feat: new nvidia and pytorch devcon post

_posts/2018-04-09-nvidia-dli-workshop.md

@@ -15,7 +15,7 @@ In this half-day Image Classification with DIGITS workshop, you will learn to
 
 Feel free to sign up [here](https://www.eventbrite.sg/e/teaching-you-to-solve-problems-with-deep-learning-by-nvidia-tickets-45464548689).
 
-![NVIDIA Poster](https://scontent-sit4-1.xx.fbcdn.net/v/t1.0-9/31301576_1895254373882033_4742497653213755937_n.jpg?_nc_cat=0&oh=470d5402c4531357ebacd8fbf0bc96d6&oe=5B507B74)
+![NVIDIA Poster](https://res.cloudinary.com/ritchieng/image/upload/v1538729603/ritchieng.com/nus_nvidia/nus_nvidia_dli_2018_1.jpg)
 
 Cheers,
 <br />Ritchie

_posts/2018-04-13-nvidia-dli-workshop-postmort.md

@@ -13,12 +13,7 @@ In this half-day Image Classification with DIGITS workshop, participants learnt
 - walk through the process of data preparation, model definition, model training and troubleshooting
 - use validation data to test and try different strategies for improving model performance using GPUs.
 
-It was a full house! Here are some recap pictures taken by helpful NVIDIA student helpers. 
-
-![](https://scontent.fsin1-1.fna.fbcdn.net/v/t1.0-9/32231597_162974757706334_5166329560932810752_n.jpg?_nc_cat=0&oh=ea89df34f632aa41ec9ad897bbbdae97&oe=5B527F1E)
-![](https://scontent.fsin1-1.fna.fbcdn.net/v/t1.0-9/32247367_162974781039665_6963445797153144832_n.jpg?_nc_cat=0&oh=b8acb791a77a5ac96b62432351f22082&oe=5B53A182)
-![](https://scontent.fsin1-1.fna.fbcdn.net/v/t1.0-9/32423168_162974784372998_2065337970352717824_n.jpg?_nc_cat=0&oh=10a6bb63f6fb277ed6c0a170fd9fb8f6&oe=5B8A156E)
-![](https://scontent.fsin1-1.fna.fbcdn.net/v/t1.0-9/32314142_162974774372999_3268960829449437184_n.jpg?_nc_cat=0&oh=06b88ab3640784e328a295b5eefe0f05&oe=5B840736)
+It was a full house! 
 
 Cheers,
 <br />Ritchie

_posts/2018-05-29-nvidia-dli-workshop2.md

@@ -15,7 +15,7 @@ Many problems have established deep learning solutions, but sometimes the proble
 
 Feel free to sign up [here](https://www.facebook.com/nus.nvidia/posts/166945523975924).
 
-![NVIDIA Poster](https://scontent-sit4-1.xx.fbcdn.net/v/t1.0-9/33900380_166945510642592_8807125676499927040_n.jpg?_nc_cat=0&oh=3a182ce62000fff8616d20291b0fceda&oe=5B7BD57D)
+![NVIDIA Poster](https://res.cloudinary.com/ritchieng/image/upload/v1538729603/ritchieng.com/nus_nvidia/nus_nvidia_dli_2018_2.jpg)
 
 Cheers,
 <br />Ritchie

_posts/2018-10-02-pytorch-devcon.md

@@ -0,0 +1,20 @@
+---
+title:  "Recap of 2018 PyTorch Developer Conference"
+published: true
+permalink: /pytorch/pytorch-devcon-2018
+summary: "Invited by PyTorch to PyTorch's first Devcon!"
+tags: [news, machine_learning]
+---
+
+Proud of how far PyTorch has come. Thanks Soumith Chintala, it is a great journey and it just started. And thanks to Adam Paszke too!
+
+Finally got to catch up with Soumith Chintala, Alfredo Canziani and Marek Bardoński and found Andrej Karpathy! Back to Singapore tonight 🤗
+
+Here are some photos of my trip! 
+
+![Soumith Chintala](https://res.cloudinary.com/ritchieng/image/upload/v1538729876/ritchieng.com/pytorch_devcon_2018/IMG_2078_02-10-2018-21-00-43.jpg)
+![Alfredo Canziani](https://res.cloudinary.com/ritchieng/image/upload/v1538729876/ritchieng.com/pytorch_devcon_2018/IMG_2081_02-10-2018-21-00-01.jpg)
+![Andrej Karpathy](https://res.cloudinary.com/ritchieng/image/upload/v1538729876/ritchieng.com/pytorch_devcon_2018/IMG_2075_02-10-2018-21-02-31.jpg)
+
+Cheers,
+<br />Ritchie

index.md

@@ -27,7 +27,8 @@ These are some of the courses I created that will gradually build up your deep l
 
 ## Upcoming Talks/Workshops
 PyTorch Developer Conference, Facebook, San Francisco, USA, October 2018
-<br />Deep Learning for Financial Time Series, NVIDIA, Singapore, October 2018 (more details coming soon)
+<br />Explainable Deep Learning, Nanjing, China, November 2018
+<br />Deep Learning for Financial Time Series, NVIDIA, Singapore, December 2018 (more details coming soon)
 
 ## Key Papers
 [Neural Optimizers with Hypergradients for Tuning Parameter-Wise Learning Rates, ICML 2017](https://sites.google.com/site/automl2017icml/accepted-papers/AutoML_2017_paper_7.pdf)