{
  "id": 18199,
  "title": "[external data] MXNet Pretrained Model",
  "url": "/competitions/yelp-restaurant-photo-classification/discussion/18199",
  "author_name": "",
  "post_date": "2015-12-31T21:23:48.277Z",
  "votes": 12,
  "comment_count": 11,
  "views": 6032,
  "content": "<p><strong>1. Inception-BN Network</strong></p>\n\n<p>This model is a pretrained model on ILSVRC2012 dataset. The model is trained by only random crop and mirror augmentation. This model is able to achieve 89.9% Top-5 accuracy on ILSVRC2012-Validation Set.</p>\n\n<p><a href=\"https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-bn.md\">https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-bn.md</a></p>\n\n<p><strong>2. Inception-V3 Network</strong></p>\n\n<p>This model is converted from TensorFlow released pretrained model. By single crop on 299 x 299 image from 384 x 384 image, this model is able to achieve 76.88% Top-1 Accuracy and 93.344% Top-5 Accuracy on ILSVRC2012-Validation Set.</p>\n\n<p><a href=\"https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-v3.md\">https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-v3.md</a></p>\n\n<p><strong>3. Full ImageNet Network</strong></p>\n\n<p>This model is a pretrained model on full imagenet dataset with 14,197,087 iamges in 21,841 classes. The model is trained by only random crop and mirror augmentation. This model is able to achieve 37.19% Top-1 accuracy on training data. This model is about 50% more complex than standard Inception-BN Network</p>\n\n<p><a href=\"https://github.com/dmlc/mxnet-model-gallery/tree/master/imagenet-21k-inception\">https://github.com/dmlc/mxnet-model-gallery/tree/master/imagenet-21k-inception</a></p>\n\n<p><em>How to use pretrained Network (Example for Inception-BN)</em></p>\n\n<p><a href=\"https://github.com/dmlc/mxnet/blob/master/example/notebooks/predict-with-pretrained-model.ipynb\">https://github.com/dmlc/mxnet/blob/master/example/notebooks/predict-with-pretrained-model.ipynb</a></p>\n\n<p><em>Extra Note</em></p>\n\n<ol>\n<li>For Full ImageNet Network, we currently use Github LFS to store it. As a group of students, we choose the free version of LFS. If you find &quot;out of quota&quot;, we are sorry for that and please wait for next month to download it.</li>\n<li>Different model uses different preprocessing. Please read README before use. </li>\n</ol>",
  "messages": [
    {
      "id": "103326",
      "postDate": "12/31/2015 21:23:48",
      "content": "<p><strong>1. Inception-BN Network</strong></p>\n\n<p>This model is a pretrained model on ILSVRC2012 dataset. The model is trained by only random crop and mirror augmentation. This model is able to achieve 89.9% Top-5 accuracy on ILSVRC2012-Validation Set.</p>\n\n<p><a href=\"https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-bn.md\">https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-bn.md</a></p>\n\n<p><strong>2. Inception-V3 Network</strong></p>\n\n<p>This model is converted from TensorFlow released pretrained model. By single crop on 299 x 299 image from 384 x 384 image, this model is able to achieve 76.88% Top-1 Accuracy and 93.344% Top-5 Accuracy on ILSVRC2012-Validation Set.</p>\n\n<p><a href=\"https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-v3.md\">https://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-v3.md</a></p>\n\n<p><strong>3. Full ImageNet Network</strong></p>\n\n<p>This model is a pretrained model on full imagenet dataset with 14,197,087 iamges in 21,841 classes. The model is trained by only random crop and mirror augmentation. This model is able to achieve 37.19% Top-1 accuracy on training data. This model is about 50% more complex than standard Inception-BN Network</p>\n\n<p><a href=\"https://github.com/dmlc/mxnet-model-gallery/tree/master/imagenet-21k-inception\">https://github.com/dmlc/mxnet-model-gallery/tree/master/imagenet-21k-inception</a></p>\n\n<p><em>How to use pretrained Network (Example for Inception-BN)</em></p>\n\n<p><a href=\"https://github.com/dmlc/mxnet/blob/master/example/notebooks/predict-with-pretrained-model.ipynb\">https://github.com/dmlc/mxnet/blob/master/example/notebooks/predict-with-pretrained-model.ipynb</a></p>\n\n<p><em>Extra Note</em></p>\n\n<ol>\n<li>For Full ImageNet Network, we currently use Github LFS to store it. As a group of students, we choose the free version of LFS. If you find &quot;out of quota&quot;, we are sorry for that and please wait for next month to download it.</li>\n<li>Different model uses different preprocessing. Please read README before use. </li>\n</ol>",
