{
  "id": 309550,
  "title": "TorchRec: PyTorch for Recommendation Systems! ",
  "url": "/competitions/h-and-m-personalized-fashion-recommendations/discussion/309550",
  "author_name": "Sanyam Bhutani",
  "post_date": "2022-02-24T04:35:28.610000",
  "votes": 32,
  "comment_count": 0,
  "views": 0,
  "content": "<p>The amazing PyTorch Team has just released a \"domain library for Recommendation Systems\":</p>\n<p>I skimmed through the announcement post, docs, and getting starting colab. There is a lot of focus on making the workflows work with Multi-GPUs, optimising operations and GPU inferencing for the large dataset sizes.</p>\n<p>Similar to <code>DDP</code>, the main API function is <code>DMP</code>: Distributed Model Parallel </p>\n<p>Here are the links and resources for getting started:</p>\n<ul>\n<li><a href=\"https://pytorch.org/blog/introducing-torchrec/\" target=\"_blank\">Announcement Blog</a></li>\n<li><a href=\"https://github.com/pytorch/torchrec\" target=\"_blank\">Github</a></li>\n<li><a href=\"https://pytorch.org/torchrec/\" target=\"_blank\">Docs</a></li>\n<li><a href=\"https://colab.research.google.com/github/pytorch/torchrec/blob/main/Torchrec_Introduction.ipynb\" target=\"_blank\">Getting started Colab</a></li>\n</ul>",
  "messages": [
    {
      "id": 1702929,
      "postDate": "2022-02-24T04:35:28.610Z",
      "content": "<p>The amazing PyTorch Team has just released a \"domain library for Recommendation Systems\":</p>\n<p>I skimmed through the announcement post, docs, and getting starting colab. There is a lot of focus on making the workflows work with Multi-GPUs, optimising operations and GPU inferencing for the large dataset sizes.</p>\n<p>Similar to <code>DDP</code>, the main API function is <code>DMP</code>: Distributed Model Parallel </p>\n<p>Here are the links and resources for getting started:</p>\n<ul>\n<li><a href=\"https://pytorch.org/blog/introducing-torchrec/\" target=\"_blank\">Announcement Blog</a></li>\n<li><a href=\"https://github.com/pytorch/torchrec\" target=\"_blank\">Github</a></li>\n<li><a href=\"https://pytorch.org/torchrec/\" target=\"_blank\">Docs</a></li>\n<li><a href=\"https://colab.research.google.com/github/pytorch/torchrec/blob/main/Torchrec_Introduction.ipynb\" target=\"_blank\">Getting started Colab</a></li>\n</ul>",
      "rawMarkdown": "The amazing PyTorch Team has just released a \"domain library for Recommendation Systems\":\n\nI skimmed through the announcement post, docs, and getting starting colab. There is a lot of focus on making the workflows work with Multi-GPUs, optimising operations and GPU inferencing for the large dataset sizes.\n\nSimilar to `DDP`, the main API function is `DMP`: Distributed Model Parallel \n\nHere are the links and resources for getting started:\n\n- [Announcement Blog](https://pytorch.org/blog/introducing-torchrec/)\n- [Github](https://github.com/pytorch/torchrec)\n- [Docs](https://pytorch.org/torchrec/)\n- [Getting started Colab](https://colab.research.google.com/github/pytorch/torchrec/blob/main/Torchrec_Introduction.ipynb)",
      "votes": 32
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1702929": "The amazing PyTorch Team has just released a \"domain library for Recommendation Systems\":\n\nI skimmed through the announcement post, docs, and getting starting colab. There is a lot of focus on making the workflows work with Multi-GPUs, optimising operations and GPU inferencing for the large dataset sizes.\n\nSimilar to `DDP`, the main API function is `DMP`: Distributed Model Parallel \n\nHere are the links and resources for getting started:\n\n- [Announcement Blog](https://pytorch.org/blog/introducing-torchrec/)\n- [Github](https://github.com/pytorch/torchrec)\n- [Docs](https://pytorch.org/torchrec/)\n- [Getting started Colab](https://colab.research.google.com/github/pytorch/torchrec/blob/main/Torchrec_Introduction.ipynb)"
  }
}