RashedDoha/fast-bpe-tokenizer
A fast C++17 implementation of the Byte Pair Encoding (BPE) tokenization algorithm. Has the ability to train on large corpora of text efficiently.
AI Engineer
A fast C++17 implementation of the Byte Pair Encoding (BPE) tokenization algorithm. Has the ability to train on large corpora of text efficiently.
Designed a ReAct style agent from scratch without any frameworks. Only using Anthropic's python SDK to make cloud LLM calls.
Finetuning an LLM on a bengali cleaned Alpaca instruction dataset.
A basic RAG implementation using langchain and chromadb
PyTorch implementation for https://arxiv.org/abs/1612.05424v2
PyTorch code for training, testing and inference on the Lyft Level5 dataset.
Personal Website
A C++ standalone library for machine learning
higher is a pytorch library allowing users to obtain higher order gradients over losses spanning training loops rather than individual training steps.
PyTorch implementation of https://arxiv.org/abs/2008.00247
Make Semantic Segmentation based Inference on an Input Video of crop rows.
Pytorch Implementation of OpenAI's PixelCNN++
TensorFlow implementation of Conditional Probability Models for Deep Image Compression, published in CVPR 2018
Repo for the website of Buet Robotics Society
The android side of my level 3 term 1 project as during my undergrad studies at BUET.