nSound event detection (SED) is the task of detecting the type as well as\nthe onset and offset times of sound events in audio streams.\n\nIn this notebook i will show how to train Sound Event Detection (SED) model with only weak annotation.\n\n![image.png](attachment:image.png)\n\nIn SED task, we need to detect sound events from continuous (long) audio clip, and provide prediction of what sound event exists from when to when.\n\nfor more details\n\n-> [Polyphonic Sound Event Detection\nwith Weak Labeling Paper](http://www.cs.cmu.edu/~yunwang/papers/cmu-thesis.pdf)\n\n-> [Introduction to Sound Event Detection Notebook](https://www.kaggle.com/hidehisaarai1213/introduction-to-sound-event-detection)\n","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{},"cell_type":"markdown","source":"### PANN Utils\n\n-> [PANNs repository](https://github.com/qiuqiangkong/audioset_tagging_cnn/)\n\n-> [PANNs paper](https://arxiv.org/abs/1912.10211)\n"},{"metadata":{"trusted":true,"_kg_hide-input":true},"cell_type":"code","source":"def init_layer(layer):\n    nn.init.xavier_uniform_(layer.weight)\n\n    if hasattr(layer, \"bias\"):\n        if layer.bias is not None:\n            layer.bias.data.fill_(0.)\n\n\ndef init_bn(bn):\n    bn.bias.data.fill_(0.)\n    bn.weight.data.fill_(1.0)\n\n\ndef init_weights(model):\n    classname = model.__class__.__name__\n    if classname.find(\"Conv2d\") != -1:\n        nn.init.xavier_uniform_(model.weight, gain=np.sqrt(2))\n        model.bias.data.fill_(0)\n    elif classname.find(\"BatchNorm\") != -1:\n        model.weight.data.normal_(1.0, 0.02)\n        model.bias.data.fill_(0)\n    elif classname.find(\"GRU\") != -1:\n        for weight in model.parameters():\n            if len(weight.size()) > 1:\n                nn.init.orghogonal_(weight.data)\n    elif classname.find(\"Linear\") != -1:\n        model.weight.data.normal_(0, 0.01)\n        model.bias.data.zero_()\n\ndef do_mixup(x: torch.Tensor, mixup_lambda: torch.Tensor):\n    \"\"\"Mixup x of even indexes (0, 2, 4, ...) with x of odd indexes\n    (1, 3, 5, ...).\n    Args:\n      x: (batch_size * 2, ...)\n      mixup_lambda: (batch_size * 2,)\n    Returns:\n      out: (batch_size, ...)\n    \"\"\"\n    out = (x[0::2].transpose(0, -1) * mixup_lambda[0::2] +\n           x[1::2].transpose(0, -1) * mixup_lambda[1::2]).transpose(0, -1)\n    return out\n\n\nclass Mixup(object):\n    def __init__(self, mixup_alpha, random_seed=1234):\n        \"\"\"Mixup coefficient generator.\n        \"\"\"\n        self.mixup_alpha = mixup_alpha\n        self.random_state = np.random.RandomState(random_seed)\n\n    def get_lambda(self, batch_size):\n        \"\"\"Get mixup random coefficients.\n        Args:\n          batch_size: int\n        Returns:\n          mixup_lambdas: (batch_size,)\n        \"\"\"\n        mixup_lambdas = []\n        for n in range(0, batch_size, 2):\n            lam = self.random_state.beta(self.mixup_alpha, self.mixup_alpha, 1)[0]\n            mixup_lambdas.append(lam)\n            mixup_lambdas.append(1. - lam)\n\n        return torch.from_numpy(np.array(mixup_lambdas, dtype=np.float32))\n\ndef interpolate(x: torch.Tensor, ratio: int):\n    \"\"\"Interpolate data in time domain. This is used to compensate the\n    resolution reduction in downsampling of a CNN.\n    Args:\n      x: (batch_size, time_steps, classes_num)\n      ratio: int, ratio to interpolate\n    Returns:\n      upsampled: (batch_size, time_steps * ratio, classes_num)\n    \"\"\"\n    (batch_size, time_steps, classes_num) = x.shape\n    upsampled = x[:, :, None, :].repeat(1, 1, ratio, 1)\n    upsampled = upsampled.reshape(batch_size, time_steps * ratio, classes_num)\n    return upsampled\n\n\ndef pad_framewise_output(framewise_output: torch.Tensor, frames_num: int):\n    \"\"\"Pad framewise_output to the same length as input frames. The pad value\n    is the same as the value of the last frame.