{
  "total": 186,
  "topics": [
    {
      "id": 43516,
      "title": "Hi from the competition organizer!",
      "comment_count": 53,
      "views": 0,
      "votes": 51
    },
    {
      "id": 47703,
      "title": "Special TensorFlow Prize",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 47631,
      "title": "Issues with Final Leader Board",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47246,
      "title": "IMPORTANT: Option to Choose 3 Final Submissions",
      "comment_count": 10,
      "views": 0,
      "votes": 10
    },
    {
      "id": 44449,
      "title": "GCP Credits Clarification",
      "comment_count": 60,
      "views": 0,
      "votes": 21
    },
    {
      "id": 43511,
      "title": "Welcome!",
      "comment_count": 21,
      "views": 0,
      "votes": 21
    },
    {
      "id": 43520,
      "title": "Hi from resident Speechgler!",
      "comment_count": 10,
      "views": 0,
      "votes": 18
    },
    {
      "id": 383668,
      "title": "Indian English Accent to American English Accent, Speech-to-Speech in realtime..",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 45682,
      "title": "Unknown unknowns",
      "comment_count": 13,
      "views": 0,
      "votes": 17
    },
    {
      "id": 47722,
      "title": "3rd place solution",
      "comment_count": 18,
      "views": 0,
      "votes": 52
    },
    {
      "id": 236644,
      "title": "Federated Learning",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 208779,
      "title": "Data is corrupted",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 185919,
      "title": "Why is there a testing_list.txt in train.7z ?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 167289,
      "title": "Where to start with Speech Recognition?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 163307,
      "title": "can any one help me out by guiding me on how to start with speech emotion detection.",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 125570,
      "title": "Can not read tensorflow Speech command data",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 113677,
      "title": "Test Data",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44091,
      "title": "List of references for newbie like me",
      "comment_count": 16,
      "views": 0,
      "votes": 74
    },
    {
      "id": 47746,
      "title": "ConvLSTM",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 44283,
      "title": "Anyone using 1-d convolutions?",
      "comment_count": 69,
      "views": 0,
      "votes": 38
    },
    {
      "id": 47540,
      "title": "Have anyone tried the capsule network?",
      "comment_count": 8,
      "views": 0,
      "votes": 0
    },
    {
      "id": 90576,
      "title": "Cleaning the data...",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 47827,
      "title": "Late experiment with Baidu's Deepspeech",
      "comment_count": 3,
      "views": 0,
      "votes": 7
    },
    {
      "id": 45069,
      "title": "Read wav files?",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 43576,
      "title": "Pre-trained models for speech recognition",
      "comment_count": 12,
      "views": 0,
      "votes": 47
    },
    {
      "id": 43556,
      "title": "Raspberry pi emulator",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 43678,
      "title": "Very slow download",
      "comment_count": 6,
      "views": 0,
      "votes": 4
    },
    {
      "id": 64373,
      "title": "Neural attention for speech recognition",
      "comment_count": 8,
      "views": 0,
      "votes": 7
    },
    {
      "id": 45037,
      "title": "Hello Edge: Keyword Spotting on Microcontrollers",
      "comment_count": 26,
      "views": 0,
      "votes": 21
    },
    {
      "id": 47674,
      "title": "Our approach [4th place]",
      "comment_count": 6,
      "views": 0,
      "votes": 48
    },
    {
      "id": 44388,
      "title": "Machine Learning Wechat group",
      "comment_count": 7,
      "views": 0,
      "votes": 3
    },
    {
      "id": 46945,
      "title": "My Tricks and Solution",
      "comment_count": 36,
      "views": 0,
      "votes": 92
    },
    {
      "id": 56074,
      "title": "An LSTM CTC solution",
      "comment_count": 5,
      "views": 0,
      "votes": 6
    },
    {
      "id": 54286,
      "title": "Test dataset released",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 48546,
      "title": "VGG19-BN scoring 0.89839, others >0.90",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 47728,
      "title": "what i have learned and moving forward",
      "comment_count": 15,
      "views": 0,
      "votes": 54
    },
    {
      "id": 51303,
      "title": "Knowledge Distillation in Keras",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 47715,
      "title": "2nd Place Solution",
      "comment_count": 11,
      "views": 0,
      "votes": 85
    },
    {
      "id": 47730,
      "title": "Newbie solution. Best private score 0.90908.",
      "comment_count": 17,
      "views": 0,
      "votes": 40
    },
    {
      "id": 49712,
      "title": "Any good solutions to the special TensorFlow competition?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 47628,
      "title": "Summary of 25 Solution",
      "comment_count": 1,
      "views": 0,
      "votes": 16
    },
    {
      "id": 49427,
      "title": "Classifying audio using One Dimensional Convolutional layers",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 49046,
      "title": "Kaggle experience--From Baseline to Top 5 percent",
      "comment_count": 1,
      "views": 0,
      "votes": -3
    },
    {
      "id": 46982,
      "title": "Let's help the beginners : LB=0.82 cnn_trad_pool2_net solution",
      "comment_count": 74,
      "views": 0,
      "votes": 52
    },
    {
      "id": 48470,
      "title": "improvements to resnet got 0.8933: single model, no data argument, no ensemble",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47947,
      "title": "Why the best score < 0.92?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47646,
