Skip to content
View ParagEkbote's full-sized avatar
🪧
https://paragekbote.github.io/
🪧
https://paragekbote.github.io/
34 followers · 34 following

Block or report ParagEkbote

Report abuse

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

Report abuse
ParagEkbote/README.md

Typing SVG

Machine Learning Engineer

Working on transformer-based models, systems and ML infra.

Core Focus Areas

  • Large Language Models and the Hugging Face ecosystem.
  • Parameter-efficient fine-tuning (LoRA / PEFT).
  • LLM inference benchmarking and optimization.
  • Reproducible ML experimentation and open-source tooling.

Open Source Contributions

I actively contribute to machine learning projects across the ecosystem.

View Contributions

Find Me Online

My Work

To know more about my work:

Portfolio

Pinned Loading

  1. quantized-containerized-models quantized-containerized-models Public

    Deploy AI models with an API through quantization and containerization.

    Python 1

  2. model-kernels model-kernels Public

    Cuda 5 1

  3. compare_tokenizer.py compare_tokenizer.py
    1 
    #!/usr/bin/env python3
    2 
    """
    3 
    Tokenizer Performance Comparison
    4 
    """
    5 
    # ------------------------------------------------
  4. inference_optimization_strategy.py inference_optimization_strategy.py
    1 
    import os
    2 
    import time
    3 
     
     
     4 
    import numpy as np
     
     
     5 
    import matplotlib.pyplot as plt
     
     
     
     
     
    5.  
  5.  
     
     
      Pruna Smash Setup for SmollM Collect... Pruna Smash Setup for SmollM Collection: https://huggingface.co/collections/AINovice2005/smollm-smashed-tiny-giants-optimized-for-speed-68e16113d90eab12bae42a34 
     
     
     1 
    import torch
     
     
     2 
    from transformers import AutoModelForCausalLM, AutoTokenizer
     
     
     3 
    from pruna import smash, SmashConfig
     
     
     4 
     
     
     5 
    # Define per-model Smash + torch.compile configs
     
     
     
     
     
    6.  
  6.  
     
     
     
      hf-model-inspector hf-model-inspector Public 
     
     
     A package and gh-action to generate quality reports on Hugging Face model stats and usage, addressing ambiguity in open model releases. 
     
       Python      1  
     
     
     
    7.