{
  "id": 100073,
  "title": "Two popular repos for PyTorch and Keras GANs",
  "url": "/competitions/generative-dog-images/discussion/100073",
  "author_name": "",
  "post_date": "2019-07-16T13:03:11.323231400Z",
  "votes": 7,
  "comment_count": 5,
  "views": 0,
  "content": "<p>For practice I have found a couple of very good repos with a lot of various GAN architectures implemented in PyTorch and Keras (separately of course).</p>\n\n<p>Very Good for anyone looking for a head start code in GANs.\n<a href=\"https://github.com/eriklindernoren/PyTorch-GAN\">https://github.com/eriklindernoren/PyTorch-GAN</a> - 3500+ stars\n<a href=\"https://github.com/eriklindernoren/Keras-GAN\">https://github.com/eriklindernoren/Keras-GAN</a> - 5000+ stars</p>\n\n<p>Another one pointed by Andrew earlier in this competition.\n<a href=\"https://www.kaggle.com/c/generative-dog-images/discussion/97772#latest-565566\">https://www.kaggle.com/c/generative-dog-images/discussion/97772#latest-565566</a></p>",
  "messages": [
    {
      "id": "577185",
      "postDate": "07/16/2019 13:03:11",
      "content": "<p>For practice I have found a couple of very good repos with a lot of various GAN architectures implemented in PyTorch and Keras (separately of course).</p>\n\n<p>Very Good for anyone looking for a head start code in GANs.\n<a href=\"https://github.com/eriklindernoren/PyTorch-GAN\">https://github.com/eriklindernoren/PyTorch-GAN</a> - 3500+ stars\n<a href=\"https://github.com/eriklindernoren/Keras-GAN\">https://github.com/eriklindernoren/Keras-GAN</a> - 5000+ stars</p>\n\n<p>Another one pointed by Andrew earlier in this competition.\n<a href=\"https://www.kaggle.com/c/generative-dog-images/discussion/97772#latest-565566\">https://www.kaggle.com/c/generative-dog-images/discussion/97772#latest-565566</a></p>",
      "rawMarkdown": "For practice I have found a couple of very good repos with a lot of various GAN architectures implemented in PyTorch and Keras (separately of course).\n\nVery Good for anyone looking for a head start code in GANs.\nhttps://github.com/eriklindernoren/PyTorch-GAN - 3500+ stars\nhttps://github.com/eriklindernoren/Keras-GAN - 5000+ stars\n\nAnother one pointed by Andrew earlier in this competition.\nhttps://www.kaggle.com/c/generative-dog-images/discussion/97772#latest-565566",
      "votes": null
    },
    {
      "id": "577229",
      "postDate": "07/16/2019 14:11:44",
      "content": "<p>They are very good material to start . Referred to them a lot while trying out a lot of different GAN . But the best \nsolution for this dataset , still remain elusive to me . \nCommon DCGAN I reached till 72 , with RaLSGAN and RaHingeGAN I probably achieved the similar in 1/3 of the epochs \nbut then they saturate :) \nStill searching for that alchemists stone . ....</p>",
      "rawMarkdown": "They are very good material to start . Referred to them a lot while trying out a lot of different GAN . But the best \nsolution for this dataset , still remain elusive to me . \nCommon DCGAN I reached till 72 , with RaLSGAN and RaHingeGAN I probably achieved the similar in 1/3 of the epochs \nbut then they saturate :) \nStill searching for that alchemists stone . ....",
      "votes": null
    },
    {
      "id": "577375",
      "postDate": "07/16/2019 16:32:55",
      "content": "<p>meanwhile, 84 is the best score I can get with RaLSGAN. After using the bounding boxes somehow my score became even worse (85)</p>",
      "rawMarkdown": "meanwhile, 84 is the best score I can get with RaLSGAN. After using the bounding boxes somehow my score became even worse (85)",
      "votes": null
    },
    {
      "id": "577426",
      "postDate": "07/16/2019 17:09:05",
      "content": "<p>Thanks for sharing .</p>",
      "rawMarkdown": "Thanks for sharing .",
      "votes": null
    },
    {
      "id": "578810",
      "postDate": "07/18/2019 07:07:05",
      "content": "<p>cool~</p>",
      "rawMarkdown": "cool~",
      "votes": null
    },
    {
      "id": "578835",
      "postDate": "07/18/2019 07:32:59",
      "content": "<p>I am still stuck at 74 using RaLSGAN and 72 using DCGAN . Lot of tries , to better the score is underway with limited success .</p>",
      "rawMarkdown": "I am still stuck at 74 using RaLSGAN and 72 using DCGAN . Lot of tries , to better the score is underway with limited success .",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 577229,
      "author_name": "phoenix9032",
      "author_url": "",
      "post_date": "07/16/2019 14:11:44",
      "content": "<p>They are very good material to start . Referred to them a lot while trying out a lot of different GAN . But the best \nsolution for this dataset , still remain elusive to me . \nCommon DCGAN I reached till 72 , with RaLSGAN and RaHingeGAN I probably achieved the similar in 1/3 of the epochs \nbut then they saturate :) \nStill searching for that alchemists stone . ....</p>",
      "votes": null,
      "replies": [
        {
          "id": 577375,
          "author_name": "timetraveller98",
          "author_url": "",
          "post_date": "07/16/2019 16:32:55",
          "content": "<p>meanwhile, 84 is the best score I can get with RaLSGAN. After using the bounding boxes somehow my score became even worse (85)</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 578835,
          "author_name": "phoenix9032",
          "author_url": "",
          "post_date": "07/18/2019 07:32:59",
          "content": "<p>I am still stuck at 74 using RaLSGAN and 72 using DCGAN . Lot of tries , to better the score is underway with limited success .</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 577426,
      "author_name": "gunishj",
      "author_url": "",
      "post_date": "07/16/2019 17:09:05",
      "content": "<p>Thanks for sharing .</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 578810,
      "author_name": "meistermorxrc",
      "author_url": "",
      "post_date": "07/18/2019 07:07:05",
      "content": "<p>cool~</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "577185": "For practice I have found a couple of very good repos with a lot of various GAN architectures implemented in PyTorch and Keras (separately of course).\n\nVery Good for anyone looking for a head start code in GANs.\nhttps://github.com/eriklindernoren/PyTorch-GAN - 3500+ stars\nhttps://github.com/eriklindernoren/Keras-GAN - 5000+ stars\n\nAnother one pointed by Andrew earlier in this competition.\nhttps://www.kaggle.com/c/generative-dog-images/discussion/97772#latest-565566",
    "577229": "They are very good material to start . Referred to them a lot while trying out a lot of different GAN . But the best \nsolution for this dataset , still remain elusive to me . \nCommon DCGAN I reached till 72 , with RaLSGAN and RaHingeGAN I probably achieved the similar in 1/3 of the epochs \nbut then they saturate :) \nStill searching for that alchemists stone . ....",
    "577375": "meanwhile, 84 is the best score I can get with RaLSGAN. After using the bounding boxes somehow my score became even worse (85)",
    "577426": "Thanks for sharing .",
    "578810": "cool~",
    "578835": "I am still stuck at 74 using RaLSGAN and 72 using DCGAN . Lot of tries , to better the score is underway with limited success ."
  },
  "source": "meta"
}