Skip to content
View manisnesan's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report manisnesan

Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
manisnesan/README.md

Hi there 👋

  • 🔭 I’m currently working as a AI Technical Strategy Lead | Senior Principal Software Engineer @ Red Hat on Search Relevance, Language Understanding and Ranking.
  • 🌱 I’m currently learning Deep Learning as a fast.ai international fellow.

Links

Talks

Papers

  • Paper - HASOC 2020 conference describes our team's approach to the solution submitted using Bert Fine-tuning for the Identification of Hate Speech and Offensive Language in Indo-European Languages.
    • Abstract: This article describes our team Chrestotes’ approach to the solution submitted to HASOC 2020: Hate Speech and Offensive Content Identification in Indo- European Languages. We demonstrate an end to end solution to the fine-grained detection of hate speech in tweets. Our solution is focused on the English Task which has been split into two subtasks. Our model achieved macro-average f1- scores of 0.4969 and 0.2652 on the subtasks A and B respectively. This solution places us in the middle of the leaderboard for subtask A and first place for subtask B.

Certifications

  • International Fellowship to attend fastai - Deep Learning for Coders course 2022 | 2020
  • Search with Machine Learning - Co:rise March 2022
  • Full Stack Deep Learning - Spring 2021 - fastclean : Experiments to find incorrect labels in the dataset and noisy training
  • [Coursera] Machine Learning Foundations: A Case Study Approach - Jan 2016
  • [Coursera] Cloud Computing Applications - Oct 2015
  • [Coursera] Functional Programming Principles in Scale - Jun 2014

Pinned Loading

  1. fastclean fastclean Public template

    [FSDL - Spring 2021] Finding noisy labels from datasets & Noisy Training

    Jupyter Notebook 2

  2. fastchai fastchai Public

    Repository capturing deep learning & nlp experiments using fastai & pytorch

    HTML 2

  3. til til Public

    collection of today i learned scripts

    Jupyter Notebook 4

  4. nlp-notes nlp-notes Public

    NLP Learning Journey

    Jupyter Notebook 1

  5. Prodigy Usage Prodigy Usage
    1
    ## Teach the model ie Annotate
    2
    `$ prodigy textcat.teach troubleshoot-sample en_core_web_sm troubleshoot_sample.jsonl `
    3
     
    4
    ## Batch Train
    5
    `$ prodigy textcat.batch-train troubleshoot-sample en_vectors_web_lg --output troubleshoot-sample-model --eval-split 0.2`
  6. JQ Tricks JQ Tricks
    1
    ## JQ to filter by value
    2
    Syntax: <strong>`cat <filename> | jq -c '.[] | select( .<key> | contains("<value>"))'`</strong>
    3
     
    4
    Example: To get json record having _id equal 611
    5
    ```bash