{
  "total": 122,
  "topics": [
    {
      "id": 456606,
      "title": "Recap of Competition - Congratulations to the Winners!",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 456379,
      "title": "Please fill our survey!",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 444828,
      "title": "Official starter notebook",
      "comment_count": 0,
      "views": 0,
      "votes": 15
    },
    {
      "id": 437673,
      "title": "What is a Tensor Processing Unit (TPU)?",
      "comment_count": 10,
      "views": 0,
      "votes": 24
    },
    {
      "id": 722428,
      "title": "Pushing MLOps Boundaries: Hybrid GNN + HistGradientBoosting Pipeline",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 718630,
      "title": "How I solved the OOM and Kernel Crashes using a Hybrid GNN + Tree Architecture",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 718390,
      "title": "[MLOps Pipeline] Hybrid GNN + HistGradientBoosting | Memory Optimized",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 583679,
      "title": "My Whitepaper: GRU-Based Runtime Prediction Pipeline using TPUGraphs Dataset",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 456579,
      "title": "33rd Solution Writeup and Discussion on GST",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 456343,
      "title": "1st Place Solution for the Google - Fast or Slow? Predict AI Model Runtime Competition",
      "comment_count": 27,
      "views": 0,
      "votes": 68
    },
    {
      "id": 456093,
      "title": "5th Place Solution: GNN with Invariant Dimension Features",
      "comment_count": 15,
      "views": 0,
      "votes": 23
    },
    {
      "id": 458511,
      "title": "58th place basic vanilla keras solution",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 458552,
      "title": "201st Place Solution for the Google - Fast or Slow? Predict AI Model Runtime",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 456462,
      "title": "4th simple mlp solution",
      "comment_count": 6,
      "views": 0,
      "votes": 25
    },
    {
      "id": 458370,
      "title": "13th Place Solution for the Google - Fast or Slow? Predict AI Model Runtime Competition",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 458329,
      "title": "84th Solution: Based on Transformer, the GraphTPU model",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 456377,
      "title": "3rd place solution write-up",
      "comment_count": 2,
      "views": 0,
      "votes": 16
    },
    {
      "id": 456092,
      "title": "11th place solution: LightGBM",
      "comment_count": 8,
      "views": 0,
      "votes": 27
    },
    {
      "id": 458257,
      "title": "Kaggle Solution Write-up",
      "comment_count": 1,
      "views": 0,
      "votes": -4
    },
    {
      "id": 456365,
      "title": "2nd Place Solution for the Google - Fast or Slow? Predict AI Model Runtime Competition",
      "comment_count": 6,
      "views": 0,
      "votes": 22
    },
    {
      "id": 457873,
      "title": "53th place solution: Fast or Slow with Data Prunning and Pretraining",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 457729,
      "title": "Discoveries through the Contest",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 456645,
      "title": "8th place solution",
      "comment_count": 1,
      "views": 0,
      "votes": 12
    },
    {
      "id": 456923,
      "title": "Congrats to new Kaggle Competitions GM",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 456565,
      "title": "forget about deep GNN .... use LLM to compile your model?",
      "comment_count": 3,
      "views": 0,
      "votes": 9
    },
    {
      "id": 456206,
      "title": "9th Place Solution: GNN with Compressed Graphs Using Dijkstra’s Algorithm",
      "comment_count": 2,
      "views": 0,
      "votes": 21
    },
    {
      "id": 456090,
      "title": "Leakage happening due to test sampling",
      "comment_count": 2,
      "views": 0,
      "votes": 13
    },
    {
      "id": 456625,
      "title": "97th place solution for Google - Fast or Slow? Predict AI Model Runtime",
      "comment_count": 7,
      "views": 0,
      "votes": 2
    },
    {
      "id": 456084,
      "title": "6th solution: Node-level Instance Norm + Residual  SageConv on 5-hop-neighbour Subgraph ",
      "comment_count": 5,
      "views": 0,
      "votes": 28
    },
    {
      "id": 456083,
      "title": "Something Interesting About Test Data.",
