There are 3 repositories under bleu-score topic.
Evaluation code for various unsupervised automated metrics for Natural Language Generation.
A neural network to generate captions for an image using CNN and RNN with BEAM Search.
A modular library built on top of Keras and TensorFlow to generate a caption in natural language for any input image.
The LSTM model generates captions for the input images after extracting features from pre-trained VGG-16 model. (Computer Vision, NLP, Deep Learning, Python)
A visual and interactive scoring environment for machine translation systems.
Deep CNN-LSTM for Generating Image Descriptions :smiling_imp:
PyTorch implementation of "Attention Is All You Need" by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, Illia Polosukhin
To evaluate machine translation, they use several methods, some of which we fully implemented
Scripts for an upcoming blog "Extractive vs. Abstractive Summarization" for RaRe Technologies.
BLEU Score in Rust
Abstractive text summarization done with the help of LSTMs using encoder-decoder model which was able to achieve accuracy of 77.27% on training set and cumulative BLEU-4 score of 0.8800 on test set.
An end-to-end image captioning project using a CNN encoder (ResNet-50) and LSTM decoder in PyTorch. Includes vocabulary building, preprocessing, training with BLEU evaluation, and inference. Generates natural language captions for images with saved metrics, model checkpoints, and visualization outputs.
Implementation for paper BLEU: a Method for Automatic Evaluation of Machine Translation
Semantic Message Extraction for Text Based Data With Deep Neural Nets
Machine learning tools for NLP programming.
Generate caption on images using CNN Encoder- LSTM Decoder structure
⚡ Seq2Seq model combines Attention mechanism
Creating RAG from Scratch . Creating RAG using the langchain. Creating RAG using llama indexing and Qdrant db
Repository containing the code to my bachelor thesis about Neural Machine Translation
This project aims to assist visually impaired individuals by providing a solution to convert images into spoken language. Leveraging deep learning and natural language processing, the system processes images, generates descriptive captions, and converts these captions into audio output.
A library for evaluating Retrieval-Augmented Generation (RAG) systems
Tensorflow implementation of "Show and Tell"
ViAG: A Novel Framework for Fine-tuning Answer Generation models ultilizing Encoder-Decoder and Decoder-only Transformers's architecture
Generate captions from images
Using Google Colab, we develop a NMT, language translator. Here, we do NMT to translate from English to Vietnamese.
In this project, we use a Deep Recurrent Architecture, which uses CNN (VGG-16 Net) pretrained on ImageNet to extract 4096-Dimensional image feature Vector and an LSTM which generates a caption from these feature vectors.
A CNN-LSTM model to generate a sentence/caption that describes the contents/scene of an image.
This project, developed at the Technion - Israel Institute of Technology, focuses on creating a custom dataset and implementing a Transformer model for German to English translation. The goal is to achieve high-quality translations with a minimum BLEU score of 35% on the validation set.
A benchmark of ChatGPT and some of its challengers on summarization task
Modern Eager TensorFlow implementation of Attention Is All You Need
Structured approach in AI and ML. Fundamentals and Advanced topics. RAG, Scoring & Profiling, LangChain & LangGraph, Certified Azure AI Engineer materials.
A model inspired from the famous Show and Tell Model is implemented for automatic image captioning.
A neural network to generate captions for an image using CNN and RNN with BEAM Search.