\n    Args:\n      framewise_output: (batch_size, frames_num, classes_num)\n      frames_num: int, number of frames to pad\n    Outputs:\n      output: (batch_size, frames_num, classes_num)\n    \"\"\"\n    pad = framewise_output[:, -1:, :].repeat(\n        1, frames_num - framewise_output.shape[1], 1)\n    \"\"\"tensor for padding\"\"\"\n\n    output = torch.cat((framewise_output, pad), dim=1)\n    \"\"\"(batch_size, frames_num, classes_num)\"\"\"\n\n    return output\n\n\nclass ConvBlock(nn.Module):\n    def __init__(self, in_channels: int, out_channels: int):\n        super().__init__()\n\n        self.conv1 = nn.Conv2d(\n            in_channels=in_channels,\n            out_channels=out_channels,\n            kernel_size=(3, 3),\n            stride=(1, 1),\n            padding=(1, 1),\n            bias=False)\n\n        self.conv2 = nn.Conv2d(\n            in_channels=out_channels,\n            out_channels=out_channels,\n            kernel_size=(3, 3),\n            stride=(1, 1),\n            padding=(1, 1),\n            bias=False)\n\n        self.bn1 = nn.BatchNorm2d(out_channels)\n        self.bn2 = nn.BatchNorm2d(out_channels)\n\n        self.init_weight()\n\n    def init_weight(self):\n        init_layer(self.conv1)\n        init_layer(self.conv2)\n        init_bn(self.bn1)\n        init_bn(self.bn2)\n\n    def forward(self, input, pool_size=(2, 2), pool_type='avg'):\n\n        x = input\n        x = F.relu_(self.bn1(self.conv1(x)))\n        x = F.relu_(self.bn2(self.conv2(x)))\n        if pool_type == 'max':\n            x = F.max_pool2d(x, kernel_size=pool_size)\n        elif pool_type == 'avg':\n            x = F.avg_pool2d(x, kernel_size=pool_size)\n        elif pool_type == 'avg+max':\n            x1 = F.avg_pool2d(x, kernel_size=pool_size)\n            x2 = F.max_pool2d(x, kernel_size=pool_size)\n            x = x1 + x2\n        else:\n            raise Exception('Incorrect argument!')\n\n        return x\n\n\nclass AttBlock(nn.Module):\n    def __init__(self,\n                 in_features: int,\n                 out_features: int,\n                 activation=\"linear\",\n                 temperature=1.0):\n        super().__init__()\n\n        self.activation = activation\n        self.temperature = temperature\n        self.att = nn.Conv1d(\n            in_channels=in_features,\n            out_channels=out_features,\n            kernel_size=1,\n            stride=1,\n            padding=0,\n            bias=True)\n        self.cla = nn.Conv1d(\n            in_channels=in_features,\n            out_channels=out_features,\n            kernel_size=1,\n            stride=1,\n            padding=0,\n            bias=True)\n\n        self.bn_att = nn.BatchNorm1d(out_features)\n        self.init_weights()\n\n    def init_weights(self):\n        init_layer(self.att)\n        init_layer(self.cla)\n        init_bn(self.bn_att)\n\n    def forward(self, x):\n        # x: (n_samples, n_in, n_time)\n        norm_att = torch.softmax(torch.clamp(self.att(x), -10, 10), dim=-1)\n        cla = self.nonlinear_transform(self.cla(x))\n        x = torch.sum(norm_att * cla, dim=2)\n        return x, norm_att, cla\n\n    def nonlinear_transform(self, x):\n        if self.activation == 'linear':\n            return x\n        elif self.activation == 'sigmoid':\n            return torch.sigmoid(x)\n        ","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Create Folds"},{"metadata":{"trusted":true},"cell_type":"code","source":"FOLDS = 5\nSEED = 42\n\ntrain = pd.read_csv(\"../input/rfcx-species-audio-detection/train_tp.csv\").sort_values(\"recording_id\")\nss = pd.read_csv(\"../input/rfcx-species-audio-detection/sample_submission.csv\")\n\ntrain_gby = train.groupby(\"recording_id\")[[\"species_id\"]].first().reset_index()\ntrain_gby = train_gby.sample(frac=1, random_state=SEED).reset_index(drop=True)\ntrain_gby.loc[:, 'kfold'] = -1\n\nX = train_gby[\"recording_id\"].values\ny = train_gby[\"species_id\"].values\n\nkfold = StratifiedKFold(n_splits=FOLDS)\nfor fold, (t_idx, v_idx) in enumerate(kfold.split(X, y)):\n    train_gby.loc[v_idx, \"kfold\"] = fold\n\ntrain = train.merge(train_gby[['recording_id', 'kfold']], on=\"recording_id\", how=\"left\")\nprint(train.kfold.value_counts())\ntrain.to_csv(\"train_folds.csv\", index=False)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### SED Model\n\n1. Model takes raw waveform and converted into log-melspectogram using `torchlibrosa`'s module\n2. spectogram converted into 3-channels input for ImageNet pretrain model to extract features from CNN's\n3. Although it's downsized through several convolution and pooling layers, the size of it's third dimension and it still contains time information. Each element of this dimension is segment. In SED model, we provide prediction for each of this.