sitthiwatdam Goto Github PK

a6_student_layer_initialization

Exploring the Student Layers Initialization via distillation using BERT Huggingface

a7_ait_gpt_chatbot

a8_instruction_tuning_with_hf-alpacaeval

backorderprediction

Implementation of a backorder prediction system, aimed at forecasting instances where products may become temporarily unavailable due to insufficient stock levels.

car-price-prediction-from-scratch

Chaky Company manufactures cars but struggles with pricing them effectively. The solution involves utilizing a machine learning model trained on car price datasets to develop a straightforward web-based car price prediction system.

car-price-prediction-linear-regression-regularization-momentum-from-scratch

Using the Car Price dataset but with the code from scratch, implement more regularization and momentum techniques on top of linear regression, perform some experiments using ML flow, and lastly deploy the docker to a prepared virtual machine.

car-price-prediction-logistic-regression

Using the Car Price dataset but treating the problem as classification. This work implements Logistic Regression and adds some more features on top, perform some experiments using ML flow, and lastly deploy to docker to a prepared virtual machine.

en_th_translator

Experimenting with different types of attention mechanisms, including general attention, multiplicative attention, and additive attention, to assess their efficacy in translating between Thai and English languages.

engine_search

Developing a search engine involves practical information retrieval in natural language processing. Implemented word embedding techniques to enhance the search capabilities of the engine. Additionally, creating a user-friendly web interface is crucial for accessibility and usability.

mathplot-vqa

resume_parser

Creating a resume parser. Users will upload their resume in PDF format, and the system will output a list of unique skills and education details in order.

sentence_bert

This project is emphasis on leveraging text embeddings and capturing semantic similarity using a powerful encoder like BERT.

text_generation

Focus on building a language model using a text dataset and training a model that can generate coherent and contextually relevant text based on a given input.