Ahan isn't taking new orders for this project right now. Here are some similar projects to explore.
You will get a Transformer based NLP Text Summarisation tool
Project details
I compare the BART model from Facebook, Pegasus from google currently available for summarisation tasks and give out good results as well. More importantly, GPT-2 and XLNet also do a good job for summarisation task but ideally GPT-2 and XLNet are mostly used for text generation, so for example if you use a text prompt as input, it often is used to add onto this text and increase the generated content and are not readily used for summarisation. To use it for summarisation, a little bit of fine-tuning the model is required but there are workarounds, eventually you can do it by reducing the max_length of the total generated tokens but in most cases, it picks most important sentences from the prompt as output without rephrasing the sentences into shorter and summarised form.
What's included
| Service Tiers |
Starter
$150
|
Standard
$250
|
Advanced
$400
|
|---|---|---|---|
| Delivery Time | 7 days | 5 days | 3 days |
Number of Revisions | 2 | 4 | 5 |
Number of Model Variations | 2 | 4 | 6 |
Model Validation/Testing | |||
Model Documentation | |||
Data Source Connectivity | - | - | |
Source Code | - |
Optional add-ons
You can add these on the next page.
Fast Delivery
+$50
Additional Graph/Chart
(+ 1 Day)
+$50
Choose your own model
(+ 1 Day)
+$50About Ahan
Machine Learning & Deep Learning Research | Data Scientist | ML Tutor
Bangalore, India - 10:57 am local time
Proficient in Natural Language Processing and Generation Tasks using GPT/GPT-2 Transformers. Fine-tuning NLP models is a job I'm good at and can help you in NLP and Data Visualisation tasks. I can also help you in building OpenNMT-PY/OpenNMT-TF Machine Translation models and help you host it as well using Streamlit and CTranslate2. I am proficient in building NLP models such as LSTM/GRU models. I've also been curating content for blogs (technical) for various technologies such as Chatbots, Image Processing, Semantic Segmentation, Deep Learning, and Data Science. I can also help you setup pipelines related to Text summarisation, Text generation, query reformulation and other NLP based pipelines. Proficient in building Huggingface based pipeline models along with deploying it using Heroku/Streamlit/Gradio and flask.
I have contributed to the following libraries:
- "Gensim": working on issues in the code for the word2vec library.
- “PyMC3”: for handling Bayesian modeling by Markov Chain Monte Carlo (MCMC);
- “TensorFlow 2.0/keras”: a later acquisition of mine, but a powerful one for dealing with Deep Learning;
- “statsmodels”: for traditional statistics (e.g., General Linear Models);
Few projects I worked on:
• Extractive and Abstractive Text Summarization using Transformers BERT, BART, T5, attention based seq2seq models
• Text Generation using GPT-3, Context Free Grammars and Markov Chain
• Keyword Extraction using text rank algorithms
• Face positioning application for ID cards operation using FLASK, HTML, CSS and Djinga
• Neural Image Captioning using CNN-RNN structure from Google Flicker 8k dataset integrated into Flask Web application as a stock photography website
• Chatbot using LSTM and tensorflow on movie conversation dataset
• Packman Game using BFS & A* algorithms and pygame module in python which is available in my introductory video!!
• Object Detection Using Fast RCNN, RCNN and YOLO on Zombies, birds, cars and tumors
• Image classification using CNN with data augmentation and transfer learning on CIFAR images
• Speech Recognition on Turkish, Hebrew and romanian data by fine tuning transformers wav2vec
• Arabic Poem Meter Classification from Acoustic Data using CNN
• Language and Gender Identification from voice records into German, Spanish and English using CNN and BiLSTM
• Topic/Concept Modeling using SVD/NMF decomposition
• Arabic tweets Sentiment Analysis using conv1d and LSTM with an overall accuracy of 94%
• Metamorphic Malware Detection from sequential malware and malware images using CNNs and LSTMs
• Speech Synthesis and Recognition using CNN-RNN architectures
• Normalizing Flows baseline architecture implementation from scratch on pytorch
• Neural Machine Translation using attention based LSTM seq2seq model from French to German
• Text Classification using neural networks, naive Bayes classifiers, Decision Tree Classifiers, random forests, etc
• Text correction recommendation system by grammar checking acceptability using T5
• A LOT OF SEARCH ENGINES USING TEXT SIMILARITY TECHNIQUES LIKE COSINE SIMILARITY and word2vec
• Neural Question Answering using Siamese Networks
• Hybrid Image from two or more images using MATLAB, Python, and OpenCV.
• Car palette detection and recognition system using opencv and machine learning models
Steps for completing your project
After purchasing the project, send requirements so Ahan can start the project.
Delivery time starts when Ahan receives requirements from you.
Ahan works on your project following the steps below.
Revisions may occur after the delivery date.
Creating an initial workflow of the project
Model Building