      "rawMarkdown": "**1. Inception-BN Network**\r\n\r\nThis model is a pretrained model on ILSVRC2012 dataset. The model is trained by only random crop and mirror augmentation. This model is able to achieve 89.9% Top-5 accuracy on ILSVRC2012-Validation Set.\r\n\r\nhttps://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-bn.md\r\n\r\n**2. Inception-V3 Network**\r\n\r\nThis model is converted from TensorFlow released pretrained model. By single crop on 299 x 299 image from 384 x 384 image, this model is able to achieve 76.88% Top-1 Accuracy and 93.344% Top-5 Accuracy on ILSVRC2012-Validation Set.\r\n\r\nhttps://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-v3.md\r\n\r\n**3. Full ImageNet Network**\r\n\r\nThis model is a pretrained model on full imagenet dataset with 14,197,087 iamges in 21,841 classes. The model is trained by only random crop and mirror augmentation. This model is able to achieve 37.19% Top-1 accuracy on training data. This model is about 50% more complex than standard Inception-BN Network\r\n\r\nhttps://github.com/dmlc/mxnet-model-gallery/tree/master/imagenet-21k-inception\r\n\r\n\r\n*How to use pretrained Network (Example for Inception-BN)*\r\n\r\nhttps://github.com/dmlc/mxnet/blob/master/example/notebooks/predict-with-pretrained-model.ipynb\r\n\r\n*Extra Note*\r\n\r\n1. For Full ImageNet Network, we currently use Github LFS to store it. As a group of students, we choose the free version of LFS. If you find \"out of quota\", we are sorry for that and please wait for next month to download it.\r\n2. Different model uses different preprocessing. Please read README before use.",
      "votes": null
    },
    {
      "id": "103394",
      "postDate": "01/02/2016 05:56:50",
      "content": "<p>Hi Bing, thanks for your explanation.  The descriptions of the layers make sense.  However, if external data is not equally accessible to all contestants, then we may not allow it.  Can you also add licensing information to this thread?</p>",
      "rawMarkdown": "Hi Bing, thanks for your explanation.  The descriptions of the layers make sense.  However, if external data is not equally accessible to all contestants, then we may not allow it.  Can you also add licensing information to this thread?",
      "votes": null
    },
    {
      "id": "103396",
      "postDate": "01/02/2016 06:00:43",
      "content": "<ol>\n<li>Inception-BN Network: CC0 Licence</li>\n<li>Inception-V3 Network: TensorFlow Apache 2.0 Licence</li>\n<li>Full ImageNet Network: CC0 Licence</li>\n</ol>",
      "rawMarkdown": "1. Inception-BN Network: CC0 Licence\r\n2. Inception-V3 Network: TensorFlow Apache 2.0 Licence\r\n3. Full ImageNet Network: CC0 Licence",
      "votes": null
    },
    {
      "id": "103418",
      "postDate": "01/02/2016 15:22:41",
      "content": "<p>Hi Bing, I tried to download the Inception-Full.zip file but got an out of quota error.</p>\n\n<p>Can you host it someplace else like Dropbox?</p>",
      "rawMarkdown": "Hi Bing, I tried to download the Inception-Full.zip file but got an out of quota error.\r\n\r\nCan you host it someplace else like Dropbox?",
      "votes": null
    },
    {
      "id": "103425",
      "postDate": "01/02/2016 18:24:56",
      "content": "<p>I am sorry recently I formatted my system and I have no backup. I am not sure when the quota will be reset. Once I find my backup I will seek a new place to host it.</p>",
      "rawMarkdown": "I am sorry recently I formatted my system and I have no backup. I am not sure when the quota will be reset. Once I find my backup I will seek a new place to host it.",
      "votes": null
    },
    {
      "id": "103432",
      "postDate": "01/02/2016 20:11:47",
      "content": "<p>MXnet rocks!</p>",
      "rawMarkdown": "MXnet rocks!",
      "votes": null
    },
    {
      "id": "104617",
      "postDate": "01/14/2016 15:36:53",
      "content": "<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>",
      "rawMarkdown": "Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.",
      "votes": null
    },
    {
      "id": "104618",
      "postDate": "01/14/2016 15:44:07",
      "content": "<p>[quote=Rob Espinosa;104617]</p>\n\n<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>\n\n<p>[/quote]</p>\n\n<p>did you see this one: <a href=\"http://stackoverflow.com/questions/33668323/clang-omp-in-xcode-under-el-capitan\">http://stackoverflow.com/questions/33668323/clang-omp-in-xcode-under-el-capitan</a>\n?</p>",