      "title": "Cost Benchmark: How much did you spend for GPUs for this competition?",
      "comment_count": 6,
      "views": 0,
      "votes": 7
    },
    {
      "id": 47616,
      "title": "Congratulations!",
      "comment_count": 24,
      "views": 0,
      "votes": 23
    },
    {
      "id": 47889,
      "title": "A feeling of incompleteness",
      "comment_count": 1,
      "views": 0,
      "votes": 6
    },
    {
      "id": 47618,
      "title": "LB Score 0.90637 Approach",
      "comment_count": 6,
      "views": 0,
      "votes": 26
    },
    {
      "id": 48046,
      "title": "Anyone else use entropy of predictions to detect silence?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 47099,
      "title": "Issue with Benchmark Script",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47660,
      "title": "single models with high scores",
      "comment_count": 14,
      "views": 0,
      "votes": 14
    },
    {
      "id": 47692,
      "title": "Mis-labeled data?",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 47650,
      "title": "CTC loss ",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 47841,
      "title": "FWIW, some code, 1D models, TF vs PyTorch, and Triplet Loss",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 47614,
      "title": "Private sharing and late merging",
      "comment_count": 14,
      "views": 0,
      "votes": 8
    },
    {
      "id": 46988,
      "title": "solution for LB=0.86 models (1d and 2d)",
      "comment_count": 30,
      "views": 0,
      "votes": 43
    },
    {
      "id": 47645,
      "title": "LB0.88171 -> Why 9 CNN-based network MEL+MFCC fully-connected ensembling didnt work?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47493,
      "title": "Why do some people share  solutions before deadline?",
      "comment_count": 21,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47687,
      "title": "Top 5% Solution Source Code ",
      "comment_count": 0,
      "views": 0,
      "votes": 12
    },
    {
      "id": 47825,
      "title": "Did anybody try siamese networks to detect unknows?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47632,
      "title": "Just one word difference between 1st/2nd/3rd place?",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 47633,
      "title": "VGG16 neural network + openSMILE VAD (LB: 0.87)",
      "comment_count": 1,
      "views": 0,
      "votes": 7
    },
    {
      "id": 47551,
      "title": "LB shakeup predictions!",
      "comment_count": 12,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47790,
      "title": "In search of best architecture: crelu, dilation, etc. ",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47574,
      "title": "Following Heng's example",
      "comment_count": 10,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47566,
      "title": "spotifyNet",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 47194,
      "title": "moving from 0.90 and beyond - competition tricks or black magic ",
      "comment_count": 21,
      "views": 0,
      "votes": 28
    },
    {
      "id": 47507,
      "title": "How to choose weights for ensembling models",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47700,
      "title": "Did anyone use Kaldi?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47625,
      "title": "Why more than half of the team disappeared?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47653,
      "title": "wavenet or gan to generate new unknowns",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 46981,
      "title": "Ad: Acoustic Scene Classification Competition",
      "comment_count": 1,
      "views": 0,
      "votes": 10
    },
    {
      "id": 47079,
      "title": "Ensembling 0.86 models to 0.89",
      "comment_count": 7,
      "views": 0,
      "votes": 36
    },
    {
      "id": 47478,
      "title": "Deep Residual Learning Benchmark (95.8% on speech commands dataset)",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 47394,
      "title": "What if winner is an AI-Bot and beats Human Kagglers",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47252,
      "title": "GCloud NVidia error?",
      "comment_count": 13,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47377,
      "title": "Will true LB labels be given post competition?",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47312,
      "title": "how is flawed pronunciation samples scored?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47388,
      "title": "pitch_shift data augmentation",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47162,
      "title": "An error about team merge. @competition organizer",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 47083,
      "title": "silence/non-silence classifier and 10-class classifier",
      "comment_count": 7,
      "views": 0,
      "votes": 6
    },
    {
      "id": 47085,
      "title": "Attention Models",
      "comment_count": 14,
      "views": 0,
      "votes": 3
    },
    {
      "id": 47351,
      "title": "All submissions scoring 0?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 47408,
      "title": "A though experiment ...",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 47395,
      "title": "Attempting to find 'outliers' with an autoencoder",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 47174,
      "title": "Question about the Tensorflow Special Prize for Competition Organizers",
      "comment_count": 7,
      "views": 0,
      "votes": 4
    },
    {
      "id": 47026,
      "title": "what should be labeled to sounds like \"five....no\"",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47221,
      "title": "Does a very quite voice consider to be unknown or silence?",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 47136,
      "title": "How to cope with files with double words?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 47021,
      "title": "Difference of private and public LB test set: beware of shakeup?",
      "comment_count": 1,
      "views": 0,
