Transfer Learning on NLP Task using Tensorflow

Поделиться
HTML-код
  • Опубликовано: 29 янв 2025

Комментарии • 19

  • @venkatesanr9455
    @venkatesanr9455 4 года назад +1

    Thanks for your knowledge sharing and innovative content as usual. Can you share this collab notebook for study purpose

    • @AIEngineeringLife
      @AIEngineeringLife  4 года назад +1

      venkatesan.. I will be uploading it in my repo in few weeks from now once I complete all of my TF related videos. My other videos code are here - github.com/srivatsan88/RUclipsLI

    • @BiranchiNarayanNayak
      @BiranchiNarayanNayak 4 года назад

      @@AIEngineeringLife Thanks

    • @venkatesanr9455
      @venkatesanr9455 4 года назад

      @@AIEngineeringLife Thanks for your kind response and knowledge sharing

    • @venkatesanr9455
      @venkatesanr9455 4 года назад

      @@AIEngineeringLife Thanks for your kind response and knowledge sharing

  • @bibhukalyandas1146
    @bibhukalyandas1146 4 года назад +1

    Sirji,can we apply preprocessing, tokenization,stemmization,stop words removal to the sentence manually give input to this pretrained model or we need to use any other pretrained model for my case?

    • @AIEngineeringLife
      @AIEngineeringLife  4 года назад

      Bibhu.. On transfer learning I have not tried it but I am sure it must be possible by passing an custom tokenizer and embedding as well. Let me try it later over the weekend and update

    • @bibhukalyandas1146
      @bibhukalyandas1146 4 года назад

      @@AIEngineeringLife sure thnks,will be waiting for your reply

    • @bibhukalyandas1146
      @bibhukalyandas1146 4 года назад

      @@AIEngineeringLife sure thnks,will be waiting for your reply

    • @bibhukalyandas1146
      @bibhukalyandas1146 4 года назад

      @@AIEngineeringLife sure thnks,will be waiting for your reply

  • @FindMultiBagger
    @FindMultiBagger 4 года назад +2

    Why you don't use keras rather than tf ?

    • @AIEngineeringLife
      @AIEngineeringLife  4 года назад

      Vinay.. I am still using tf.keras to add on to existing model.. You mean why not use pure keras?
      I feel either was it must work

    • @FindMultiBagger
      @FindMultiBagger 4 года назад

      @@AIEngineeringLife I mean , pure keras as api for everything ! It might easier to grasph rather than core tf.

  • @vishalkesti6131
    @vishalkesti6131 4 года назад +1

    Sir, is it ok if we pass our data as it is ? I am just wondering should I preprocess my data and then pass it for the pretrained model. Kindly let me know if this is right?

    • @AIEngineeringLife
      @AIEngineeringLife  4 года назад

      In this case you can pass as it is as the pretrained model accepts sentence. But you might want to tokenize it in some cases or in some simple models remove stop words and all

  • @valerysalov8208
    @valerysalov8208 4 года назад

    How to make your dataset gold standard?

  • @gokulyc
    @gokulyc 4 года назад +1

    can i get notebook link?

    • @AIEngineeringLife
      @AIEngineeringLife  4 года назад

      Must be in this repo - github.com/srivatsan88/Natural-Language-Processing or in my RUclipsLI repo

  • @arpanghosh3801
    @arpanghosh3801 4 года назад

    Please share the code