      "rawMarkdown": "[quote=Rob Espinosa;104617]\r\n\r\nHas anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.\r\n\r\n\r\n\r\n[/quote]\r\n\r\n\r\ndid you see this one: http://stackoverflow.com/questions/33668323/clang-omp-in-xcode-under-el-capitan\r\n?",
      "votes": null
    },
    {
      "id": "104724",
      "postDate": "01/15/2016 21:31:17",
      "content": "<ol>\n<li>one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: <a href=\"https://github.com/dmlc/mxnet/issues/324\">https://github.com/dmlc/mxnet/issues/324</a></li>\n<li>link the path according to the official installation guide <a href=\"https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\">https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx</a></li>\n<li>optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution <a href=\"https://github.com/dmlc/mxnet/issues/645\">https://github.com/dmlc/mxnet/issues/645</a>\n[quote=Rob Espinosa;104617]</li>\n</ol>\n\n<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>\n\n<p>[/quote]</p>",
      "rawMarkdown": "1. one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: https://github.com/dmlc/mxnet/issues/324\r\n2. link the path according to the official installation guide https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\r\n3. optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution https://github.com/dmlc/mxnet/issues/645\r\n[quote=Rob Espinosa;104617]\r\n\r\nHas anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.\r\n\r\n\r\n\r\n[/quote]",
      "votes": null
    },
    {
      "id": "104969",
      "postDate": "01/18/2016 13:07:40",
      "content": "<p>Thank you. Fixed it.</p>\n\n<p>[quote=phunter;104724]</p>\n\n<ol>\n<li>one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: <a href=\"https://github.com/dmlc/mxnet/issues/324\">https://github.com/dmlc/mxnet/issues/324</a></li>\n<li>link the path according to the official installation guide <a href=\"https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\">https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx</a></li>\n<li>optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution <a href=\"https://github.com/dmlc/mxnet/issues/645\">https://github.com/dmlc/mxnet/issues/645</a>\n[quote=Rob Espinosa;104617]</li>\n</ol>\n\n<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>\n\n<p>[/quote]</p>\n\n<p>[/quote]</p>",
      "rawMarkdown": "Thank you. Fixed it.\r\n\r\n[quote=phunter;104724]\r\n\r\n1. one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: https://github.com/dmlc/mxnet/issues/324\r\n2. link the path according to the official installation guide https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\r\n3. optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution https://github.com/dmlc/mxnet/issues/645\r\n[quote=Rob Espinosa;104617]\r\n\r\nHas anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.\r\n\r\n\r\n\r\n[/quote]\r\n\r\n\r\n[/quote]",
      "votes": null
    },
    {
      "id": "106724",
      "postDate": "02/03/2016 14:52:20",
      "content": "<p>Hi Bing,</p>\n\n<p>did you manage to get a copy of the models that you could upload?</p>\n\n<p>Thank you\nLudwig</p>",
      "rawMarkdown": "Hi Bing,\r\n\r\ndid you manage to get a copy of the models that you could upload?\r\n\r\nThank you\r\nLudwig",
      "votes": null
    },
    {
      "id": "112609",
      "postDate": "03/22/2016 10:15:47",
      "content": "<p>So, the inception-21k is pretty bad right ? ,  would it be the same if we use the latest libraries that were released by google to train the inception v3 with 21k synstes ? or would there be a significant difference ? </p>\n\n<p>has someone done it ? </p>",
      "rawMarkdown": "So, the inception-21k is pretty bad right ? ,  would it be the same if we use the latest libraries that were released by google to train the inception v3 with 21k synstes ? or would there be a significant difference ? \r\n\r\nhas someone done it ?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 103394,
      "author_name": "",
      "author_url": "",
      "post_date": "01/02/2016 05:56:50",
      "content": "<p>Hi Bing, thanks for your explanation.  The descriptions of the layers make sense.  However, if external data is not equally accessible to all contestants, then we may not allow it.  Can you also add licensing information to this thread?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 103396,
      "author_name": "binghsu",
      "author_url": "",
      "post_date": "01/02/2016 06:00:43",