\n\n![image.png](attachment:image.png)\n\n4. This figure gives us an intuitive explanation what is weak annotation and what is strong annotation in terms of sound event detection. For this competition, we only have weak annotation (clip level annotation). Therefore, we need to train our SED model in weakly-supervised manner.\n\n5. In weakly-supervised setting, we only have clip-level annotation, therefore we also need to aggregate that in time axis. Hense, we at first put classifier that outputs class existence probability for each time step just after the feature extractor and then aggregate the output of the classifier result in time axis. In this way we can get both clip-level prediction and segment-level prediction (if the time resolution is high, it can be treated as event-level prediction). Then we train it normally by using BCE loss with clip-level prediction and clip-level annotation.","attachments":{"image.png":{"image/png":"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"}}},{"metadata":{"trusted":true},"cell_type":"code","source":"encoder_params = {\n    \"tf_efficientnet_b0_ns\": {\n        \"features\": 1280,\n        \"init_op\": partial(tf_efficientnet_b0_ns, pretrained=True, drop_path_rate=0.2)\n    }\n}\n\n\nclass AudioSEDModel(nn.Module):\n    def __init__(self, encoder, sample_rate, window_size, hop_size, mel_bins, fmin, fmax, classes_num):\n        super().__init__()\n\n        window = 'hann'\n        center = True\n        pad_mode = 'reflect'\n        ref = 1.0\n        amin = 1e-10\n        top_db = None\n        self.interpolate_ratio = 30  # Downsampled ratio\n\n        # Spectrogram extractor\n        self.spectrogram_extractor = Spectrogram(n_fft=window_size, hop_length=hop_size, \n            win_length=window_size, window=window, center=center, pad_mode=pad_mode, \n            freeze_parameters=True)\n\n        # Logmel feature extractor\n        self.logmel_extractor = LogmelFilterBank(sr=sample_rate, n_fft=window_size, \n            n_mels=mel_bins, fmin=fmin, fmax=fmax, ref=ref, amin=amin, top_db=top_db, \n            freeze_parameters=True)\n\n        # Spec augmenter\n        self.spec_augmenter = SpecAugmentation(time_drop_width=64, time_stripes_num=2, \n            freq_drop_width=8, freq_stripes_num=2)\n        \n        # Model Encoder\n        self.encoder = encoder_params[encoder][\"init_op\"]()\n        self.fc1 = nn.Linear(encoder_params[encoder][\"features\"], 1024, bias=True)\n        self.att_block = AttBlock(1024, classes_num, activation=\"sigmoid\")\n        self.bn0 = nn.BatchNorm2d(mel_bins)\n        self.init_weight()\n    \n    def init_weight(self):\n        init_layer(self.fc1)\n        init_bn(self.bn0)\n    \n    def forward(self, input, mixup_lambda=None):\n        \"\"\"Input : (batch_size, data_length)\"\"\"\n\n        x = self.spectrogram_extractor(input)\n        # batch_size x 1 x time_steps x freq_bins\n        x = self.logmel_extractor(x)\n        # batch_size x 1 x time_steps x mel_bins\n\n        frames_num = x.shape[2]\n\n        x = x.transpose(1, 3)\n        x = self.bn0(x)\n        x = x.transpose(1, 3)\n        #print(x.shape)\n\n        if self.training:\n            x = self.spec_augmenter(x)\n        \n        # Mixup on spectrogram\n        if self.training and mixup_lambda is not None:\n            x = do_mixup(x, mixup_lambda)\n        \n        # Output shape (batch size, channels, time, frequency)\n        x = x.expand(x.shape[0], 