Share your V2 projects here

nchukaobah · July 4, 2020, 3:13am

Do you have a paper reference perharps? it sounds really interesting.

kofi · July 17, 2020, 11:47am

One of the further research questions in chapter 17 of the fastai book is to use the unfold function in PyTorch to create a CNN module and train a model with it. I tried my hands on it and I would be happy to hear your thoughts about it and how you think it could be made better.

github.com

asiedubrempong/fastai-experiments/blob/master/foundations/cnn_from_scratch.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Implementing a CNN module from scratch using the unfold function in PyTorch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai2.vision.all import *"
   ]
  },
  {
   "cell_type": "code",

This file has been truncated. show original

pierreguillou · July 19, 2020, 11:19am

Hello. Based on the Transformers tutorial of @sgugger, I published a “from English to any language” fine-tuning method of models based on generative English pre-trained transformers like GPT-2 using Hugging Face libraries (Tokenizers and Transformers) and fastai v2.

There are a medium post + notebook in github + model in Hugging face model hub.

As proof of concept, I fine-tuned a GPorTuguese-2 (Portuguese GPT-2 small), a language model for Portuguese text generation (and more NLP tasks…), from an English pre-trained GPT-2 small downloaded from Hugging Face Transformers library. Here are examples of what it can generate:

If you want to test it online (without running my notebook), you can thanks to the Hugging Face model hub at https://huggingface.co/pierreguillou/gpt2-small-portuguese

morgan · July 21, 2020, 10:29am

This is brilliant, thanks for sharing, so many great little tips and tricks!

Not sure if I missed it in the article but I’m curious if you think the Portuguese text is as good as the English text it can generate?

Also, is there much research on fine-tuning text generation models on different languages? I know cross-lingual models can help for translation, but I don’t think I had seen it for text generation before…

Nice work!

muellerzr · July 21, 2020, 4:05pm

I’ve just released an update to fastinference, and two new libraries:

fastinference_pytorch and fastinference_onnx

The goal of these two is to be lightweight modules to run your fastai models within a familiar API. Currently it just supports tabular models but vision and NLP are on the way! See below for a numpy example (there is zero fastai code being used )

Not to mention there is a speed boost here as well

For more, the documentation is available:

Note: To use this you must export your model from fastinference with learn.to_fastinference

pierreguillou · July 29, 2020, 11:55am

Hello @morgan. Well, it was the objective of my work

Yes, I think that a small generative model (like GP-2 small) fine-tuned from English to another language like Portuguese allows to get a working model with a relatively small fine-tuning dataset (I used a bit more than 1 GB from Portuguese Wikipedia).

Well, great if I opened up some research but I’m sure there was already.

About applying the same fine-tuning method on encoder transformer-based models like BERT (RoBERTa, ALBERT, etc.), I’m currently testing my method on your FastHugs code. It works very well (because you did a great work Morgan!) I will publish soon.

By the way, I have 2 questions on your code that I will publish in the FastHugs thread. Many thanks!

Brainkite · July 30, 2020, 10:22am

Hey guys, recently took the time to write about an experimental study I did few months back about Edge Detection models.

Also it’s on my brand new blog where I publish annotated research papers that I read.

jimmiemunyi · August 25, 2020, 4:31pm

My first ever two models. Just finished Chapter Two of the book

and

Santosh1 · August 26, 2020, 4:44am

vijish · August 26, 2020, 7:31am

I have been working on low light image enhancement using GAN and Perceptual/Feature loss. Got inspiration from Deoldify to try this up.
I will share the Github repo soon.

Feras · August 26, 2020, 8:19am

Simple crowd estimation using image regression on “Mall” data set.

https://www.kaggle.com/ferasoughali/crowd-estimation-fastai-v2

ilovescience · August 28, 2020, 3:22am

I have been researching the applications of CycleGANs (and related models) in pathology. I have written a paper that was presented at the ICML Computational Biology 2020 workshop:

I used fastai v2 and nbdev to accelerate my development. I collaborated with @ashaw and @danielvoconnor as part of the WAMRI initiative.

Recently, I have been working on refactoring my code into a package for unpaired image-to-image translation. Check it out here!:

Also, I will be presenting my research, talking about CycleGANs, and sharing code, TOMORROW August 28th 9am PDT:

Check it out!

joedockrill · August 28, 2020, 5:42am

17 and 2nd yr PhD??? Bro…

morgan · August 28, 2020, 7:56am

Amazing! Looking forward to this!

KevinB · August 28, 2020, 2:39pm

Is this going to be recorded?

ilovescience · August 28, 2020, 3:30pm

Yes it will be available afterwards as a YouTube video at the same link.

vijish · August 29, 2020, 3:56pm

Hi All,

I tried to generate Face sketch with Super resolution, Self attention and GAN. Here are some of the results that I got,Please have a look.

Thank you Jeremy for the wonderful course.

k7shreyas · August 30, 2020, 4:55pm

Made a very basic dog classifier based on the knowledge of Lessons 1 & 2. It can identify 17 different breeds of dogs. I used RESNET 101 architecture and 4 epochs with data set size of 150 images per breed to train the model with approx. 96% accuracy on validation set.

Screenshot from 2020-08-30 21-56-46 Screenshot from 2020-08-30 22-35-06

Find github repo here
Try it here(It’s bit slow maybe due to heavy architecture, please bear with it

Unplug_Charger · August 30, 2020, 5:58pm

Hello

I did the same for butterflies…

Lovkush · August 30, 2020, 7:21pm

Hello all! I wrote a blog post describing a little bit of playing around I did with the movie review / NLP sentiment analysis at the end of lesson 1. Some surprises were found, e.g., the model thought ‘I disliked the movie’ and ‘I hated the movie’ were positive reviews!