      "content": "<ol>\n<li>Inception-BN Network: CC0 Licence</li>\n<li>Inception-V3 Network: TensorFlow Apache 2.0 Licence</li>\n<li>Full ImageNet Network: CC0 Licence</li>\n</ol>",
      "votes": null,
      "replies": []
    },
    {
      "id": 103418,
      "author_name": "emolson",
      "author_url": "",
      "post_date": "01/02/2016 15:22:41",
      "content": "<p>Hi Bing, I tried to download the Inception-Full.zip file but got an out of quota error.</p>\n\n<p>Can you host it someplace else like Dropbox?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 103425,
      "author_name": "binghsu",
      "author_url": "",
      "post_date": "01/02/2016 18:24:56",
      "content": "<p>I am sorry recently I formatted my system and I have no backup. I am not sure when the quota will be reset. Once I find my backup I will seek a new place to host it.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 103432,
      "author_name": "phunter",
      "author_url": "",
      "post_date": "01/02/2016 20:11:47",
      "content": "<p>MXnet rocks!</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 104617,
      "author_name": "bytyey",
      "author_url": "",
      "post_date": "01/14/2016 15:36:53",
      "content": "<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 104618,
      "author_name": "abhishek",
      "author_url": "",
      "post_date": "01/14/2016 15:44:07",
      "content": "<p>[quote=Rob Espinosa;104617]</p>\n\n<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>\n\n<p>[/quote]</p>\n\n<p>did you see this one: <a href=\"http://stackoverflow.com/questions/33668323/clang-omp-in-xcode-under-el-capitan\">http://stackoverflow.com/questions/33668323/clang-omp-in-xcode-under-el-capitan</a>\n?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 104724,
      "author_name": "phunter",
      "author_url": "",
      "post_date": "01/15/2016 21:31:17",
      "content": "<ol>\n<li>one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: <a href=\"https://github.com/dmlc/mxnet/issues/324\">https://github.com/dmlc/mxnet/issues/324</a></li>\n<li>link the path according to the official installation guide <a href=\"https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\">https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx</a></li>\n<li>optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution <a href=\"https://github.com/dmlc/mxnet/issues/645\">https://github.com/dmlc/mxnet/issues/645</a>\n[quote=Rob Espinosa;104617]</li>\n</ol>\n\n<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>\n\n<p>[/quote]</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 104969,
      "author_name": "bytyey",
      "author_url": "",
      "post_date": "01/18/2016 13:07:40",
      "content": "<p>Thank you. Fixed it.</p>\n\n<p>[quote=phunter;104724]</p>\n\n<ol>\n<li>one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: <a href=\"https://github.com/dmlc/mxnet/issues/324\">https://github.com/dmlc/mxnet/issues/324</a></li>\n<li>link the path according to the official installation guide <a href=\"https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\">https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx</a></li>\n<li>optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution <a href=\"https://github.com/dmlc/mxnet/issues/645\">https://github.com/dmlc/mxnet/issues/645</a>\n[quote=Rob Espinosa;104617]</li>\n</ol>\n\n<p>Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:</p>\n\n<p>./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with  include; use &quot;quotes&quot;\n      instead</p>\n\n<p>I ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.</p>\n\n<p>[/quote]</p>\n\n<p>[/quote]</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 106724,
      "author_name": "shackenberg",
      "author_url": "",
      "post_date": "02/03/2016 14:52:20",
      "content": "<p>Hi Bing,</p>\n\n<p>did you manage to get a copy of the models that you could upload?</p>\n\n<p>Thank you\nLudwig</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 112609,
      "author_name": "as1525",
      "author_url": "",
      "post_date": "03/22/2016 10:15:47",