3, x.shape[2], x.shape[3])\n        #print(x.shape)\n        x = self.encoder.forward_features(x)\n        #print(x.shape)\n        x = torch.mean(x, dim=3)\n        #print(x.shape)\n\n        x1 = F.max_pool1d(x, kernel_size=3, stride=1, padding=1)\n        x2 = F.avg_pool1d(x, kernel_size=3, stride=1, padding=1)\n        x = x1 + x2\n        #print(x.shape)\n\n        x = F.dropout(x, p=0.5, training=self.training)\n        x = x.transpose(1, 2)\n        x = F.relu_(self.fc1(x))\n        x = x.transpose(1, 2)\n        x = F.dropout(x, p=0.5, training=self.training)\n        #print(x.shape)\n\n        (clipwise_output, norm_att, segmentwise_output) = self.att_block(x)\n        logit = torch.sum(norm_att * self.att_block.cla(x), dim=2)\n        segmentwise_output = segmentwise_output.transpose(1, 2)\n\n        # Get framewise output\n        framewise_output = interpolate(segmentwise_output,\n                                       self.interpolate_ratio)\n        framewise_output = pad_framewise_output(framewise_output, frames_num)\n\n        output_dict = {\n            'framewise_output' : framewise_output,\n            'logit' : logit,\n            'clipwise_output' : clipwise_output\n        }\n\n        return output_dict","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Dataset"},{"metadata":{"_kg_hide-input":true,"trusted":true},"cell_type":"code","source":"def crop_or_pad(y, sr, period, record, mode=\"train\"):\n    len_y = len(y)\n    effective_length = sr * period\n    rint = np.random.randint(len(record['t_min']))\n    time_start = record['t_min'][rint] * sr\n    time_end = record['t_max'][rint] * sr\n    if len_y > effective_length:\n        # Positioning sound slice\n        center = np.round((time_start + time_end) / 2)\n        beginning = center - effective_length / 2\n        if beginning < 0:\n            beginning = 0\n        beginning = np.random.randint(beginning, center)\n        ending = beginning + effective_length\n        if ending > len_y:\n            ending = len_y\n        beginning = ending - effective_length\n        y = y[beginning:ending].astype(np.float32)\n    else:\n        y = y.astype(np.float32)\n        beginning = 0\n        ending = effective_length\n\n\n    beginning_time = beginning / sr\n    ending_time = ending / sr\n    label = np.zeros(24, dtype='f')\n\n    for i in range(len(record['t_min'])):\n        if (record['t_min'][i] <= ending_time) and (record['t_max'][i] >= beginning_time):\n            label[record['species_id'][i]] = 1\n    \n    return y, label","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"class SedDataset:\n    def __init__(self, df, period=10, stride=5, audio_transform=None, data_path=\"train\", mode=\"train\"):\n\n        self.period = period\n        self.stride = stride\n        self.audio_transform = audio_transform\n        self.data_path = data_path\n        self.mode = mode\n\n        self.df = df.groupby(\"recording_id\").agg(lambda x: list(x)).reset_index()\n\n    def __len__(self):\n        return len(self.df)\n    \n    def __getitem__(self, idx):\n        record = self.df.iloc[idx]\n\n        y, sr = sf.read(f\"{self.data_path}/{record['recording_id']}.flac\")\n        \n        if self.mode != \"test\":\n            y, label = crop_or_pad(y, sr, period=self.period, record=record, mode=self.mode)\n\n            if self.audio_transform:\n                y = self.audio_transform(samples=y, sample_rate=sr)\n        else:\n            y_ = []\n            i = 0\n            effective_length = self.period * sr\n            stride = self.stride * sr\n            y = np.stack([y[i:i+effective_length].astype(np.float32) for i in range(0, 60*sr+stride-effective_length, stride)])\n            label = np.zeros(24, dtype='f')\n            if self.mode == \"valid\":\n                for i in record['species_id']:\n                    label[i] = 1\n        \n        return {\n            \"image\" : y,\n            \"target\" : label,\n            \"id\" : record['recording_id']\n        }","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Augmentations"},{"metadata":{"_kg_hide-input":true,"trusted":true},"cell_type":"code","source":"import audiomentations as AA\n\ntrain_audio_transform = AA.Compose([\n    AA.AddGaussianNoise(p=0.5),\n    AA.AddGaussianSNR(p=0.5),\n    #AA.AddBackgroundNoise(\"../input/train_audio/\", p=1)\n    #AA.AddImpulseResponse(p=0.1),\n    #AA.AddShortNoises(\"../input/train_audio/\", p=1)\n    #AA.FrequencyMask(min_frequency_band=0.0,  max_frequency_band=0.2, p=0.1),\n    #AA.TimeMask(min_band_part=0.0, max_band_part=0.2, p=0.1),\n    #AA.PitchShift(min_semitones=-0.5, max_semitones=0.5, p=0.1),\n    #AA.Shift(p=0.1),\n    #AA.Normalize(p=0.1),\n    #AA.ClippingDistortion(min_percentile_threshold=0, max_percentile_threshold=1, p=0.05),\n    #AA.PolarityInversion(p=0.05),\n    #AA.Gain(p=0.2)\n])","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Utils"},{"metadata":{"_kg_hide-input":true,"trusted":true},"cell_type":"code","source":"def _lwlrap_sklearn(truth, scores):\n    \"\"\"Reference implementation from https://colab.research.google.com/drive/1AgPdhSp7ttY18O3fEoHOQKlt_3HJDLi8\"\"\"\n    sample_weight = np.sum(truth > 0, axis=1)\n    nonzero_weight_sample_indices = np.flatnonzero(sample_weight > 0)\n    overall_lwlrap = metrics.label_ranking_average_precision_score(\n        truth[nonzero_weight_sample_indices, :] > 0, \n        scores[nonzero_weight_sample_indices, :], \n        sample_weight=sample_weight[nonzero_weight_sample_indices])\n    return overall_lwlrap\n\nclass AverageMeter(object):\n    \"\"\"Computes and stores the average and current value\"\"\"\n\n    def __init__(self):\n        self.reset()\n\n    def reset(self):\n        self.val = 0\n        self.avg = 0\n        self.sum = 0\n        self.count = 0\n\n    def update(self, val, n=1):\n        self.val = val\n        self.sum += val * n\n        self.count += n\n        self.avg = self.sum / self.count\n\nclass MetricMeter(object):\n    def __init__(self):\n        self.reset()\n    \n    def reset(self):\n        self.y_true = []\n        self.y_pred = []\n    \n    def update(self, y_true, y_pred):\n        self.y_true.extend(y_true.cpu().detach().numpy().tolist())\n        self.y_pred.extend(y_pred.cpu().detach().numpy().tolist())\n\n    @property\n    def avg(self):\n        #score_class, weight = lwlrap(np.array(self.y_true), np.array(self.y_pred))\n        self.score = _lwlrap_sklearn(np.array(self.y_true), np.array(self.y_pred)) #(score_class * weight).sum()\n        return {\n            \"lwlrap\" : self.score\n        }\n\ndef seed_everithing(seed):\n    random.seed(seed)\n    os.environ['PYTHONHASHSEED'] = str(seed)\n    np.random.seed(seed)\n    torch.manual_seed(seed)\n    torch.cuda.manual_seed(seed)\n    torch.backends.cudnn.deterministic = True\n","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Losses"},{"metadata":{"trusted":true},"cell_type":"code","source":"from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss\n\nclass PANNsLoss(nn.Module):\n    def __init__(self):\n        super().__init__()\n\n        self.bce = nn.BCELoss()\n\n    def forward(self, input, target):\n        input_ = input[\"clipwise_output\"]\n        input_ = torch.where(torch.isnan(input_),\n                             torch.zeros_like(input_),\n                             input_)\n        input_ = torch.where(torch.isinf(input_),\n                             torch.zeros_like(input_),\n                             input_)\n\n        target = target.float()\n\n        return self.bce(input_, target)","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Functions"},{"metadata":{"_kg_hide-input":true,"trusted":true},"cell_type":"code","source":"def train_epoch(args, model, loader, criterion, optimizer, scheduler, epoch):\n    losses = AverageMeter()\n    scores = MetricMeter()\n\n    model.train()\n    t = tqdm(loader)\n    for i, sample in enumerate(t):\n        optimizer.zero_grad()\n        input = sample['image'].to(args.device)\n        target = sample['target'].to(args.device)\n        output = model(input)\n        