Merge remote-tracking branch 'origin/main'

schedule.md

@@ -106,6 +106,18 @@ schedule:
   - date: Oct.<br>16, 11
     topics:
       - name: Introduction to Huggingface & Kaggle Challenge (Notebook)
+    readings:
+      - name: Training a causal language model from scratch (practical guides)
+        url: https://huggingface.co/learn/nlp-course/chapter7/6?fw=pt
+    references:
+      - name: transfer learning (T5)
+        url: https://arxiv.org/abs/1910.10683
+      - name: auto-regressive or causal language model (GPT-2)
+        url: https://d4mucfpksywv.cloudfront.net/better-language-models/language_models_are_unsupervised_multitask_learners.pdf
+      - name: Transformers library
+        url: https://huggingface.co/docs/transformers/index
+      - name: Accelerate library
+        url: https://huggingface.co/docs/accelerate/index
     homeworks:
       - name: HW2 (to be announced)
   #        url: https://www.dropbox.com/s/j2w4cpq14jypkbe/HW2.pdf?dl=0
@@ -161,28 +173,51 @@ schedule:
       - name: Generative Models I
     #        url: https://www.dropbox.com/s/nf4ohrqjqg7rb66/Week10-Graph-part2.pdf?dl=0
     readings:
-      - name: GAN
-        url: https://arxiv.org/abs/1406.2661
+      - name: Evidence Lower Bound ELBO — What & Why
+        url: https://yunfanj.com/blog/2021/01/11/ELBO.html
+      - name: From Autoencoder to Beta-VAE
+        url: https://lilianweng.github.io/posts/2018-08-12-vae/
+      - name: From GAN to WGAN
+        url: https://lilianweng.github.io/posts/2017-08-20-gan/
+      - name: Understanding Failure Modes of GAN Training
+        url: https://medium.com/game-of-bits/understanding-failure-modes-of-gan-training-eae62dbcf1dd
+    references:
       - name: VAE
         url: https://arxiv.org/abs/1312.6114
-    references:
-      - name: TBD
+      - name: GAN
+        url: https://arxiv.org/abs/1406.2661
   - date: Nov.<br>18, 15
     topics:
       - name: Diffusion, text-to-image generation
     #        url: https://www.dropbox.com/s/nf4ohrqjqg7rb66/Week10-Graph-part2.pdf?dl=0
     readings:
-      - name: DALL·E3
-        url: https://cdn.openai.com/papers/dall-e-3.pdf
+      - name: Hands-on Notebooks "Diffusion Models from Scratch"
+        url: https://colab.research.google.com/github/huggingface/diffusion-models-class/blob/main/unit1/02_diffusion_models_from_scratch.ipynb
+      - name: How to Use DALL-E 3 Tips, Examples, and Features
+        url: https://www.datacamp.com/tutorial/an-introduction-to-dalle3
       - name: Sora
         url: https://openai.com/index/sora/
       - name: Kling
         url: https://kling.kuaishou.com/en
+    references:
+      - name: Diffusion Probabilistic Model
+        url: https://arxiv.org/abs/2006.11239
+      - name: DALL·E3
+        url: https://cdn.openai.com/papers/dall-e-3.pdf
+      - name: Sora
+        url: https://arxiv.org/pdf/2402.17177
   - date: Nov.<br>25, 22
     topics:
       - name: Graph Representation Learning
     #        url: https://www.dropbox.com/s/3e09x5i9wyn8q3c/Week9-Graph-part1.pdf?dl=0
     readings:
+      - name: Design Space for Graph Neural Networks
+        url: https://arxiv.org/pdf/2011.08843
+      - name: A Comprehensive Survey on Graph Neural Networks
+        url: https://arxiv.org/abs/1901.00596
+      - name: Graph Neural Networks Implementation Tutorial 
+        url: https://uvadlc-notebooks.readthedocs.io/en/latest/tutorial_notebooks/tutorial7/GNN_overview.html
+    references:
       - name: DeepWalk
         url: https://arxiv.org/pdf/1403.6652
       - name: LINE