RuiLGTM commited on
Commit
7f8083b
·
verified ·
1 Parent(s): fc3b1cc

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +0 -16
README.md CHANGED
@@ -41,22 +41,6 @@ The synthetic training data is generated by the training code.
41
 
42
  The models train on this data together with synthetic series.
43
 
44
- <p align="center">
45
- <img class="dark:hidden"
46
- src="https://huggingface.co/datasets/abel-lab/Simons-Pretrain/resolve/main/assets/training-data.png" alt="Training data in observations"
47
- width="720" />
48
- <img class="hidden dark:block"
49
- src="https://huggingface.co/datasets/abel-lab/Simons-Pretrain/resolve/main/assets/training-data-dark.png" alt="Training data in observations"
50
- width="720" />
51
- </p>
52
-
53
- Compared with other models, such as
54
- [Toto 2.0](https://www.datadoghq.com/blog/ai/toto-2/),
55
- [TimesFM-3](https://research.google/blog/timesfm-3-a-zero-shot-foundation-model-for-multivariate-forecasting/)
56
- and Timer-S1, Simons is efficient with data and depends little on real-world data. Simons-297M trains
57
- on 19B real observations, 11.3B of them released here, and 800B synthetic ones that the training code
58
- generates. Toto 2.0 trains on 2.1T real and 2.9T synthetic data points; TimesFM-3 and Timer-S1
59
- state corpora of 1T or more.
60
 
61
  ## Usage
62
 
 
41
 
42
  The models train on this data together with synthetic series.
43
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
44
 
45
  ## Usage
46