{
  "id": 100667,
  "title": "Triplet Loss: how to select triplets?",
  "url": "/competitions/recursion-cellular-image-classification/discussion/100667",
  "author_name": "",
  "post_date": "2019-07-20T02:25:45.134859800Z",
  "votes": 5,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Is there a good way of selecting triplets efficiently? Any paper or github repo is welcome! To me it is quite a difficult task of selecting proper triplets... </p>\n\n<p>And is triplet loss useful? Thanks!</p>",
  "messages": [
    {
      "id": "580328",
      "postDate": "07/20/2019 02:25:45",
      "content": "<p>Is there a good way of selecting triplets efficiently? Any paper or github repo is welcome! To me it is quite a difficult task of selecting proper triplets... </p>\n\n<p>And is triplet loss useful? Thanks!</p>",
      "rawMarkdown": "Is there a good way of selecting triplets efficiently? Any paper or github repo is welcome! To me it is quite a difficult task of selecting proper triplets... \n\nAnd is triplet loss useful? Thanks!",
      "votes": null
    },
    {
      "id": "1949389",
      "postDate": "09/21/2022 16:40:59",
      "content": "<pre><code>from sentence_transformers import SentenceTransformer,  SentencesDataset, LoggingHandler, losses\nfrom sentence_transformers.readers import InputExample\n\nmodel = SentenceTransformer('distilbert-base-nli-mean-tokens')\ntrain_examples = [InputExample(texts=['Anchor 1', 'Positive 1', 'Negative 1']),\n    InputExample(texts=['Anchor 2', 'Positive 2', 'Negative 2'])]\ntrain_dataset = SentencesDataset(train_examples, model)\ntrain_dataloader = DataLoader(train_dataset, shuffle=True, batch_size=train_batch_size)\ntrain_loss = losses.TripletLoss(model=model)\n</code></pre>",
      "rawMarkdown": "```\nfrom sentence_transformers import SentenceTransformer,  SentencesDataset, LoggingHandler, losses\nfrom sentence_transformers.readers import InputExample\n\nmodel = SentenceTransformer('distilbert-base-nli-mean-tokens')\ntrain_examples = [InputExample(texts=['Anchor 1', 'Positive 1', 'Negative 1']),\n    InputExample(texts=['Anchor 2', 'Positive 2', 'Negative 2'])]\ntrain_dataset = SentencesDataset(train_examples, model)\ntrain_dataloader = DataLoader(train_dataset, shuffle=True, batch_size=train_batch_size)\ntrain_loss = losses.TripletLoss(model=model)\n```",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1949389,
      "author_name": "mustafawaheed91",
      "author_url": "",
      "post_date": "09/21/2022 16:40:59",
      "content": "<pre><code>from sentence_transformers import SentenceTransformer,  SentencesDataset, LoggingHandler, losses\nfrom sentence_transformers.readers import InputExample\n\nmodel = SentenceTransformer('distilbert-base-nli-mean-tokens')\ntrain_examples = [InputExample(texts=['Anchor 1', 'Positive 1', 'Negative 1']),\n    InputExample(texts=['Anchor 2', 'Positive 2', 'Negative 2'])]\ntrain_dataset = SentencesDataset(train_examples, model)\ntrain_dataloader = DataLoader(train_dataset, shuffle=True, batch_size=train_batch_size)\ntrain_loss = losses.TripletLoss(model=model)\n</code></pre>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "580328": "Is there a good way of selecting triplets efficiently? Any paper or github repo is welcome! To me it is quite a difficult task of selecting proper triplets... \n\nAnd is triplet loss useful? Thanks!",
    "1949389": "```\nfrom sentence_transformers import SentenceTransformer,  SentencesDataset, LoggingHandler, losses\nfrom sentence_transformers.readers import InputExample\n\nmodel = SentenceTransformer('distilbert-base-nli-mean-tokens')\ntrain_examples = [InputExample(texts=['Anchor 1', 'Positive 1', 'Negative 1']),\n    InputExample(texts=['Anchor 2', 'Positive 2', 'Negative 2'])]\ntrain_dataset = SentencesDataset(train_examples, model)\ntrain_dataloader = DataLoader(train_dataset, shuffle=True, batch_size=train_batch_size)\ntrain_loss = losses.TripletLoss(model=model)\n```"
  },
  "source": "meta"
}