United States

Ashim Dahal

Elite
@ashimdahal

making sand decipher its surrounding

torchy. A pytroch wrapper that adds more functionality and reduces code.

15

jelly. A chatbot that replies to romanized Nepali.

9

ViT-vs-CNN-Image-Segmentation. Comparision of performance mass trained and fine tuned ViT with a generic CNN model using a novel combined loss function

6

clip-shift-analysis. [MIV at CVPR 2025 (Proceedings Track)] analyzing clip embeddings shift with different augmentation to the same input image averaged over 200k images

6

attachment-scraper. Scrapes emails from given list of emails and efficiently stores attachments in user specific folder.

4

ashimdahal. Config files for my GitHub profile.

2

VR-Dyslexia-with-Llama. A simple AR solution to help people with dyslexia practice writing and reading dyslexia unfriendly words

2

SchoolManagment-System. It is a complete school managment system project for the Balmiki Education Foundation For a Exhibition

2

FRIDA-TEAM_AIR. Github repo for submission of Frida application from TEAM AIR(Artificial Intelligence Realm) for Kathmandu University Hackfest

2

zero_day_ATTACK_detection. A project by Team Cyber Innovations Lab under the supervision of Dr. Nick Rahimi

1

Document-Scanner. Python

1

ADNet-in-PyTorch. Implementation of ADNet for March machine learning madness competition in Kaggle

1

Anytranslate. Anytranslate is an API wrapper for google translate

1

nepse-cli. This is a CLI app for NEPSE data

1

DRCFS_Chitwan_ML_Bootcamp. 7 day long ML bootcamp organized by DRCFS, Sungava College and Sungava IT Club

1

Gfootball-agent. Python

1

ashimdahal.github.io. CSS

1

TWL_DRCFS_30DaysPythonBootcamp. This is the repo for 3o Days Python Bootcamp by TWL and DRCFS

1

pydub. Manipulate audio with a simple and easy high level interface

1

Chicken-Recipe-Generator. A study of three different models that generate chicken recipies.

1

Dotfiles. dotfiles for my config

1

Nepali-Ho-Ni_NASA-spaceapp. Jupyter Notebook

1

Neural-Network-From-scratch-for-MNIST-handwritten-dataset. MNIST handwritten classifier neural network, it is a MLP

1

flappy-bird-using-pygame. a simple pygame demonstration with the evergreen game flappy bird

1

updated-school. CSS

1

linear_regression_gradient_descent. this is a simple linear regression model from scrarch without any dependencies except numpy

1

Live_post_load. Loads the posts live and also enters them to the server without refresh

1

Keras-side-practice. CNN form keras

1

hotel-website-for-hotel-Moontower-inn. html/css for hotel with backend

1

dogs-vs-cats-keras-kernel-edition. score in kaggle is 0.116 with this pretrained model , could be increased with increasing the number of epochs

1

football. Check out the new game server:

1

Sudoku-solver. Python

1
32
Apply