natural language process 2

Please follow the instructions: - We aim to build a model that generates accurate topics/headlines of articles. This task can be seen as a text summarization task. You have to train a dataset i.e, NEWS. When we run the code, it should do a training on the dataset. When the training is done, it will generate a text file that includes all the output results ”news headlines from dataset thats been trained” Users don’t have to input at all, they only need to run. And then the text file will be generated giving the output results. - There's no limitation whatsoever. You can use open-source but you MUST implement some components to show that you worked in this project. But I prefer if you code from scratch without having to install libraries. If you used some open sources, please share the license & sources. Also, you have to address what modifications and adjustments you have made clearly. - Some of the preferred approach methods (if applicable to our idea): word embedding, word2vec, cbow, skip gram, hierarchical softmax, negative sampling, fastText, doc2vec, subword embedding. - What I expect you to do for this work: 1- writing report (idea introduction, how to approach with some figures/ architecture/data, components to be used for coding… etc. 2- coding & training (including dataset). [Important] - I expect you to walk me through setting up everything and how to train - The project is done after training the dataset. No need for application nor website. - I don’t have a specific dataset. You could use anything. - We expect a quick response & update, so please give us a regular update about the project.