      "comment_count": 19,
      "views": 0,
      "votes": 19
    },
    {
      "id": 456129,
      "title": "10th place solution: Fast or Slow with Graph Transformers",
      "comment_count": 1,
      "views": 0,
      "votes": 25
    },
    {
      "id": 456450,
      "title": "17th place solution: GNN + TransformerEncoderLayer w/ BCELoss.",
      "comment_count": 0,
      "views": 0,
      "votes": 8
    },
    {
      "id": 456489,
      "title": "16th place solution",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 456673,
      "title": "7th Place Solution Write-up",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 456074,
      "title": "19th Place Solution",
      "comment_count": 3,
      "views": 0,
      "votes": 14
    },
    {
      "id": 456105,
      "title": "14th Place Solution for the Google - Fast or Slow? Predict AI Model Runtime Competition",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 455915,
      "title": "Meme Thread",
      "comment_count": 0,
      "views": 0,
      "votes": 10
    },
    {
      "id": 456484,
      "title": "What graphs are in Layout-XLA test - see here",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 455672,
      "title": "RMSE > MSE",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 453824,
      "title": "Crikey! Data Loading Videos",
      "comment_count": 4,
      "views": 0,
      "votes": 13
    },
    {
      "id": 453760,
      "title": "usage of data, e.g. train on \"valid+train\" or pretrain on \"test\"?",
      "comment_count": 3,
      "views": 0,
      "votes": 10
    },
    {
      "id": 453715,
      "title": "Do operations run in parallel?",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 454021,
      "title": "Multiple tensor input configs?",
      "comment_count": 1,
      "views": 0,
      "votes": 5
    },
    {
      "id": 454177,
      "title": "Larger bounds = faster?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 451958,
      "title": "CV / LB Thread",
      "comment_count": 7,
      "views": 0,
      "votes": 14
    },
    {
      "id": 453985,
      "title": "Evaluation for test dataset same config files",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 452540,
      "title": "K-fold predictions",
      "comment_count": 1,
      "views": 0,
      "votes": 8
    },
    {
      "id": 453134,
      "title": "Test data includes private and public LB data?",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 452840,
      "title": "anyone know how to visualise the graph from npz file?",
      "comment_count": 1,
      "views": 0,
      "votes": 9
    },
    {
      "id": 451732,
      "title": "Predict AI Model Runtime: Research Papers ",
      "comment_count": 1,
      "views": 0,
      "votes": 6
    },
    {
      "id": 448367,
      "title": "Increasing Training Loss with Improving Validation Performanc",
      "comment_count": 41,
      "views": 0,
      "votes": 9
    },
    {
      "id": 443581,
      "title": "[IMPORTANT] Data update – retrain your models!",
      "comment_count": 17,
      "views": 0,
      "votes": 33
    },
    {
      "id": 451798,
      "title": "Is this a supposed leak? Some models in the test data seem to be able to identify",
      "comment_count": 3,
      "views": 0,
      "votes": 2
    },
    {
      "id": 451472,
      "title": "CV and LB are not consistent",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 449137,
      "title": "Wrong data in layout/xla/random",
      "comment_count": 9,
      "views": 0,
      "votes": 11
    },
    {
      "id": 447859,
      "title": "Is it important to clean the data?",
      "comment_count": 3,
      "views": 0,
      "votes": 6
    },
    {
      "id": 450988,
      "title": "What types of features are most informative?",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 448402,
      "title": "Help! Adding new features",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 449161,
      "title": "Whether raw protobuf contains NodeLayoutConfig",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 436741,
      "title": "[place holder] my experimental results",
      "comment_count": 14,
      "views": 0,
      "votes": 20
    },
    {
      "id": 443265,
      "title": "Understanding the \"layout propagation\" technique",
      "comment_count": 11,
      "views": 0,
      "votes": 8
    },
    {
      "id": 445225,
      "title": "Extracting additional features from protobuf",
      "comment_count": 4,