      "votes": 10
    },
    {
      "id": 47211,
      "title": "Can the organizer publish the public LB test script at end of the competition?",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 47134,
      "title": "Mixup aug simple drop in code ",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 46967,
      "title": "Probe submissions or Super fast GPU?",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 47060,
      "title": "Breaking Speech Recognition Model using Adversarial Examples",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 46659,
      "title": "How do you split the train data?",
      "comment_count": 22,
      "views": 0,
      "votes": 5
    },
    {
      "id": 47025,
      "title": "How to use Jupyter / X11 on GCP?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 46839,
      "title": "Adding noise for data augmentation",
      "comment_count": 8,
      "views": 0,
      "votes": 1
    },
    {
      "id": 46962,
      "title": "Post prediction logic",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 46881,
      "title": "Pretrained classical CV models, like Resnets are not allowed?",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 46770,
      "title": "Model evaluation on Raspberry Pi3",
      "comment_count": 1,
      "views": 0,
      "votes": 4
    },
    {
      "id": 46880,
      "title": "Recipe for running on Google Cloud Platform / ML Engine",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 46298,
      "title": "LB 0.80 while I predict 47% labels as \"unknown\"",
      "comment_count": 6,
      "views": 0,
      "votes": 14
    },
    {
      "id": 46723,
      "title": "How are you generating noise?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 46635,
      "title": "0.84(LB) csv file here",
      "comment_count": 8,
      "views": 0,
      "votes": -2
    },
    {
      "id": 46618,
      "title": "Which submission are used for the final score ? ",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 46046,
      "title": "Up, off",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44897,
      "title": "WavFileWarning: Chunk (non-data) not understood, skipping it.",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 46497,
      "title": "Silence Class Generation in Validation Set",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 46439,
      "title": "Anyone using recurrent neural networks?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 46427,
      "title": "Choosing best 2 submissions for final leaderboard?",
      "comment_count": 5,
      "views": 0,
      "votes": null
    },
    {
      "id": 46456,
      "title": "Anyone experiencing errors trying to run the speech example.",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 45362,
      "title": "Does anyone use tf's 'mfccs_from_log_mel_spectrograms'?",
      "comment_count": 6,
      "views": 0,
      "votes": 4
    },
    {
      "id": 45968,
      "title": "How to Win a Data Science Competition: Learn from Top Kagglers",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 46229,
      "title": "Additional data (83,698 samples)",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 45547,
      "title": "High level Google keyword spotting architecture",
      "comment_count": 3,
      "views": 0,
      "votes": 7
    },
    {
      "id": 46080,
      "title": "A less-than-honest approach to the special task",
      "comment_count": 6,
      "views": 0,
      "votes": null
    },
    {
      "id": 45987,
      "title": "blending noise",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 45824,
      "title": "Reading from a Windows directory (\\ back-slash)",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 45642,
      "title": "A  0.09  submission , which part is wrong ?",
      "comment_count": 6,
      "views": 0,
      "votes": 0
    },
    {
      "id": 44250,
      "title": "Mismatch between validation and test accuracy",
      "comment_count": 34,
      "views": 0,
      "votes": 13
    },
    {
      "id": 44754,
      "title": "File Corrupted ",
      "comment_count": 10,
      "views": 0,
      "votes": 6
    },
    {
      "id": 45548,
      "title": "Suggestion regarding features to the neural network",
      "comment_count": 2,
      "views": 0,
      "votes": 12
    },
    {
      "id": 45840,
      "title": "Tensorflow issues",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 45004,
      "title": "Any data normalization or standardization ?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 45768,
      "title": "Can we use external data for unknown？",
      "comment_count": 1,
      "views": 0,
      "votes": 0
    },
    {
      "id": 44039,
      "title": "Data partitioning and local CV score",
      "comment_count": 3,
      "views": 0,
      "votes": 4
    },
    {
      "id": 45580,
      "title": "Password for Test data archive?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 45526,
      "title": "Got the GCP Credit but can't use GPU",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 45653,
      "title": "Some files make cost down to 0",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 45555,
      "title": "no charge",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 45515,
      "title": "Hard to classify 'on' vs 'off'",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 45508,
      "title": "Raspberry Pi 3 kit selection  ",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 45504,
      "title": "Create a set of inputs to frozen TensorFlow GraphDef file for Raspberry Pi 3",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44687,
      "title": "One training file repeated 3709 times in test!",
      "comment_count": 3,
      "views": 0,
      "votes": 21
    },
    {
      "id": 45324,
      "title": "Unknown or silence",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44640,
      "title": "Mozilla Open Sourced a Speech Recognition Model Yesterday",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 44239,
      "title": "Is all synthetic data allowed?",
