Yassine commited on
Commit
f0dda28
1 Parent(s): d74f8b9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -0
README.md CHANGED
@@ -13,6 +13,18 @@ A VQ-VAE [1,2] was used to heavily compress each frame into 128 "tokens" of 10 b
13
 
14
  Note that the compressor is extremely lossy on purpose. It makes the dataset smaller and easy to play with (train GPT with large context size, fast autoregressive generation, etc.). We might extend the dataset to a less lossy version when we see fit.
15
 
 
 
 
 
 
 
 
 
 
 
 
 
16
  # References
17
  [1] Van Den Oord, Aaron, and Oriol Vinyals. "Neural discrete representation learning." Advances in neural information processing systems 30 (2017).
18
 
 
13
 
14
  Note that the compressor is extremely lossy on purpose. It makes the dataset smaller and easy to play with (train GPT with large context size, fast autoregressive generation, etc.). We might extend the dataset to a less lossy version when we see fit.
15
 
16
+ <video title="source" controls>
17
+ <source src="https://github.com/commaai/commavq/assets/29985433/91894bf7-592b-4204-b3f2-3e805984045c" type="video/mp4">
18
+ </video>
19
+
20
+ <video title="compressed" controls>
21
+ <source src="https://github.com/commaai/commavq/assets/29985433/3a799ac8-781e-461c-bf14-c15cea42b985" type="video/mp4">
22
+ </video>
23
+
24
+ <video title="imagined" controls>
25
+ <source src="https://github.com/commaai/commavq/assets/29985433/f6f7699b-b6cb-4f9c-80c9-8e00d75fbfae" type="video/mp4">
26
+ </video>
27
+
28
  # References
29
  [1] Van Den Oord, Aaron, and Oriol Vinyals. "Neural discrete representation learning." Advances in neural information processing systems 30 (2017).
30