loss = criterion(output, target)\n        loss.backward()\n        optimizer.step()\n        if scheduler and args.step_scheduler:\n            scheduler.step()\n\n        bs = input.size(0)\n        scores.update(target, torch.sigmoid(torch.max(output['framewise_output'], dim=1)[0]))\n        losses.update(loss.item(), bs)\n\n        t.set_description(f\"Train E:{epoch} - Loss{losses.avg:0.4f}\")\n    t.close()\n    return scores.avg, losses.avg\n        \ndef valid_epoch(args, model, loader, criterion, epoch):\n    losses = AverageMeter()\n    scores = MetricMeter()\n    model.eval()\n    with torch.no_grad():\n        t = tqdm(loader)\n        for i, sample in enumerate(t):\n            input = sample['image'].to(args.device)\n            target = sample['target'].to(args.device)\n            output = model(input)\n            loss = criterion(output, target)\n\n            bs = input.size(0)\n            scores.update(target, torch.sigmoid(torch.max(output['framewise_output'], dim=1)[0]))\n            losses.update(loss.item(), bs)\n            t.set_description(f\"Valid E:{epoch} - Loss:{losses.avg:0.4f}\")\n    t.close()\n    return scores.avg, losses.avg\n\ndef test_epoch(args, model, loader):\n    model.eval()\n    pred_list = []\n    id_list = []\n    with torch.no_grad():\n        t = tqdm(loader)\n        for i, sample in enumerate(t):\n            input = sample[\"image\"].to(args.device)\n            bs, seq, w = input.shape\n            input = input.reshape(bs*seq, w)\n            id = sample[\"id\"]\n            output = model(input)\n            output = torch.sigmoid(torch.max(output['framewise_output'], dim=1)[0])\n            output = output.reshape(bs, seq, -1)\n            output = torch.sum(output, dim=1)\n            #output, _ = torch.max(output, dim=1)\n            output = output.cpu().detach().numpy().tolist()\n            pred_list.extend(output)\n            id_list.extend(id)\n    \n    return pred_list, id_list","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Main Function"},{"metadata":{"trusted":true},"cell_type":"code","source":"def main(fold):\n    seed_everithing(args.seed)\n\n    args.fold = fold\n    args.save_path = os.path.join(args.output_dir, args.exp_name)\n    os.makedirs(args.save_path, exist_ok=True)\n\n    train_df = pd.read_csv(args.train_csv)\n    sub_df = pd.read_csv(args.sub_csv)\n    if args.DEBUG:\n        train_df = train_df.sample(200)\n    train_fold = train_df[train_df.kfold != fold]\n    valid_fold = train_df[train_df.kfold == fold]\n\n    train_dataset = SedDataset(\n        df = train_fold,\n        period=args.period,\n        audio_transform=train_audio_transform,\n        data_path=args.train_data_path,\n        mode=\"train\"\n    )\n\n    valid_dataset = SedDataset(\n        df = valid_fold,\n        period=args.period,\n        stride=5,\n        audio_transform=None,\n        data_path=args.train_data_path,\n        mode=\"valid\"\n    )\n\n    test_dataset = SedDataset(\n        df = sub_df,\n        period=args.period,\n        stride=5,\n        audio_transform=None,\n        data_path=args.test_data_path,\n        mode=\"test\"\n    )\n\n    train_loader = torch.utils.data.DataLoader(\n        train_dataset,\n        batch_size=args.batch_size,\n        shuffle=True,\n        drop_last=True,\n        num_workers=args.num_workers\n    )\n\n    valid_loader = torch.utils.data.DataLoader(\n        valid_dataset,\n        batch_size=args.batch_size,\n        shuffle=False,\n        drop_last=False,\n        num_workers=args.num_workers\n    )\n\n    test_loader = torch.utils.data.DataLoader(\n        test_dataset,\n        batch_size=args.batch_size,\n        shuffle=False,\n        drop_last=False,\n        num_workers=args.num_workers\n    )\n\n    model = AudioSEDModel(**args.model_param)\n    model = model.to(args.device)\n\n    if args.pretrain_weights:\n        print(\"---------------------loading pretrain weights\")\n        model.load_state_dict(torch.load(args.pretrain_weights, map_location=args.device), strict=False)\n        