Graph Machine Learning
AnemoI
English
hcookie129 commited on
Commit
36f414b
1 Parent(s): 3a74f23

Rework to assets dir

Browse files
README.md CHANGED
@@ -16,7 +16,7 @@ Here, we introduce the **Artificial Intelligence Forecasting System (AIFS)**, a
16
  model developed by the European Centre for Medium-Range Weather Forecasts (ECMWF).
17
 
18
  <div style="display: flex; justify-content: center;">
19
- <img src="aifs_10days.gif" alt="AIFS 10 days Forecast" style="width: 50%;"/>
20
  </div>
21
 
22
 
@@ -34,8 +34,8 @@ AIFS is based on a graph neural network (GNN) encoder and decoder, and a sliding
34
  and is trained on ECMWF’s ERA5 re-analysis and ECMWF’s operational numerical weather prediction (NWP) analyses.
35
 
36
  <div style="display: flex; justify-content: center;">
37
- <img src="encoder_graph.jpeg" alt="Encoder graph" style="width: 50%;"/>
38
- <img src="decoder_graph.jpeg" alt="Decoder graph" style="width: 50%;"/>
39
  </div>
40
 
41
  It has a flexible and modular design and supports several levels of parallelism to enable training on
@@ -94,7 +94,7 @@ AIFS is trained to produce 6-hour forecasts. It receives as input a representati
94
  at \\(t_{−6h}\\), \\(t_{0}\\), and then forecasts the state at time \\(t_{+6h}\\).
95
 
96
  <div style="display: flex; justify-content: center;">
97
- <img src="aifs_diagram.png" alt="AIFS 2m Temperature" style="width: 80%;"/>
98
  </div>
99
 
100
  The full list of input and output fields is shown below:
@@ -158,7 +158,7 @@ of the metrics, such as ACC (ccaf), RMSE (rmsef) and forecast activity (standard
158
  sdaf) can be found in e.g Ben Bouallegue et al. ` [2024].
159
 
160
  <div style="display: flex; justify-content: center;">
161
- <img src="aifs_v021_scorecard.png" alt="Scorecard comparing forecast scores of AIFS versus IFS (2022)" style="width: 80%;"/>
162
  </div>
163
 
164
 
@@ -223,4 +223,4 @@ Lang, S., Alexe, M., Chantry, M., Dramsch, J., Pinault, F., Raoult, B., ... & Ra
223
 
224
  ## More Information
225
 
226
- [More Information Needed](https://arxiv.org/pdf/2406.01465)
 
16
  model developed by the European Centre for Medium-Range Weather Forecasts (ECMWF).
17
 
18
  <div style="display: flex; justify-content: center;">
19
+ <img src="assets/aifs_10days.gif" alt="AIFS 10 days Forecast" style="width: 50%;"/>
20
  </div>
21
 
22
 
 
34
  and is trained on ECMWF’s ERA5 re-analysis and ECMWF’s operational numerical weather prediction (NWP) analyses.
35
 
36
  <div style="display: flex; justify-content: center;">
37
+ <img src="assets/encoder_graph.jpeg" alt="Encoder graph" style="width: 50%;"/>
38
+ <img src="assets/decoder_graph.jpeg" alt="Decoder graph" style="width: 50%;"/>
39
  </div>
40
 
41
  It has a flexible and modular design and supports several levels of parallelism to enable training on
 
94
  at \\(t_{−6h}\\), \\(t_{0}\\), and then forecasts the state at time \\(t_{+6h}\\).
95
 
96
  <div style="display: flex; justify-content: center;">
97
+ <img src="assets/aifs_diagram.png" alt="AIFS 2m Temperature" style="width: 80%;"/>
98
  </div>
99
 
100
  The full list of input and output fields is shown below:
 
158
  sdaf) can be found in e.g Ben Bouallegue et al. ` [2024].
159
 
160
  <div style="display: flex; justify-content: center;">
161
+ <img src="assets/aifs_v021_scorecard.png" alt="Scorecard comparing forecast scores of AIFS versus IFS (2022)" style="width: 80%;"/>
162
  </div>
163
 
164
 
 
223
 
224
  ## More Information
225
 
226
+ [Find the paper here](https://arxiv.org/pdf/2406.01465)
2t_aifs_v021.png → assets/2t_aifs_v021.png RENAMED
File without changes
aifs_10days.gif → assets/aifs_10days.gif RENAMED
File without changes
aifs_diagram.png → assets/aifs_diagram.png RENAMED
File without changes
aifs_v021_scorecard.png → assets/aifs_v021_scorecard.png RENAMED
File without changes
decoder_graph.jpeg → assets/decoder_graph.jpeg RENAMED
File without changes
encoder_graph.jpeg → assets/encoder_graph.jpeg RENAMED
File without changes