This repository contains results from my MSc. thesis on "Test Case Generation from User Stories using Generative AI Techniques with LLM Models." Each folder includes generated test cases in PDF, detailed metrics scores of data in Excel sheets, and visual graphs, offering a comprehensive view of the experiments in images folder and their outcomes.

Semantically consistent regularizer for zero-shot learning

This repository hosts the code for Jan Hadl's Master Thesis at TU Wien: GS-VQA, a zero-shot visual questions answering (VQA) pipeline that uses vision-language models (VLMs) for visual perception and answer-set programming (ASP) for symbolic reasoning.

This repository contains official implementation of the paper "Training-Free Zero-Shot Semantic Segmentation with LLM Refinement" (BMVC 2024).

zero-shot keyword spotting with KWS test dataset using ImageBind

Limit the use of end-to-end data for Speech Translation (by leveraging Automatic Speech Recognition and Machine Translation data instead) using zero-shot multilingual text translation techniques.

This repository highlights the LLMs reasoning capabilities of ✨ Mistral / LLaMA-3 / Phi-3 / Gemma / Flan-T5 / GPT-4o ✨ in Targeted Sentiment Analysis in Russian / Translated to English mass-media 📊

In this project, I created a simple Python using the HuggingFace's Transformers library which us an application that is answering "What is the capital of a country?".

Leveraging Codebook Knowledge with NLI and ChatGPT for Zero-Shot Political Relation Classification (ACL 2024)

Choice Fusion as Knowledge for Zero-Shot Dialogue State Tracking (ICASSP 2023)

Show Less, Instruct More: Enriching Prompts with Definitions and Guidelines for Zero-Shot NER

Multilingual Neural Machine Translation for Low Resource Languages - a Ph.D. Thesis.

First Zero-Shot Topic Classification for Persian

Fine-grained zero-shot or zero-shot learning

text-only training or language-free training for multimodal tasks (image/audio/video caption, retrieval, text2image)

An Instruction-tuned LLM for zero-shot NER on Italian

Unofficial Pytorch Implementation of StyleSpeech

SmartFew is your swiss knife for semi-supervised structuring of unlabeled data using Few Shot Learning.