      "views": 0,
      "votes": 14
    },
    {
      "id": 436629,
      "title": "some youtube video on the competition you should not missed!",
      "comment_count": 7,
      "views": 0,
      "votes": 45
    },
    {
      "id": 447031,
      "title": "(Rookie in this competition)Why did the submission take so long to run?",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 443415,
      "title": "What is your kendal tau for layour:xla",
      "comment_count": 8,
      "views": 0,
      "votes": 7
    },
    {
      "id": 446795,
      "title": "GPU RAM issues for inference in Kaggle kernel",
      "comment_count": 3,
      "views": 0,
      "votes": 3
    },
    {
      "id": 446460,
      "title": "layout datasets have configs with the same node_config_feat but different config_runtime",
      "comment_count": 2,
      "views": 0,
      "votes": 5
    },
    {
      "id": 445869,
      "title": "Question on how to find the output file",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 445573,
      "title": "Building GNN from scratch",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 435631,
      "title": "Understanding Layout and Tile Configurations in AI Compilers",
      "comment_count": 16,
      "views": 0,
      "votes": 101
    },
    {
      "id": 433784,
      "title": "New to Machine Learning or Kaggle?",
      "comment_count": 12,
      "views": 0,
      "votes": 13
    },
    {
      "id": 438765,
      "title": "Interest to join my team",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 438074,
      "title": "Understanding the layout config features of a convolutional node",
      "comment_count": 11,
      "views": 0,
      "votes": 17
    },
    {
      "id": 443386,
      "title": "Understanding dot node layout configurations",
      "comment_count": 12,
      "views": 0,
      "votes": 4
    },
    {
      "id": 444206,
      "title": "Range of OPA values",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 443731,
      "title": "What does `Baseline on Updated Data` on Leaderboard mean?",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 436977,
      "title": "Anticipate a decent shakeup",
      "comment_count": 19,
      "views": 0,
      "votes": 18
    },
    {
      "id": 436680,
      "title": "Explanations of Data Folder Structure",
      "comment_count": 2,
      "views": 0,
      "votes": 12
    },
    {
      "id": 442258,
      "title": "Understanding the Tile Configuration",
      "comment_count": 7,
      "views": 0,
      "votes": 4
    },
    {
      "id": 442643,
      "title": "Understanding 'is_root' feature in node_feats of tiles",
      "comment_count": 5,
      "views": 0,
      "votes": 2
    },
    {
      "id": 438498,
      "title": "How to deal with layout nlp data?",
      "comment_count": 11,
      "views": 0,
      "votes": 10
    },
    {
      "id": 439002,
      "title": "Building a model to learn layout data",
      "comment_count": 10,
      "views": 0,
      "votes": 7
    },
    {
      "id": 441930,
      "title": "Question about training time of baseline models",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 442910,
      "title": "Absence of Conda Environment in TPU Kernels",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 441445,
      "title": "Understanding the Layout Configuration",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 438819,
      "title": "Understanding the data collection process",
      "comment_count": 4,
      "views": 0,
      "votes": 3
    },
    {
      "id": 435640,
      "title": "Can we extract additional features from the graph?",
      "comment_count": 16,
      "views": 0,
      "votes": 14
    },
    {
      "id": 441799,
      "title": "Data processing is slow, any tips?",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 440869,
      "title": "Why is it so quiet here in this competition...",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 440878,
      "title": "New to Kaggle, why are we not able to get same score from the notebook we copied and working on",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 437530,
      "title": "Share competition related resources",
      "comment_count": 4,
      "views": 0,
      "votes": 5
    },
    {
      "id": 436786,
      "title": "Question about `edge_index`",
      "comment_count": 7,
      "views": 0,
      "votes": 5
    },
    {
      "id": 436224,
      "title": "Power BI EDA - [01.09.2023, Topology Analysis & Runtime analysis]",
