Skip to content
View Aashutoshh01's full-sized avatar
๐Ÿ’ญ
I may be slow to respond.
๐Ÿ’ญ
I may be slow to respond.

Block or report Aashutoshh01

Report abuse

Contact GitHub support about this userโ€™s behavior. Learn more about reporting abuse.

Report abuse
Aashutoshh01/README.md

Hey I'm Aashutosh Joshi

Currently undergrad at Indian Institute of Technology, Kharagpur

Aashutoshh01's profile views

Banner

๐Ÿ“Œ About Me

  • ๐ŸŒฑ I learn and build across advanced AI/ML systems, transformer architectures, small language models, and complex-valued deep learning.
  • ๐Ÿค I collaborate on open-source AI projects, research prototypes, and high-performance model training pipelines.
  • ๐Ÿ” I work on scalable agentic AI systems, distributed training setups, and multimodal model design.
  • ๐Ÿš€ I build models from scratch, GPT style LLMs, SLMs, ViTs, and multi-agent systems, and design full training and evaluation workflows.
  • ๐Ÿง  I research SAR processing, complex-valued neural networks, and efficient adaptation methods such as LoRA and CV-LoRA.
  • ๐Ÿ’ฌ Ask me about transformers, PyTorch, AutoGen, RAG systems, model optimization, quantization, or H100-class GPU training.
  • โšก I love creating end-to-end AI systems, from dataset pipelines to inference, deployment, and interactive UIs.

๐Ÿง  My Focus Areas

  • AI/ML Research
  • Transformer Architectures
  • Small & Large Language Models (SLMs/LLMs)
  • Complex-Valued Deep Learning
  • LLM Pretraining & Fine-tuning
  • Agentic AI Systems (LangGraph & AutoGen)
  • Retrieval-Augmented Generation (RAG)
  • Open-Source Contribution
  • Model Optimization & Quantization
  • Vision Models (ViT, CNNs)
  • Time-Series Modeling
  • Multi-Agent Automation Systems

๐Ÿ“Š GitHub Stats & Trophies

Aashutoshh01's GitHub Stats Aashutoshh01's GitHub Streak

Aashutoshh01's Activity Graph

๐Ÿ”— Connect with Me

LinkedIN Twitter Gmail

๐Ÿ’ฌ Quote

SLM>>>LLM

Bottom Line

Pinned Loading

  1. paligemma-pytorch paligemma-pytorch Public

    End-to-end reconstruction of PaliGemma in PyTorch, featuring SigLIP ViT, Gemma decoder with RoPE & GQA, KV caching, and autoregressive multimodal inference.

    Python 1

  2. ArXivGPT ArXivGPT Public

    A 124M-parameter GPT-2 model trained from scratch on ArXiv abstracts, featuring fully custom Transformer components implemented in PyTorch.

    Jupyter Notebook 1

  3. VisionSmith VisionSmith Public

    A clean, from-scratch Vision Transformer (ViT) implementation in PyTorch, trained on MNIST and based on the โ€œAn Image is Worth 16ร—16 Wordsโ€ paper. Includes full patch embedding, positional encodingโ€ฆ

    Jupyter Notebook 1

  4. wav2vec2-pytorch wav2vec2-pytorch Public

    End-to-end PyTorch implementation of Wav2Vec2 with convolutional feature encoding, Transformer contextualization, Gumbel vector quantization, and full pipelines for contrastive pre-training and CTCโ€ฆ

    Python 1

  5. ScratchSpeakSLM ScratchSpeakSLM Public

    A tiny GPT-style language model built completely from scratch in PyTorch, inspired by TinyStories, demonstrating how small transformer models can learn to generate coherent text.

    Jupyter Notebook 1

  6. conformer-pytorch conformer-pytorch Public

    End-to-end Conformer ASR in PyTorch featuring Macaron-style blocks, relative multi-head attention, depthwise convolutions, and an LSTM decoder with CTC loss.

    Python 1