 lars
|
87233e0f56
|
Edit predictions
|
2025-12-16 18:52:40 +01:00 |
|
 Lars Bogner
|
b2d8ba1810
|
edit predit
|
2025-12-16 18:52:14 +01:00 |
|
|
Lars Bogner
|
565c8e4855
|
Run predictions
|
2025-12-16 18:41:11 +01:00 |
|
|
lars
|
e015da6334
|
Write report
|
2025-12-16 18:24:44 +01:00 |
|
|
Lars Bogner
|
f351bd66dd
|
Fix res
|
2025-11-27 20:11:31 +01:00 |
|
|
Lars Bogner
|
f384ce0556
|
Run 4$
|
2025-11-27 20:07:42 +01:00 |
|
|
lars
|
070b6b4070
|
Add mode 4
|
2025-11-27 17:09:51 +01:00 |
|
|
Lars Bogner
|
4fa5afb060
|
Run scan 3
|
2025-11-27 16:48:10 +01:00 |
|
|
Lars Bogner
|
7d50637396
|
Increase epochs for mode 3
|
2025-11-27 10:25:11 +01:00 |
|
|
lars
|
51094364b2
|
Implement nan failsafe for training
|
2025-11-27 10:24:02 +01:00 |
|
|
lars
|
3931c14791
|
Revert unnecessary prints
|
2025-11-27 08:55:37 +01:00 |
|
|
lars
|
c120deb9cc
|
Found bug?!
|
2025-11-27 08:54:37 +01:00 |
|
|
lars
|
4be4bd79a2
|
Debugging of mode 3
|
2025-11-27 08:52:43 +01:00 |
|
|
lars
|
2cd518321f
|
Prepare run 3
|
2025-11-27 08:41:43 +01:00 |
|
|
Lars Bogner
|
ed13aeb822
|
Finish run 2
|
2025-11-27 08:34:08 +01:00 |
|
|
lars
|
5951475586
|
New hyperparamter scan results
|
2025-11-25 22:38:54 +01:00 |
|
|
lars
|
5fb29df06c
|
Limit batch size for long context
|
2025-11-25 11:06:10 +01:00 |
|
|
lars
|
8098d6282d
|
Fix to cuda error
|
2025-11-25 10:06:46 +01:00 |
|
|
lars
|
c0b2545671
|
Add more hyperparameter scan modes
|
2025-11-25 09:38:25 +01:00 |
|
|
lars
|
b012a86768
|
Fix bug in steps for dataset
|
2025-11-23 20:59:42 +01:00 |
|
|
lars
|
30e21eb5e9
|
Add validation dataset to hyperparameter scan
|
2025-11-23 13:02:22 +01:00 |
|
|
lars
|
a77fd08c5d
|
Add some finess to the hyperparameter scan
|
2025-11-23 12:45:39 +01:00 |
|
|
lars
|
b47bc1e297
|
File for hyperparameter_scan
|
2025-11-23 12:03:03 +01:00 |
|
|
lars
|
7960c05bef
|
Add GRU as potential RNN layer
|
2025-11-23 11:54:11 +01:00 |
|
|
lars
|
8a153691b5
|
Move to tanh activation and fix bug in altitude loss term
|
2025-11-21 15:15:23 +01:00 |
|
|
lars
|
b2ce697abe
|
Allow saving of datasets
|
2025-11-21 09:24:14 +01:00 |
|
|
lars
|
056bc94b16
|
Modify models
|
2025-11-21 09:08:31 +01:00 |
|
|
lars
|
0ca64a6687
|
Streamline model for better training
|
2025-11-21 08:33:44 +01:00 |
|
|
lars
|
de28c4a880
|
Fix bug with device change
|
2025-11-21 08:09:36 +01:00 |
|
|
lars
|
240ed75e3a
|
Add device to models
|
2025-11-21 08:06:08 +01:00 |
|
|
lars
|
681a9974b6
|
Fix dtypes in df loading
|
2025-11-21 07:44:31 +01:00 |
|
|
lars
|
5f8493d59e
|
New ideas
|
2025-11-21 07:16:05 +01:00 |
|
|
Lars Bogner
|
1538b0acf7
|
with '#' will be ignored, and an empty message aborts the commit.
Minor changes, run autoencoder on ml node
|
2025-11-20 14:15:06 +01:00 |
|
|
lars
|
4b071e68b5
|
Upload current version
|
2025-11-20 12:36:30 +01:00 |
|
|
lars
|
dfdd200867
|
Succesful data exploration
|
2025-11-19 11:57:59 +01:00 |
|
|
lars
|
3e47dcd52f
|
Limit number of threads in processing step to balance system load
|
2025-11-12 17:53:34 +01:00 |
|
|
lars
|
1212c2deac
|
Use async for data downloading
|
2025-11-12 14:23:31 +01:00 |
|
|
lars
|
bc6d436b85
|
v1 of data fetching
|
2025-11-12 12:35:45 +01:00 |
|