model = model.to(args.device)\n\n    criterion = PANNsLoss() #BCEWithLogitsLoss() #MaskedBCEWithLogitsLoss() #BCEWithLogitsLoss()\n    optimizer = torch.optim.Adam(model.parameters(), lr=args.lr)\n    num_train_steps = int(len(train_loader) * args.epochs)\n    num_warmup_steps = int(0.1 * args.epochs * len(train_loader))\n    scheduler = get_linear_schedule_with_warmup(optimizer, num_warmup_steps=num_warmup_steps, num_training_steps=num_train_steps)\n\n    best_lwlrap = -np.inf\n    early_stop_count = 0\n\n    for epoch in range(args.start_epcoh, args.epochs):\n        train_avg, train_loss = train_epoch(args, model, train_loader, criterion, optimizer, scheduler, epoch)\n        valid_avg, valid_loss = valid_epoch(args, model, valid_loader, criterion, epoch)\n\n        if args.epoch_scheduler:\n            scheduler.step()\n        \n        content = f\"\"\"\n                {time.ctime()} \\n\n                Fold:{args.fold}, Epoch:{epoch}, lr:{optimizer.param_groups[0]['lr']:.7}\\n\n                Train Loss:{train_loss:0.4f} - LWLRAP:{train_avg['lwlrap']:0.4f}\\n\n                Valid Loss:{valid_loss:0.4f} - LWLRAP:{valid_avg['lwlrap']:0.4f}\\n\n        \"\"\"\n        print(content)\n        with open(f'{args.save_path}/log_{args.exp_name}.txt', 'a') as appender:\n            appender.write(content+'\\n')\n        \n        if valid_avg['lwlrap'] > best_lwlrap:\n            print(f\"########## >>>>>>>> Model Improved From {best_lwlrap} ----> {valid_avg['lwlrap']}\")\n            torch.save(model.state_dict(), os.path.join(args.save_path, f'fold-{args.fold}.bin'))\n            best_lwlrap = valid_avg['lwlrap']\n            early_stop_count = 0\n        else:\n            early_stop_count += 1\n        #torch.save(model.state_dict(), os.path.join(args.save_path, f'fold-{args.fold}_last.bin'))\n\n        if args.early_stop == early_stop_count:\n            print(\"\\n $$$ ---? Ohoo.... we reached early stoping count :\", early_stop_count)\n            break\n    \n    model.load_state_dict(torch.load(os.path.join(args.save_path, f'fold-{args.fold}.bin'), map_location=args.device))\n    model = model.to(args.device)\n\n    target_cols = sub_df.columns[1:].values.tolist()\n    test_pred, ids = test_epoch(args, model, test_loader)\n    print(np.array(test_pred).shape)\n\n    test_pred_df = pd.DataFrame({\n        \"recording_id\" : sub_df.recording_id.values\n    })\n    test_pred_df[target_cols] = test_pred\n    test_pred_df.to_csv(os.path.join(args.save_path, f\"fold-{args.fold}-submission.csv\"), index=False)\n    print(os.path.join(args.save_path, f\"fold-{args.fold}-submission.csv\"))\n    ","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### Config"},{"metadata":{"trusted":true},"cell_type":"code","source":"class args:\n    DEBUG = False\n\n    exp_name = \"SED_E0_5F_BASE\"\n    pretrain_weights = None \n    model_param = {\n        'encoder' : 'tf_efficientnet_b0_ns',\n        'sample_rate': 48000,\n        'window_size' : 512, #* 2, # 512 * 2\n        'hop_size' : 512, #345 * 2, # 320\n        'mel_bins' : 128, # 60\n        'fmin' : 0,\n        'fmax' : 48000 // 2,\n        'classes_num' : 24\n    }\n    period = 10\n    seed = 42\n    start_epcoh = 0 \n    epochs = 50\n    lr = 1e-3\n    batch_size = 16\n    num_workers = 4\n    early_stop = 15\n    step_scheduler = True\n    epoch_scheduler = False\n\n    device = ('cuda' if torch.cuda.is_available() else 'cpu')\n    train_csv = \"train_folds.csv\"\n    test_csv = \"test_df.csv\"\n    sub_csv = \"../input/rfcx-species-audio-detection/sample_submission.csv\"\n    output_dir = \"weights\"\n    train_data_path = \"../input/rfcx-species-audio-detection/train\"\n    test_data_path = \"../input/rfcx-species-audio-detection/test\"\n","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"### train folds"},{"metadata":{"trusted":true},"cell_type":"code","source":"main(fold=0)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"","execution_count":null,"outputs":[]}],"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}