Checkin 2 Reflection:
Introduction: This can be copied from the proposal. For our final project, we plan on implementing the following paper that describes an NLP model in pytorch using tensorflow. This model essentially takes in different English tweets and outputs the suicidal intent of the tweet (Suicidal Intent Present or Suicidal Intent Absent) based on the content presented and past history of user tweets. Suicide is a huge problem in today’s world, and this paper attempts to use social media historical activity to determine suicidal ideation and provide preliminary screening of suicidal risk. This is a classification problem because we are trying to distinguish between tweets that demonstrate suicidal intent and those that don’t.
Challenges: What has been the hardest part of the project you’ve encountered so far? So far, most of the work we have done has been centered around converting the code from PyTorch to TensorFlow. This has mostly been smooth, except for some parts of the code that were substantially different in Torch. For example, the Torch function for binary cross entropy loss had an additional weights parameter that TensorFlow didn’t have, so we needed to write separate code to handle that. Furthermore, we haven’t run everything yet, but we predict the hardest part will be debugging the code after we initially run everything.
Insights: Are there any concrete results you can show at this point? How is your model performing compared with expectations? We are not yet at the point of running everything yet, but we are very nearly there. We do have most of the code converted into TensorFlow, but after that we need to run everything, and we hope the model will perform well after some debugging. We were able to run all the PyTorch code in Google Colab and it worked seamlessly.
Plan: Are you on track with your project? What do you need to dedicate more time to? What are you thinking of changing, if anything? We are on track with this project, and after finishing the conversions we will start running all the code very shortly. We think that after running everything, we will need to spend a lot of time on debugging and that will be our main time-consuming activity. Other than that, we don’t think there’s anything we want to change, unless the model is very inaccurate and then we might change some hyperparameters.
Log in or sign up for Devpost to join the conversation.