      "comment_count": 7,
      "views": 0,
      "votes": 9
    },
    {
      "id": 435577,
      "title": "Tensor Tile, Tensor Shard, Simulated Annealing and Langevin Dynamics. ",
      "comment_count": 15,
      "views": 0,
      "votes": 51
    },
    {
      "id": 437867,
      "title": "Installing torch geometric on GPU command",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 437418,
      "title": "New to GNNs this stuff may help u for better understand ",
      "comment_count": 2,
      "views": 0,
      "votes": 3
    },
    {
      "id": 437068,
      "title": "What Do Isolated Nodes Represent in a Computational Graph ?",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 436985,
      "title": "i am new and i have ni idea of what i am doing right now. i need guidlines. thank you ",
      "comment_count": 2,
      "views": 0,
      "votes": 4
    },
    {
      "id": 437313,
      "title": "how to deal with  layout_nlp_default dataset",
      "comment_count": 2,
      "views": 0,
      "votes": 2
    },
    {
      "id": 437125,
      "title": "torch-geometric torch-scatter compatibility issue",
      "comment_count": 6,
      "views": 0,
      "votes": 1
    },
    {
      "id": 437039,
      "title": "🚀 Embrace Learning with 📓 Public Notebooks! 📈",
      "comment_count": 1,
      "views": 0,
      "votes": 3
    },
    {
      "id": 437223,
      "title": "This Competition has an Official Discord Channel",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 437384,
      "title": "Running the Model in Google Colab",
      "comment_count": 5,
      "views": 0,
      "votes": 1
    },
    {
      "id": 436710,
      "title": "npz taking a long time to load?",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 435785,
      "title": "Files converted to parquet and csv",
      "comment_count": 6,
      "views": 0,
      "votes": 19
    },
    {
      "id": 436435,
      "title": "An introduction to graph theory - edition 2.09.2023",
      "comment_count": 2,
      "views": 0,
      "votes": 6
    },
    {
      "id": 436214,
      "title": "Layout Runtimes And Configurations",
      "comment_count": 6,
      "views": 0,
      "votes": 7
    },
    {
      "id": 435744,
      "title": "Understanding Competition Evaluation Metrics ",
      "comment_count": 4,
      "views": 0,
      "votes": 11
    },
    {
      "id": 437067,
      "title": "Looking for buddies?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 436569,
      "title": "Looking for Teammate?",
      "comment_count": 1,
      "views": 0,
      "votes": 1
    },
    {
      "id": 436962,
      "title": "Team search",
      "comment_count": 0,
      "views": 0,
      "votes": -1
    },
    {
      "id": 435700,
      "title": "When we use npz file instend of csv? What is the benefit of using npz file?",
      "comment_count": 5,
      "views": 0,
      "votes": 3
    },
    {
      "id": 436829,
      "title": "CSV format tile-xla : Train, Validation and Test dataset | Data Understanding | EDA | Baseline Model Notebook",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 436204,
      "title": "Configuration is at the graph-level",
      "comment_count": 4,
      "views": 0,
      "votes": 4
    },
    {
      "id": 436294,
      "title": "Understanding the problem",
      "comment_count": 1,
      "views": 0,
      "votes": 2
    },
    {
      "id": 435981,
      "title": "Related Paper : Runtime Performance Prediction for Deep Learning Models with Graph Neural Network.",
      "comment_count": 3,
      "views": 0,
      "votes": 5
    },
    {
      "id": 436230,
      "title": "Looking for teammate",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 436358,
      "title": "What do two example figures in \"Description\" mean?",
      "comment_count": 3,
      "views": 0,
      "votes": 1
    },
    {
      "id": 436123,
      "title": "More robust validation needed?",
      "comment_count": 2,
      "views": 0,
      "votes": 1
    },
    {
      "id": 435611,
      "title": " Finding the Kaggle Competition Teammate.",
      "comment_count": 5,
      "views": 0,
      "votes": -4
    },
    {
      "id": 435548,
      "title": "only the first 5 entries will be considered?",
      "comment_count": 6,
      "views": 0,
      "votes": 3
    },
    {
      "id": 435567,
      "title": "Looking for a team",
      "comment_count": 7,
      "views": 0,
      "votes": -6
    }
  ],
  "errors": []
}