[
  {
    "Bits per dim": 3.43,
    "code_links": [
      {
        "title": "lucidrains/routing-transformer",
        "url": "https://github.com/lucidrains/routing-transformer"
      },
      {
        "title": "lucidrains/local-attention",
        "url": "https://github.com/lucidrains/local-attention"
      }
    ],
    "date": "2020-03-12",
    "date2": 20200312,
    "model": "Routing Transformer",
    "paper": {
      "title": "Efficient Content-Based Sparse Attention with Routing Transformers",
      "url": "https://cknow.io/lib/4ca8cce5b31c4199"
    },
    "paper_data_uoa": "4ca8cce5b31c4199"
  },
  {
    "Bits per dim": 3.44,
    "code_links": [
      {
        "title": "openai/sparse_attention",
        "url": "https://github.com/openai/sparse_attention"
      }
    ],
    "date": "2018-04-23",
    "date2": 20180423,
    "model": "Sparse Transformer 152M (strided)",
    "paper": {
      "title": "Generating Long Sequences with Sparse Transformers",
      "url": "https://cknow.io/lib/a158142bbf282e3a"
    },
    "paper_data_uoa": "a158142bbf282e3a"
  },
  {
    "Bits per dim": 3.57,
    "code_links": [
      {
        "title": "openai/pixel-cnn",
        "url": "https://github.com/openai/pixel-cnn"
      },
      {
        "title": "rampage644/wavenet",
        "url": "https://github.com/rampage644/wavenet"
      },
      {
        "title": "EugenHotaj/pytorch-generative",
        "url": "https://github.com/EugenHotaj/pytorch-generative"
      },
      {
        "title": "sarus-tech/tf2-published-models",
        "url": "https://github.com/sarus-tech/tf2-published-models/tree/master/gated_pixelcnn"
      },
      {
        "title": "anordertoreclaim/PixelCNN",
        "url": "https://github.com/anordertoreclaim/PixelCNN"
      },
      {
        "title": "SJHNJU/gated-pixelcnn",
        "url": "https://github.com/SJHNJU/gated-pixelcnn"
      },
      {
        "title": "SJHNJU/Gated_PixelCNN",
        "url": "https://github.com/SJHNJU/Gated_PixelCNN"
      },
      {
        "title": "vocong25/gated_pixelcnn",
        "url": "https://github.com/vocong25/gated_pixelcnn"
      },
      {
        "title": "chandlersupple/Tensorflow-PixelCNN",
        "url": "https://github.com/chandlersupple/Tensorflow-PixelCNN"
      }
    ],
    "date": "2016-06-16",
    "date2": 20160616,
    "model": "Gated PixelCNN (van den Oord et al., [2016c])",
    "paper": {
      "title": "Conditional Image Generation with PixelCNN Decoders",
      "url": "https://cknow.io/lib/dbcdb26607457982"
    },
    "paper_data_uoa": "dbcdb26607457982"
  },
  {
    "Bits per dim": 3.757,
    "code_links": [
      {
        "title": "rtqichen/residual-flows",
        "url": "https://github.com/rtqichen/residual-flows"
      }
    ],
    "date": "2019-06-06",
    "date2": 20190606,
    "model": "Residual Flow",
    "paper": {
      "title": "Residual Flows for Invertible Generative Modeling",
      "url": "https://cknow.io/lib/09cd07ba48f3953a"
    },
    "paper_data_uoa": "09cd07ba48f3953a"
  },
  {
    "Bits per dim": 3.81,
    "code_links": [
      {
        "title": "openai/glow",
        "url": "https://github.com/openai/glow"
      },
      {
        "title": "ikostrikov/pytorch-flows",
        "url": "https://github.com/ikostrikov/pytorch-flows"
      },
      {
        "title": "rosinality/glow-pytorch",
        "url": "https://github.com/rosinality/glow-pytorch"
      },
      {
        "title": "y0ast/Glow-PyTorch",
        "url": "https://github.com/y0ast/Glow-PyTorch"
      },
      {
        "title": "ex4sperans/variational-inference-with-normalizing-flows",
        "url": "https://github.com/ex4sperans/variational-inference-with-normalizing-flows"
      },
      {
        "title": "musyoku/generative-flow",
        "url": "https://github.com/musyoku/generative-flow"
      },
      {
        "title": "musyoku/chainer-glow",
        "url": "https://github.com/musyoku/chainer-glow"
      },
      {
        "title": "chrischute/glow",
        "url": "https://github.com/chrischute/glow"
      },
      {
        "title": "eyalbetzalel/GLOW",
        "url": "https://github.com/eyalbetzalel/GLOW"
      }
    ],
    "date": "2018-07-09",
    "date2": 20180709,
    "model": "Glow (Kingma and Dhariwal, 2018)",
    "paper": {
      "title": "Glow: Generative Flow with Invertible 1x1 Convolutions",
      "url": "https://cknow.io/lib/424ca5c82c1a2f75"
    },
    "paper_data_uoa": "424ca5c82c1a2f75"
  }
]