      "content": "<p>So, the inception-21k is pretty bad right ? ,  would it be the same if we use the latest libraries that were released by google to train the inception v3 with 21k synstes ? or would there be a significant difference ? </p>\n\n<p>has someone done it ? </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "103326": "**1. Inception-BN Network**\r\n\r\nThis model is a pretrained model on ILSVRC2012 dataset. The model is trained by only random crop and mirror augmentation. This model is able to achieve 89.9% Top-5 accuracy on ILSVRC2012-Validation Set.\r\n\r\nhttps://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-bn.md\r\n\r\n**2. Inception-V3 Network**\r\n\r\nThis model is converted from TensorFlow released pretrained model. By single crop on 299 x 299 image from 384 x 384 image, this model is able to achieve 76.88% Top-1 Accuracy and 93.344% Top-5 Accuracy on ILSVRC2012-Validation Set.\r\n\r\nhttps://github.com/dmlc/mxnet-model-gallery/blob/master/imagenet-1k-inception-v3.md\r\n\r\n**3. Full ImageNet Network**\r\n\r\nThis model is a pretrained model on full imagenet dataset with 14,197,087 iamges in 21,841 classes. The model is trained by only random crop and mirror augmentation. This model is able to achieve 37.19% Top-1 accuracy on training data. This model is about 50% more complex than standard Inception-BN Network\r\n\r\nhttps://github.com/dmlc/mxnet-model-gallery/tree/master/imagenet-21k-inception\r\n\r\n\r\n*How to use pretrained Network (Example for Inception-BN)*\r\n\r\nhttps://github.com/dmlc/mxnet/blob/master/example/notebooks/predict-with-pretrained-model.ipynb\r\n\r\n*Extra Note*\r\n\r\n1. For Full ImageNet Network, we currently use Github LFS to store it. As a group of students, we choose the free version of LFS. If you find \"out of quota\", we are sorry for that and please wait for next month to download it.\r\n2. Different model uses different preprocessing. Please read README before use.",
    "103394": "Hi Bing, thanks for your explanation.  The descriptions of the layers make sense.  However, if external data is not equally accessible to all contestants, then we may not allow it.  Can you also add licensing information to this thread?",
    "103396": "1. Inception-BN Network: CC0 Licence\r\n2. Inception-V3 Network: TensorFlow Apache 2.0 Licence\r\n3. Full ImageNet Network: CC0 Licence",
    "103418": "Hi Bing, I tried to download the Inception-Full.zip file but got an out of quota error.\r\n\r\nCan you host it someplace else like Dropbox?",
    "103425": "I am sorry recently I formatted my system and I have no backup. I am not sure when the quota will be reset. Once I find my backup I will seek a new place to host it.",
    "103432": "MXnet rocks!",
    "104617": "Has anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.",
    "104618": "[quote=Rob Espinosa;104617]\r\n\r\nHas anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.\r\n\r\n\r\n\r\n[/quote]\r\n\r\n\r\ndid you see this one: http://stackoverflow.com/questions/33668323/clang-omp-in-xcode-under-el-capitan\r\n?",
    "104724": "1. one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: https://github.com/dmlc/mxnet/issues/324\r\n2. link the path according to the official installation guide https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\r\n3. optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution https://github.com/dmlc/mxnet/issues/645\r\n[quote=Rob Espinosa;104617]\r\n\r\nHas anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.\r\n\r\n\r\n\r\n[/quote]",
    "104969": "Thank you. Fixed it.\r\n\r\n[quote=phunter;104724]\r\n\r\n1. one should stay with clang in Mac and turn off USE_OPENMP, since opencv from brew requires this. Ref: https://github.com/dmlc/mxnet/issues/324\r\n2. link the path according to the official installation guide https://github.com/dmlc/mxnet/blob/master/doc/build.md#building-on-osx\r\n3. optional: clang from Xcode doesn't support openmp, but clang-omp supports, please follow this issue and solution https://github.com/dmlc/mxnet/issues/645\r\n[quote=Rob Espinosa;104617]\r\n\r\nHas anyone tried installing MXnet on OS X El Capitan?  I'm getting the following error:\r\n\r\n./dmlc-core/include/dmlc/omp.h:9:10: error: 'omp.h' file not found with <angled> include; use \"quotes\"\r\n      instead\r\n\r\nI ran into several other errors that I was able to solve with the help of prior StackExchange threads. But have had no luck with this one.\r\n\r\n\r\n\r\n[/quote]\r\n\r\n\r\n[/quote]",
    "106724": "Hi Bing,\r\n\r\ndid you manage to get a copy of the models that you could upload?\r\n\r\nThank you\r\nLudwig",
    "112609": "So, the inception-21k is pretty bad right ? ,  would it be the same if we use the latest libraries that were released by google to train the inception v3 with 21k synstes ? or would there be a significant difference ? \r\n\r\nhas someone done it ?"
  },
  "source": "meta"
}