      "comment_count": 9,
      "views": 0,
      "votes": 21
    },
    {
      "id": 45329,
      "title": " GCP credit application",
      "comment_count": 4,
      "views": 0,
      "votes": null
    },
    {
      "id": 45019,
      "title": "Error: \"Data too short when trying to read value\"",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 45364,
      "title": "Google Speech Recognition Tutorial",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44611,
      "title": "More precise leaderboard score",
      "comment_count": 5,
      "views": 0,
      "votes": 23
    },
    {
      "id": 45333,
      "title": "Unable to download speech_commands. Please help!",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 44885,
      "title": "How to Submit? Help!",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 45313,
      "title": "Load test data into notebook",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 43624,
      "title": "FYI: 87.8% reference implementation here",
      "comment_count": 10,
      "views": 0,
      "votes": 39
    },
    {
      "id": 43713,
      "title": "Sudden accuracy drop when training on all labels",
      "comment_count": 4,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44234,
      "title": "Public/Private Leaderboard Split",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 43599,
      "title": "Unsupervised learning on the test set",
      "comment_count": 10,
      "views": 0,
      "votes": 12
    },
    {
      "id": 44970,
      "title": "Getting a list of files, labels",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 44946,
      "title": "Sound samples shorter than one sec?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44923,
      "title": "Violation of terms of the competition?",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 44852,
      "title": "Can I write code in python?",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 44634,
      "title": "Raspberry Pi Questions",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 44814,
      "title": "Validation accuracy greater than training accuracy",
      "comment_count": 3,
      "views": 0,
      "votes": null
    },
    {
      "id": 44327,
      "title": "Can I use Custom Operation on LasberryPi for Special Prize?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 44724,
      "title": "no team but have some thoughts on approaching this if folks are interested",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 43549,
      "title": "Check MD5 & SHA-1 hashsum",
      "comment_count": 5,
      "views": 0,
      "votes": 4
    },
    {
      "id": 44463,
      "title": "unable to view my kernel changes",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 44458,
      "title": "How to use test data",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 43925,
      "title": "Newbie: Help with Starting Out & Setting up Environment",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44411,
      "title": "Not able to get test dataset in kernal",
      "comment_count": 1,
      "views": 0,
      "votes": null
    },
    {
      "id": 44396,
      "title": "Multiple keywords in a wav clip",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 44218,
      "title": "Silence vs Unknown",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 44017,
      "title": "File train/audio/bird/3e7124ba_nohash_0.wav is silent",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 44288,
      "title": "Template to split data by user",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 44223,
      "title": "Evaluation metric",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 44028,
      "title": "Difference between librosa and scipy for data loading",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 43829,
      "title": "How to get $500 credit?",
      "comment_count": 6,
      "views": 0,
      "votes": 4
    },
    {
      "id": 44109,
      "title": "One extra test file?",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 43733,
      "title": "Raspberry Pi Leaderboard? ",
      "comment_count": 5,
      "views": 0,
      "votes": 9
    },
    {
      "id": 43694,
      "title": "It should be 12 classes?",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 43773,
      "title": "uncertainty about unknown",
      "comment_count": 7,
      "views": 0,
      "votes": 4
    },
    {
      "id": 43770,
      "title": "data error: the archive has been broken",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 43871,
      "title": "For the Raspberry Pi track, what does it mean that the only dependency must be Tensorflow?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 43693,
      "title": "Tutorial on installing TF on the Rpi3",
      "comment_count": 1,
      "views": 0,
      "votes": 15
    },
    {
      "id": 43776,
      "title": "Confusion about Raspberry Pi",
      "comment_count": 2,
      "views": 0,
      "votes": null
    },
    {
      "id": 43696,
      "title": "How was silence data generated in test set?",
      "comment_count": 2,
      "views": 0,
      "votes": 7
    },
    {
      "id": 43667,
      "title": "Whats stopping a team from manually annotating the test set?",
      "comment_count": 5,
      "views": 0,
      "votes": 7
    },
    {
      "id": 43601,
      "title": "Admins please add  'librosa' to the kernels",
      "comment_count": 7,
      "views": 0,
      "votes": 6
    },
    {
      "id": 43685,
      "title": "The big PyTorch Thread",
      "comment_count": 4,
      "views": 0,
      "votes": 2
    },
    {
      "id": 43729,
      "title": "webrtcvad voice activity detection example",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 43632,
      "title": "Other labeled audio sets",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 43575,
      "title": "Tensorflow Eager Wavenet starter kit",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 43514,
      "title": "Sound Format",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    }
  ],
  "errors": []
}